So, I finally made up my mind: it's time to bid farewell to the internal combustion engine era and switch to a pure electric vehicle.

But once I dove in, I realized the car world is no longer in the era of comparing "genuine leather seats" and "sunroof size." In some car forums and video comment sections, I saw a very interesting saying: "Today's EVs are divided into 2K, 4K, and 8K versions."

At first, I thought this referred to the specifications of "TVs, fridges, and large sofas," since 2K, 4K, and 8K are terms usually mentioned when talking about video resolution or the displays of TVs, monitors, and phones. But upon closer inspection, I found that 4K and 8K are actually core indicators for electric vehicles.

Electric Vehicle Indicators

Undoubtedly, just like with monitors and video resolution, "K" stands for "Kilo" (thousand). Currently, many core hardware indicators for electric vehicles are targeting 1000 as a milestone:

Category 1: Physical Performance

• 1000 Horsepower: Once exclusive to supercars, speed has now become a cheap dividend of EVs.
• 1000 N·m Torque: The system is not just powerful; it can precisely control every vibration of the car body.
• 1000A Response Current: The "conduction speed" of the nerves, determining the quickness of system anticipation and hazard avoidance.
• 1000 Hz Suspension Adjustment: The chassis scans and adjusts damping 1000 times per second, letting the EV truly "walk on flat ground."

Category 2: Energy Supplement & Range

• 1000V Platform: Full-domain high-voltage architecture; charging officially enters the "second-level charge" era.
• 1000 km Range: Thoroughly erases the embarrassment of the "Electric Grandpa" label; even with a 50% discount, it's enough for cross-province travel.
• 1000kW Charging Power: The peak of megawatt-level charging infrastructure; as long as the pile is strong enough, charging is faster than refueling.
• 1000 Seconds to Full Charge: Regardless of extreme heat or cold, it regains full power within 1000 seconds (approx. 16 minutes).
• 1000 Lossless Cycles: After 1000 charge-discharge cycles, the battery health remains above 90%.

Category 3: Intelligent Soul

• 1000 TOPS Computing Power: The "brain capacity" supporting high-order smart driving in complex urban road conditions.
• 1000-Line Grade LiDAR: The "resolution" of the eyes, seeing distant dangers clearly even in heavy rain or dark nights.
• 1000 Mbps Communication Bandwidth: In-car gigabit-level "neural network," ensuring massive smart driving data stays online.

Category 4: Other Conditions

• 1000 Nit Peak Brightness: Even under direct high-noon sunlight, the map on the screen remains clear.
• 1000 Watt Audio Power: Driving 20+ speakers to achieve panoramic sound, turning the car interior into a theater.
• 1000 mm Wading Depth: The "tough guy" confidence of an EV; it becomes your life-saving boat during urban waterlogging.

Looking closely at these indicators, many are actually the basic needs of ordinary people. They represent the "final straw" on the psychological defense line.

For instance, 1000 km range is the final hurdle that fuel car owners long for. A 1000 km range means that even with a significant discount, it's enough for a cross-province trip in most provinces. And once these indicators cross the "1000" mark, the EV is no longer a "precision electronic toy" that needs careful pampering, but an industrial product truly capable of competing with fuel cars in hardcore survival capabilities.

How to View 4K/8K Cars

Based on the coverage of these "K-level indicators," the 1K/2K, 4K, and 8K standards popular in the car circle are essentially "entry tickets" for different tiers.

1K / 2K Version

Not everyone needs to buy a "maxed-out account" worth hundreds of thousands. In the 100,000 to 150,000 yuan price range, car companies are very smart; they don't need to pursue perfection in everything, but rather play the "long-board effect."

For example, some cars focus on 1000km range. They might lack LiDAR and have an average interior, but they solve the owner's "long-distance anxiety." Others might offer a 1000V platform and 1000A response current simultaneously. While not luxurious, they charge incredibly fast and start with extreme power. Like a "specialized student," as long as one or two of these long boards hit your core needs, it is your king of cost-performance.

4K Version

If your budget reaches the 200,000 to 300,000 yuan level, it's no longer about choosing a long board, but looking at "comprehensive quality." The 4K version usually refers to meeting or approaching the 1000-level standard across four physical dimensions: voltage, range, horsepower, and power. A car that reaches 4K is already in the global first tier of mechanical quality. This is why many people comment on short video platforms that in the 2026 car market, 4K is merely the entry threshold for flagship cars.

8K Version

If 4K solves whether the "body" is strong, then 8K solves whether the "soul" is smart. This adds four intelligence dimensions—computing power, radar, torque, and current—on top of the four physical indicators. If all 8 indicators are maxed out, it is undoubtedly a top-tier "all-rounder," achieving the leap from "transportation tool" to "intelligent living entity."

Of course, overall, no car manufacturers are currently directly promoting concepts like 4K or 8K, because after looking around, there seem to be very few cars that can achieve all these indicators simultaneously. But this doesn't mean the concept won't become popular, as it accurately hits the technological trajectory of "electric vehicles."

As long as it is a technology product, it will inevitably be quantified and parameterized.

In the past, when buying a car, everyone only looked at displacement, horsepower, and torque—the physical mechanical aesthetics of the internal combustion engine. Furthermore, parameters back then were relatively "sluggish"; a V6 engine could be sold for ten years without the power specs needing to change.

But electric vehicles are completely different; they are more like smartphones or computers on wheels. Now when you buy a car, you look at computing power (TOPS), architecture voltage (V), charging rate (C), and LiDAR line count. The iteration of these parameters is measured in "months," not "decades."

The emergence of these "K-level indicators" is essentially us trying to find a unified set of weights and measures amidst chaotic marketing jargon. If a car can only achieve 1K or 2K, it remains a good tool that solves your actual pain points through single long boards (such as extreme range or high-speed charging). If a car can reach 4K or even 8K, it is no longer just a commuting tool, but an evolving intelligent terminal. This quantitative standard, though harsh, is a good thing for consumers—it brings an end to the era of charging premiums based on "brand storytelling."

However, the process of choosing a car has been far more agonizing than I imagined. I meticulously studied almost every mainstream pure EV priced from tens of thousands to three or four hundred thousand RMB, watched countless video analyses and spec comparisons, yet I still had no idea where to start. Later, my wife suggested that we simply sell our current car first and buy a new one with whatever we get for it. After going through that ordeal, I realized that selling a car is also a massive pitfall.

While dealing with car dealers to sell the car, my wife mentioned that her colleague's Mazda EZ-6 looked quite nice, so I went online to check it out. To be honest, when I was looking for cars before, I hadn't paid any attention to foreign brands other than Tesla. As a result of my search, I stumbled upon some Quora pages. Once I started browsing, I couldn't stop; I ended up logging into an account I hadn't used in 10 years and wrote a few answers.

How will Jaguar convince its traditional petrol-head customers to buy into a completely electric, ultra-luxury future?

My wife recently started looking for a new car. She was initially drawn to the Mazda EZ-6 because a colleague of hers bought one. Its design is undeniably stunning—sharp, futuristic, and sleek.

As a car enthusiast, I was confused. Mazda? Renowned for the "Jinba Ittai" (horse and rider as one) handling of the MX-5 and RX-7? Since when did they become a leader in EVs?

I did some digging, and the "Oho!" moment arrived: The Mazda EZ-6 is essentially a rebadged Deepal SL03. Mazda, despite its heritage, lacked the R&D capability to build a competitive EV on its own. They had to turn to their Chinese partner, Changan Automobile, who assigned their sub-brand, Deepal, to handle the engineering.

The irony is palpable:

• The Mazda EZ-6 sold about 19,182 units in China last year (ranked #155 in NEV sales).
• Its "twin," the Deepal SL03, only sold 14,000 units.
• However, the Deepal SL03 is just the baseline. Its siblings, the Deepal S05 and S07, are both in the Top 100, selling 102,000 and 61,000 units respectively.

If you ask me how a Mazda EZ-6 feels, I’d say it feels "weird." It’s a Japanese body with a Chinese heart. But here is the cold truth: If Mazda didn't have this car, they wouldn't even be at the "New Energy" table. And if you aren't at the table in 2026, you don't have a seat in the future.

Now, look at Jaguar.

Jaguar is facing the exact same existential crisis. They’ve announced a pivot to an "Ultra-Luxury" all-electric brand, with cars starting at $130,000 to $150,000. They want to compete with Bentley and Aston Martin.

But where is the technology coming from? In June 2024, JLR signed a strategic deal to use Chery’s E0X electric platform (the same one powering the Exeed Exlantix and Huawei’s Luxeed). Just like Mazda, Jaguar is effectively "borrowing a soul" from a Chinese manufacturer to stay alive.

How competitive is the Chinese market compared to the US/Europe?

Five years ago, my wife purchased a Lexus ES200 for 360,000 RMB (approximately $50,000 USD). This is a somewhat "peculiar" model, primarily sold only in China and a few parts of Eastern Europe. It is equipped with a 2.0L naturally aspirated engine, achieving 0-100 km/h in a sluggish 12 seconds. It was sarcastically nicknamed "The Highway Lightning" on Chinese social media for being notoriously slow. Why didn't we buy the more powerful ES300h hybrid version back then? Because it was a seller's market; the ES300h cost 450,000 RMB, and dealers even demanded substantial markups (additional fees) on top of that.

Fast forward to today, the car has done 70,000 kilometers, and my wife wants to trade it in. Before we look at new options, let's talk about its residual value. I finally sold it for 140,000 RMB, which is only 38.9% of its original purchase price. The core reason for this drastic depreciation is simple: a brand-new Lexus ES200 in China now costs only about 245,000 RMB (including taxes and insurance) at the dealership—a 32% drop in price from five years ago.

For the new car, my wife set a budget of 200,000 to 300,000 RMB ($28,000 - $42,000 USD). If we are looking for a sedan (or coupe) in this price range today, the sheer number of options is dizzying:

• The Trendsetters & Performance Kings: We have the Xiaomi SU7 (the first car from the smartphone giant, an instant blockbuster), the Zeekr 007, and the Zeekr 001 (Geely's premium EV performance benchmarks known for driving dynamics).
• The Pioneers of Smart Driving & Futuristic Design: We have the Xpeng P7i (leader in autonomous driving), the Avatr 12 (a futuristic coupe co-developed by Changan, Huawei, and CATL), the Stelato S7 (Huawei's first premium sedan in its Smart Selection business), and the IM L6 (featuring advanced digital chassis and rear-wheel steering technology).
• New Energy Counterattacks from Established Giants: We see the BYD Han EV (the long-standing champion of Chinese C-segment EV sales), the Lynk & Co Z10 (the brand's first flagship pure EV), the Denza Z9GT (BYD's premium D-segment GT coupe), and the NIO ET5T (arguably the world's most beautiful touring car, supporting 3-minute battery swapping).
• More Value-Oriented Choices: Options like the Arcfox Alpha S (boasting Magna manufacturing quality), the Avatr 06, and the ONVO L60 (NIO's sub-brand focusing on efficiency and family use).

If we switched our search to family SUVs, the choices would be even more numerous, as Chinese consumers generally prefer SUVs. However, what truly causes "physical suffocation" for traditional brands is the explosion of PHEVs and EREVs (Extended-Range EVs). In this segment, range anxiety is a thing of the past. Models like the BYD Seal 06 DM-i and Qin L have combined ranges exceeding 2,100 km (1,300 miles).

But the most "dangerous" competitors are Li Auto and the Huawei Alliance:

• Li Auto: Dominating the family SUV market with its L6, L7, L8, and L9 series. Their "Fridge-TV-Sofa" luxury combined with range-extending technology has wiped out the premium once enjoyed by foreign luxury SUVs.
• Huawei and its "Five Realms" Alliance (HIMA): Huawei infuses its top-tier IT DNA into traditional manufacturers through five distinct brands:
  • AITO (Wanjie): High-tech SUVs (M7, M9) co-developed with Seres.
  • Luxeed (Zhijie): Performance sedans like the S7 with Chery.
  • Stelato (Xiangjie): Executive luxury sedans like the S9 with BAIC.
  • Shangjie: Mid-to-high-end family vehicles with SAIC.
  • Maextro (Zunjie): Ultra-luxury vehicles aiming at the Rolls-Royce/Maybach segment with JAC.

Most of these brands are unknown outside China. Yet, domestically, they are engaged in a "bloodbath." Why is this competition "Desperate"? When cars become "smart mobile spaces" defined by IT giants like Huawei, Xiaomi, and Li Auto, traditional logic fails. Features that were expensive options 5 years ago—massaging seats, air suspension, 800V fast charging, and Level 2+ autonomous driving—are now standard. The essence of this is a violent fusion of industrial capacity and internet-speed iteration. For traditional automakers used to a 5-to-7-year facelift cycle, iterating software every few months and upgrading hardware annually is a relentless nightmare.

If you think the competition in the $30k-$50k range is fierce, wait until you look at the $10k-$20k (80,000 - 150,000 RMB) segment. This is the true "meat grinder" of China's automobile industry. This price range used to be the "cash cow" for global giants like Volkswagen (Lavida) and Toyota (Corolla/Levin). Today, it has transformed into a brutal battlefield where domestic New Energy Vehicles (NEVs) and internal combustion engine (ICE) cars fight for every single inch of market share.

Take the 2026 Changan Qiyuan Q05 as a prime example. Starting at roughly $11k USD, this compact SUV offers a spec sheet that seems like a typo to anyone living in Europe:

1. Battery & Charging: It comes standard with CATL (world-leader) batteries supporting 3C fast charging, adding 200km of range in just 15 minutes. (If it were a BYD, this time could be shortened to 3 minutes.)
2. The Democratization of LiDAR: It is mind-blowing, but this $11k car features a LiDAR sensor, 3 millimeter-wave radars, and 11 high-def cameras. It’s powered by the Horizon Journey 6M chip with 128 TOPS of computing power. In the West, LiDAR is a luxury reserved for $60,000+ vehicles; here, it’s a budget standard.
3. The Tech Ecosystem: It features Huawei HiCar 3D Control, allowing users to manage the vehicle via a 3D digital twin on a 15.6-inch 2.5K ultra-clear screen. Furthermore, it integrates with Midea and Haier smart home ecosystems, allowing you to control your home appliances from your car.
4. Luxury as Default: The driver’s seat offers 10-way power adjustment with integrated massage, heating, and ventilation. It even includes a panoramic sunroof with an electric sunshade—features that are high-cost "optional packages" in most German or Japanese cars.

If you look at the Top 50 Sales Data from March 2026 for the $10k-$20k range, the level of competition is suffocating:

• The Total Takeover by EV/PHEV: The Geely Geome Xingyuan (priced at a mere $9,500 - $13,700) topped the charts with over 30,000 units sold in a single month. It is followed by a flood of BYD models like the Yuan UP, Dolphin, and the legendary Qin PLUS, which have driven the price of Plug-in Hybrids and EVs down to—or even below—the price of equivalent gasoline cars.
• Premium Tech at Budget Prices: Models like the Deepal S05, Leapmotor C10, and Xpeng MONA M03 offer cabin space and smart features at the $15,000 mark that would typically cost $40,000+ in overseas markets.
• The Collapse of Legacy Premiums: While names like Sylphy (Nissan) and Lavida (VW) still appear on the list, they are barely surviving by slashing prices. A car that sold for $18,000 five years ago now starts at less than $11,000.

Why is this the most desperate competition? Because in this segment, manufacturers are not just competing on features; they are competing on their ability to survive on razor-thin margins. Through a fully localized battery supply chain and massive automation, Chinese automakers have achieved a level of cost efficiency that traditional global giants simply cannot match.

The Chinese market isn't just building "smarter" cars; it is redefining the very value of an automobile through an industrial miracle. From a $10,000 commuter car to a $50,000 luxury executive sedan, every single price bracket is crowded with rivals ready to replace you.

As I felt when I sold that Lexus ES200: the era of resting on brand prestige and easy profits in China is dead. The current Chinese market is an "Evolutionary Laboratory", where a few months of stagnation means total elimination by the next wave of innovation. If you want to see the future of global manufacturing, this is it.

Chinese electric cars will be available in Canada soon. Fellow Canadian EV lovers, will you buy it or just stick to Tesla?

Most Westerners assume Chinese EVs are just "Tesla-lite"—cheap bodies with an iPad in the middle. They couldn't be more wrong.

By 2026, while Tesla sits comfortably on its high margins with incremental updates, Chinese automakers have entered a radical engineering race. I call it the "4K and 8K Engineering Standard." It's not about screen resolution (though they have 8K screens, too); it's about pushing the physical limits of mobility.

Look at the numbers below and ask yourself: When was the last time Tesla updated its core architecture to match these?

Why do people buy Hyundai or Kia cars when Toyota are much better?

To answer this, we need to look at the China market—the world’s largest and most competitive automotive "Gladiator Arena." The rise and fall of Hyundai/Kia (HMG) compared to Toyota's current struggle reveals a hard truth: "Better" is no longer defined by reliability alone, but by intelligence and value.

1. The Era of Glory: When Hyundai/Kia Ruled the Streets Hyundai and Kia entered China in 2002 (Beijing Hyundai and Dongfeng Yueda Kia). For over a decade, they were unstoppable. In fact, Hyundai-Kia once held a massive market share, with Beijing Hyundai consistently ranking among the top sellers.

In the 2010s, China had a famous nickname for the three most popular mid-size sedans: "The Three Treasures of Diors" (屌丝三宝)—referring to the Kia K5, Hyundai Sonata 8, and Chevrolet Malibu. While the nickname was a bit tongue-in-cheek, it proved that HMG had mastered the formula of "High Design + High Configuration + Affordable Price."

2. 2026: The Brutal Reality Check Fast forward to today, and the landscape has shifted seismically. Looking at the 2025-2026 sales data in China, Hyundai and Kia have essentially become "Others":

• Hyundai: The only model still "breathing" is the Elantra (Ranked 124th), selling about 58,000 units at a heavily discounted price ($8,000 - $15,000 USD).
• Kia: You have to scroll past the top 200 to find the KX1 (Yipao) at 244th place, selling a meager 22,000 units.

They didn't fail because they were "unreliable." They failed because they got stuck in the middle—not as prestigious as German brands, and not as tech-forward as Chinese EVs.

3. Toyota’s "Fortress" is Cracking Is Toyota "better"? In terms of brand loyalty, yes. Toyota still has 3 models in the Top 20 (Camry, RAV4, and Frontlander). But look closer at the former "King of Cars," the Corolla—it has plummeted to 72nd place, selling only 88,000 units.

The reason? The definition of "Better" has changed.

• In Sedans: The Xiaomi SU7 and Tesla Model 3 are selling at nearly double the price of a Camry, yet their sales are catching up at lightning speed.
• In SUVs: The Tesla Model Y, Geely Boyue L, and Xiaomi YU7 have already overtaken the RAV4.
• In MPVs: Even the legendary Sienna is under siege from high-end Chinese rivals like the Denza D9 and Voyah Dreamer.

4. The Price War of 2026: Buying "Time" Toyota is still selling because of massive price cuts. To stay relevant in 2026, Toyota has slashed prices by nearly 30% across the board to compensate for their lag in "Smart Cockpit" and "Autonomous Driving" technology.

They aren't alone—German giants (BBA: Benz, BMW, Audi) are even more desperate, with some price tags chopped by 50%.

The Final Nail in the Coffin If we look strictly at the New Energy Vehicle (NEV) track, honestly, neither Hyundai/Kia nor Toyota even have a seat at the table. However, unlike the Koreans who seem somewhat lost, the Japanese automakers have proven to be more "flexible"—or perhaps, more desperate.

In 2025 and 2026, we’ve seen a bizarre trend in China: Japanese giants are now "rebadging" Chinese tech to stay alive.

• Toyota partnered with GAC to produce the Bozi 3X, which is essentially a re-skinned Aion V.
• Mazda collaborated with Changan Deepal to turn the Deepal SL03 into the Mazda EZ-6.
• Nissan teamed up with Dongfeng eπ to transform the eπ 007 into the Nissan N7.

Let’s be real: the Chinese platforms these Japanese brands are "borrowing" are mostly entry-level, budget-tier architectures. These cars solve the problem of existence (having an EV to sell), but they don't solve the problem of competitiveness.

When a consumer looks at a "Toyota" EV that is actually a GAC Aion underneath, they ask: "Why should I pay a premium for the Toyota badge when the original Chinese version is cheaper, has better software, and more frequent OTA updates?"

What are your thoughts on the new Mercedes-Benz EQS electric sedan?

If you view the 2027 Mercedes-Benz EQS through a global lens, it looks like a solid step forward. But through the lens of the 2026/2027 Chinese market, it looks like a student finally finishing a 2024 assignment. For a car that will undoubtedly carry a "million-yuan" price tag, its specifications aren't just late—they are arguably derivative.

Here is why the 2027 EQS is struggling to maintain its "luxury" status in the world's most competitive EV market:

1. Battery and Architecture Mercedes is touting the 800V architecture and the 122kWh battery as major breakthroughs. In China, this isn't a breakthrough; it’s the bare minimum for survival. By 2026, even mid-range Chinese cars priced around $20,000 (140,000 RMB) have adopted 800V. Players like BYD have moved toward 1000V+ platforms with "Flash Charging" capabilities that make Mercedes' 350kW peak look average.

Mercedes and BYD have a long-standing history via the Denza joint venture. The Denza Z9GT, priced at a fraction of the EQS, already utilizes a 122.2kWh battery. Given the supply chain logic in China, it’s highly probable that the EQS is using the same—or very similar—battery technology. When your $140,000 luxury flagship shares its "heart" with a $40,000 local GT, the "German Engineering" premium starts to evaporate.

2. Smart Driving In 2026, if a car costs more than $20,000 (140,000 RMB) in China and cannot perform Urban NOA (Navigate on Autopilot) or advanced automated parking with LiDAR, it is effectively obsolete. Mercedes’ in-house "Drive Pilot" (L3) is impressive on German Autobahns but historically struggles with the "chaotic" complexity of Chinese urban traffic.

To stay relevant, Mercedes has reportedly deepened its partnership with Chinese tech firms like Momenta (and rumors suggest potential Huawei integration). If the intelligent "soul" of the EQS is actually provided by Chinese vendors, what exactly are you paying Mercedes for? In a market where Xiaomi and Xpeng iterate their AI models every few weeks, Mercedes’ traditional development cycles are a liability.

3. Interior and Luxury Mercedes used to define luxury with leather and wood. Today, Chinese consumers define luxury by "Human-Centric Intelligence." Brands like Li Auto, NIO, and Xiaomi offer "Full-House" configurations as the default. Features like 10-point seat massage, 800V fast-cooling fridges, and seamless "Home-to-Car" IoT ecosystems are standard.

Mercedes still clings to the old-world model of "base price + expensive options." In 2026, asking a customer to pay extra for ventilated seats or premium audio in a luxury EV is seen as an insult—especially when an $11,000 Changan Qiyuan Q05 offers those features by default.

The new EQS is undoubtedly a magnificent masterpiece of traditional hardware. It boasts exceptional aerodynamic design, world-class suspension tuning, and the iconic "Three-Pointed Star." However, within the specific context of the 2026 Chinese market, it has officially abdicated its throne as a "Technology Leader."

Of course, thanks to the immense global reputation of German automotive engineering, Mercedes will likely still find success with this car in Europe and the US. Let's hope it succeeds.

During the session, she shared an interesting legislative detail regarding the meaning of the term "ecological environment." For instance, after the institutional reforms in 2018, the newly established Ministry of Ecological Environment was initially translated into English as Ministry of Ecological Environment (ecological nature of the environment). However, not long after, this translation quietly changed to Ministry of Ecology and Environment (ecology and environment).

Do not underestimate this "and." In the constitutional text, "living environment" and "ecological environment" were originally in a parallel relationship. But as the context evolved, "ecological environment" took precedence, evolving into a grand narrative of "ecology + environment." Wittgenstein's famous philosophical proposition—"The limits of my language mean the limits of my world"—is particularly cruel in the legal field: a subtle adjustment in legal language often determines the depth and breadth of public power's intervention into private life.

On the surface, legal provisions are written in black and white, understandable by everyone. But in actual execution, the "boundary lines" behind the words are far more complex than we imagine.

Square Dancing Loudspeakers and the "Jurists" on Douyin

Recently, my residential community's owner group chat became lively again, mainly due to a group of square-dancing aunties next door who blasted their loudspeakers every night. Someone in the group shared a Douyin short video, excitedly telling everyone: "The newly revised 'Public Security Administration Punishments Law' has come into effect. The police can directly detain those making noise disturbances! Everyone, call 110 to report them!"

In the context of short video platforms, the law is extremely simplified into "instant gratification and revenge." But as a legal practitioner, I know deeply: Detention, as a coercive measure restricting personal freedom, is never something that can be executed on a whim. Noise disturbance, in the underlying logic of the law, is first and foremost a issue of civil tort or administrative management, and only lastly an issue of public security violation.

Deconstructing Article 88 of the Public Security Administration Punishments Law

The reason many people feel that "a change in the law means people can be arrested" is because they saw the mention of "detention for up to five days" in Article 88 of the new version of the Public Security Administration Punishments Law. However, many overlook the prerequisite path triggering this penalty, specifically the overlooked word "after" (经) in the provision:

Article 88 Those who violate laws and regulations regarding the prevention and control of social life noise pollution, generating social life noise, and after being dissuaded, mediated, and handled by grassroots mass self-governance organizations, owners' committees, property service providers, or relevant departments in accordance with the law, fail to stop, continuing to interfere with the normal life, work, and study of others, shall be detained for up to five days or fined up to 1,000 yuan; if the circumstances are serious, they shall be detained for more than five days but less than ten days, and may also be fined up to 1,000 yuan.

This word after serves as the "procedural firewall" for public security organs during law enforcement.

1. It is not parallel, but progressive

The public security organs' understanding of this article is usually that public security punishment is the last line of defense. If you call 110 directly, the police will most likely ask you: "Have you contacted the property management? Have you contacted the residents' committee? Have you reported to urban management?" This is not shirking responsibility, but rather the "principle of exhausting remedies" set by the law. The police need to confirm that before employing coercive measures, society's self-regulation mechanisms (property management, residents' committees) and administrative law enforcement mechanisms (handling by relevant departments) have been declared failures.

2. Who are the "relevant departments"?

Judging from the "Ecological Environment Code" scheduled to take effect in August 2026, the regulatory authority for social life noise usually lies with departments designated by local governments (such as urban management). If you cannot produce a "Notice Ordering Correction" issued by the urban management department, proving that the other party "failed to stop" even after administrative warnings, the public security bureau can hardly directly characterize the other party as "violating public security administration."

A "Clearance Guide" for Noise Rights Protection

To make the "teeth" of Article 88 of the Public Security Administration Punishments Law truly bite, we need to construct a closed-loop chain of evidence:

1. Activate Grassroots Responsibilities: According to Article 599 of the "Ecological Environment Code," in "noise-sensitive areas" such as residential zones, it is a statutory duty for property management and residents' committees to dissuade and mediate. Keep records of your complaints; this is the first step.
2. Trigger Administrative Intervention: Call 12345 and request intervention from "relevant departments" such as urban management. As long as they dispatch officers and leave a record of handling (even if it is just a verbal warning recorded on file), you have obtained the key to initiating public security punishment.
3. Public Security as the Safety Net: When the above methods fail and the other party continues to "interfere," then call 110. At this point, you can confidently cite Article 88 and tell the police: "The procedures have been completed, please punish in accordance with the law."

Conclusion

Returning to Wittgenstein's statement. If our understanding of the law remains only at the level of slogans in short videos, then our world of rights protection will be narrow and full of frustration. Only by clarifying the complex procedural boundaries behind the provisions can we precisely wield the scalpel of law in our noisy lives, cutting to the heart of the problem. Law is not a satisfying story for venting anger; it is a chess game of rights protection that requires careful planning.

Reference legal provisions:

"Ecological Environment Code" Article 599 For acts of social life noise disturbance in areas concentrated with noise-sensitive buildings, grassroots mass self-governance organizations, owners' committees, and property service providers shall promptly dissuade and mediate; if dissuasion and mediation are ineffective, they may report or complain to the departments responsible for the supervision and management of social life noise pollution prevention and control, or to the departments designated by the local people's government. The departments receiving the reports or complaints shall handle them in accordance with the law.

Trade-offs Between Efficiency and Accuracy in Small Models

Initially, I pursued extreme lightweight solutions and tried pure frontend small models around 20MB, such as ckiplab-bert-tiny-chinese-ner and the even more streamlined ckiplab-albert-tiny-chinese-ner. They were indeed satisfying to use; when I performed batch extraction with 100 test data entries, I found that these small models could almost complete processing on a PC within seconds. However, the processing accuracy was somewhat lacking: many personal names and organization names in the data body were not accurately identified; moreover, some three-character personal names were simultaneously recognized as two different people. Especially when facing cases with colloquial expressions, many colloquial entity names could not be recognized either.

Of course, during the design of this application, I had already considered this situation, so I added a very user-friendly manual review mechanism. The desensitized data from earlier is broken down into individual Chinese characters; clicking on uncompleted desensitization terms automatically adds them to the desensitization database.

However, the processing efficiency of small models is indeed not high; my actual testing showed that nearly 30% of personal names, abbreviated place names, and company names might fail to be recognized. Especially for some regional abbreviations and organization acronyms mentioned in cases, such NER models can hardly recognize them. And replacing these characters and words all at once can easily cause semantic confusion in the main text. Manual operation is also quite troublesome.

Large Models Place Significant Pressure on Hardware

Subsequently, I turned to the recently popular "small-parameter large models," namely Qwen3.5:0.8B, and for comparison, also pulled last year's DeepSeek-R1:1.5B for testing. After all, in a pure CPU environment, this is already the limit of size I can tolerate. I originally thought that the semantic understanding of large models could handle everything, but as it turned out, in entity extraction tasks, they performed extremely unstably. Not only could they not overcome the logical threshold, but long texts also easily lost commands. The only slight improvement was that after disabling Qwen's think mode, the output finally returned to normal, but during each round of testing, the resulting data remained highly unstable.

However, the process was not without gains. Unwilling to give up, I tested with Qwen3.5:4B and found that the 4B model performed very well for my long-text desensitization needs—almost as if it were a completely different model from the 0.8B version. However, on the laptop I used for testing, the Ollama version of Qwen3.5:4b had an output efficiency of only 5 tokens/s; even switching to the gguf version could only reach 7 tokens/s, which was still quite uncomfortable to use—a single data entry of under 100 characters took 2 minutes.

Medium-sized NER Models Generally Offer a Better Balance Between Performance and Efficiency

During the development of this project, I also comparatively tested solutions such as RaNER, HanLP, SpaCy, and Siamese-UniNLU. Detailed comparative data was not retained during the process, but based on subjective actual testing, none seemed to fully meet my requirements. For instance, RaNER encountered issues right from the start with long texts, and Siamese-UniNLU also failed on long texts. SpaCy and HanLP repeatedly had problems with text truncation; when judging certain entity names, they often got tangled with surrounding characters.

After repeated trials and adjustments, I finally selected two solutions that appeared to offer a relatively balanced performance and effectiveness:

Pure Local CPU Environment: Using RaNER and REX-UniNLU

After going through numerous iterations, I ultimately returned to focus intensively on two specialized models from the Alibaba ecosystem: RaNER and REX-UniNLU.

RaNER (Base-Generic) was originally the first model I downloaded, but it has significant shortcomings in long-text processing, with default support limited to only 512 Tokens. However, looking across the Chinese NLP community, various evaluation datasets show that RaNER delivers the strongest performance; the 400MB general-purpose Base version can achieve 91% accuracy in recognizing personal names, place names, and organizations. Although during my testing I found that it might not reach such a high ratio in the legal domain, considering that my final application approach adopts a "NER desensitization + manual confirmation + guardrail review" method, in practice, only two NER models with different logics are needed to achieve higher accuracy.

REX-UniNLU (DeBERTa-v3) was something I accidentally discovered while testing Siamese-UniNLU; it should serve as a replacement project for Siamese and natively better addresses the long-text input problem. However, since this model is also a 400M general-purpose model, it can not only handle NER tasks but also perform relation, fact, and sentiment extraction, making it quite suitable for use as a data guardrail reviewer in later stages. It is highly sensitive in capturing colloquial expressions and complex nested entities, effectively compensating for the rigidity of traditional NER models in boundary determination. Moreover, during this process, I introduced the strictest entity extraction mode, which can almost extract all names of entities that are "suspected" to be entities.

Online VPS Environment: Using ckiplab-albert-tiny-chinese-ner and SpaCy

If RaNER + REX aims to pursue ultimate accuracy on a PC with 16GB of RAM, then when I shift my focus to a Synology NAS with 4GB of RAM, or an inexpensive VPS with only 1 core and 2GB, this "heavy armor" setup simply cannot run. In such extreme environments, my favorite combination is ckiplab-bert-tiny-chinese-ner paired with SpaCy.

ckiplab-albert-tiny is less than 15MB in size; its role is to leverage BERT's residual semantic capabilities to "pre-filter" the most obvious, well-formatted personal names and place names in the text within milliseconds. Although it may miss complex colloquial expressions, its extremely low computational cost frees up resources for subsequent processing stages.

Next, I introduced SpaCy (zh_core_web_md). The advantage of SpaCy lies not in how "smart" it is, but in its extremely high industrial stability. As an NLP framework implemented based on Cython, it performs very robustly when handling administrative divisions and basic organization names. By chaining these two models together—one responsible for rapid initial processing, the other for maintaining baseline accuracy—combined with my "human-machine review" logic, a solid data security guardrail can still be constructed even in low-power environments.

Minimalist Installation

Compared to the hassle I had previously when installing on Ubuntu or Docker, this installation process on Feishu can be said to be revolutionary.

Download the App and register, and as soon as you open the interface, a prompt for one-click installation of Little Lobster will automatically pop up. Just click once, wait two minutes, and it's done.

The Default AI Is a Bit Silly

I hadn't even had a chance to ask which model this built-in AI actually uses. First, let me test the network connection by asking it to search online for interesting news on HackerNews today.

As a result, there was no response for a long time. After waiting for about 5 minutes, a bunch of messages came out consecutively, saying the network timed out.

Then I changed the request, asking it to check what's trending on Weibo Hot Search today. But it kept getting stuck in the previous task, still repeatedly trying to access HackerNews without success.

I reminded it for the third time to check Weibo Hot Search instead of continuing with the previous task, and it finally got the idea.

But obviously, after messing around for a while, it still couldn't find the Weibo Hot Search. It gave a bunch of reasons, which all made sense, but obviously, Weibo Hot Search is not only available on Weibo itself—many other websites also pull this information.

After much deliberation, it finally found a mirror version of Weibo Hot Search.

Then I asked it to follow the same approach and search for mirror information of HackerNews again, and this time it actually found it.

Since it had already found the information, I then proposed a third request, wanting to test its ability to analyze article content, as well as test whether it could search shopping websites.

And then, there was no "and then." The 4 million Tokens had already been used up!!!

Since I hadn't had a chance to ask what model it is, here I let Doubao roughly calculate the price of 4 million Tokens.

A Game for the Rich

To be honest, the effective time of this trial from start to finish didn't exceed 5 minutes. I didn't expect that the model consumption could be this fast. Previously, when I used my self-hosted Little Lobster for one or two hours, it seemed to only consume three or four million Tokens, and that was with intensive conversations and large amounts of long text input and output. According to this consumption rate, not many people can really afford to keep a Little Lobster.

The Vanished "Word Count Anxiety" Is Back

I remember 20 years ago, mobile phone screens weren't even as large as the camera island on some phones today. On one side, you had a 5 RMB 30MB data plan; on the other, a 5 RMB 100-message SMS bundle. Every month, you had to plan meticulously, counting every character to get by.

As it turns out, 20 years later, that feeling of being stretched thin has returned. Once your "phone credit" is gone, Claude stops replying to your messages.

Generally speaking, there are two mainstream ways to use AI Tokens today: monthly subscription plans and pre-paid pay-as-you-go.

As for monthly plans, the entry-level price on major domestic platforms is basically around 40 RMB/month. For example, the Zhipu GLM plan I’m using costs 49 RMB per month and allows for sending and receiving 400 messages per week. Calculating it out, it’s not that expensive—only about 3 cents per message.

The problem is, this plan limits you to 80 messages every 5 hours, which is 16 messages per hour. If you are a high-intensity AI user, you'll basically enter an "AI cooling-off period" after just half an hour of use. If you really want to make the most of Claude's functions, you have to upgrade to a monthly plan costing 149 RMB or even 469 RMB to meet the demand.

So, after all that, we’re back to where we started: meticulous calculation, trying to pack as much content as possible into every message, striving to "fill up all 70 characters" before sending the next one.

Why I Suggest Translating Tokens as "Ai-Xin"

Our generation has lived through SMS (plain text), MMS (with images), Feixin (cross-terminal communication), and WeChat (mobile internet). We thought that in 2026, with data long since being "all-you-can-eat" and video streaming being ubiquitous, billing units would start at GBs. Instead, the emergence of AI has dragged us right back to the Stone Age of counting conversation turns.

I thought about it, and calling Tokens something like "Lingpai" (Link) feels too rigid—after all, the APIs over there haven't been named yet—and calling them "Dianshu" (Points) feels too much like an online game. After much consideration, I feel "Ai-Xin" is the most vivid. Phonetically, it combines with "AI"; a message sent by an AI is an "Ai-Xin." Semantically, it is communication driven by intelligence, with every message carrying the warmth of computing power. Most importantly, it captures the pain point: watching consumption rise rapidly, it’s a case of "Ai-Xin-Bu-Xin" (Believe it or not/Love it or not)—the money is being deducted regardless.

AI Image Generation Today Is the "Extravagant MMS" of Yesteryear

If AI text dialogue is a 10-cent text message, then AI image and video generation is the "MMS" that everyone felt they couldn't afford back in the day.

I checked, and for mainstream image generation consumption, if converted based on Tokens or computing points, the cost of generating a high-quality image is often dozens or even hundreds of times that of a plain text dialogue. This sense of consumption is staggering.

Back then, an MMS was a 300KB image so blurry you couldn't even see a face, costing 50 cents.

Now, an AI-generated 4K ultra-realistic portrait might cost between 1 to 2 RMB in computing power.

In actual use, it’s almost a case of not using it unless absolutely necessary—saving wherever possible.

What’s the worst part?

AI image and video generation is an endless demand. You have to constantly revise prompts and fine-tune to get the result you want. This process, in terms of consumption, is even more painful than when an MMS failed to send or receive but you were still charged.

Of course, the "text-to-image" discussed here mainly refers to the current top-tier AI models. As for the free trials or "downgraded" models launched by various companies, they aren't even in the consideration. After all, a truly useful "AI" has never been cheap.

From Mao's Selected Works

The phrase "All reactionaries are paper tigers" in Mao's Selected Works is something almost every Chinese person should know, and the widespread use of the three characters "paper tiger" is largely thanks to this. Therefore, most people naturally take it as the source without much thought, which is quite understandable. However, from a "textual research" perspective, I still want to seek the true origin of this term.

From a Cantonese Dictionary

There is a common online claim that the source of "paper tiger" is the 1828 book A Vocabulary of the Canton Dialect compiled by the British missionary Robert Morrison. I searched around online but couldn't find a PDF version of this book. However, from common sense, if a term is included in a dictionary, it obviously means the term was already "widely circulated" before that. Taking a dictionary as the origin of a term clearly doesn't make sense.

From Ming-Qing Novels

When I queried this question on Tencent Yuanbao, Yuanbao suggested that the earliest appearance of "paper tiger" can be found in the Qing novel Guwangyan. In Chapter 15 of that novel, there is the sentence: "Up to now my master is just a paper tiger, originally fake, only good for scaring children and country folk." I checked carefully and found that this claim likely originates from an article on the Huayu Bridge website titled "Also Discussing the Origin of 'Paper Tiger'." But in the same article, the author also mentions finding this expression in Dream of the Red Chamber. Additionally, the author traces it to the Ming dynasty novels Jin Ping Mei Cihua and Chanzhen Yishi, with the former likely completed around the Wanli era, circa 1600, and the latter around the end of the Tianqi period, that is, about 1627.

From Late Yuan and Early Ming Novels

At this point I actually felt a bit stuck. After all, previous researchers had already traced it back to the Wanli period. But I was still unwilling to give up and switched to another AI to continue asking. Surprisingly, Qianwen concluded that the origin of "paper tiger" is in Water Margin. Now this became interesting.

I immediately searched the full text of Water Margin, but unfortunately, I didn't find the three characters "纸老虎". However, I did find the concept of "纸虎". Although "paper tiger" and "paper tiger" (纸虎) differ by one character, they essentially carry the same meaning. In Chapter 25 of Water Margin, when Wu Dalang catches Ximen Qing in the act of adultery, Pan Jinlian curses Ximen Qing loudly: "Seeing a paper tiger, you get so scared you fall over!"

Wu Dalang rushed to the door and pushed it with his hand, but it wouldn't open. He could only shout: "What a fine deed!" The woman held the door shut, panicking, and said: "Usually you talk big like a bird's beak, showing off your great boxing skills, but when it's time to act you become useless. Seeing a paper tiger, you fall flat on your face!" These words from the woman were clearly instructing Ximen Qing to strike Wu Dalang and escape. Hearing this from under the bed, Ximen Qing was reminded of the idea. He crawled out and said: "Lady, it's not that I lack ability, but for a moment I lost my wits." He then pulled open the door and shouted: "Don't come near!" Wu Dalang was about to grab him when Ximen Qing swiftly kicked out with his right foot.

Following this line of thought, I also looked at novels from the same period as Water Margin, especially those by Luo Guanzhong and Feng Menglong. After all, Luo Guanzhong was the disciple of Shi Nai'an, and the novels of these three have the widest circulation and the most content.

So in Luo Guanzhong's Pingyao Zhuan I also found the concept of "paper tiger", and it even appears directly in the title of Chapter 14: "Holy Aunt Hall's Paper Tiger Guards JinShan, Shuijing Garden Zhang Luan Meets Mei'er."

"Paper Tiger" Before the Song Dynasty

Having traced it to the late Yuan and early Ming, I basically already had the answer in my mind. To verify, I searched several ancient text databases, but the "paper tiger" results were all about actual paper-cut tigers, with no metaphorical usage like the modern one.

However, the process wasn't entirely fruitless. I suddenly saw a Tang dynasty reference possibly related to Buddhist scriptures mentioning "paper tiger". But upon clicking, oh boy, it turned out to be an OCR recognition error. The original text was "二十六纸" and "虎耳太子经", completely unrelated to "paper tiger". It was just that ancient people didn't use punctuation, so the two characters happened to be adjacent.

But it was this Buddhist scripture-related search result that made me think further—perhaps I could search Buddhist scriptures. After all, general ancient text search engines only include a small portion of Buddhist literature. For a broader search, a specialized database is needed.

Unfortunately, I didn't find any information related to "paper tiger" or "paper tiger" in Buddhist scriptures before the Song dynasty. However, I did find a Ming dynasty text San Yi Ming Meng Shuo, which clearly records a Chan master saying to his disciple after the disciple bowed and then acted arrogantly: "You are just one paper tiger." This usage is exactly the same as the modern one.

A monk bowed and rose. The master said: "Wrong." The monk gave a shout. The master raised his staff in a striking pose and said: "This order should be carried out by the senior seat." The monk swung his sitting mat in a circle. The master grabbed him and said: "Why not speak a sentence?" The monk said: "I'm afraid it will become a mess." The master struck repeatedly and said: "One paper tiger." — [Ming] San Yi Ming Meng · Shuo, compiled by [Ming] Jing Fan

Conclusion

So, if I must give a conclusion, I prefer this understanding: The concept of "paper tiger" formed during the novel-writing period of the late Yuan, Ming, and Qing dynasties, possibly absorbing and borrowing from folktales circulating at the time, or perhaps simply invented by the authors themselves. Regardless, this concept already appeared during the time Shi Nai'an (c. 1296–c. 1370) was writing Water Margin, since Shi Nai'an was only compiling and editing. The specific events inside trace back to the Liangshan uprising in the Northern Song period in 1119. By the early Ming period when Luo Guanzhong lived (c. 1330–c. 1400), the concept of "paper tiger" had already been used multiple times. By the late Ming, even Buddhist literature began using the "paper tiger" concept. Later, in the Qing dynasty, whether in Dream of the Red Chamber or various other novels, the "paper tiger" concept became increasingly widespread, to the point that even regional dialect dictionaries included it. Then in 1946, phrases like "All reactionaries are paper tigers" and later "U.S. imperialism is a paper tiger" gave the term comprehensive global promotion, so much so that foreigners eventually started using it too.

Sometimes the answer to a question isn't that important, but peeling back these layers one by one is itself very interesting.

There is a typical phenomenon that illustrates this. For example, I used to be accustomed to using Bing to search for many technical problems, but it seems I haven't used it much for a long time recently. Looking closely at my search history, it is basically direct tool demands, such as "SVG to PNG Online Tool", "QR Code Online Generator", "Random Password Generator", and regarding technical problem searches, it is more like "Claude Account Ban Rules", "Gemini Image Prompt Collection", "ChatGPT Long Conversation Browser Freeze" these kinds of specific usage questions about AI tools.

Furthermore, because I use search engines less, I haven't redeemed Bing Rewards points for a very long time. Looking at the last redemption, it was a year ago. Before this, every year I could redeem two to three 100 Yuan shopping cards during normal use of Bing. This also indirectly shows that compared to a year ago, the number of times I use search engines has dropped by more than half.

And these demands have all been replaced by AI.

The Problem with the Technical Blog Route

For many years, the technical blog circle has mostly focused on the underlying interaction usage of the IT and Internet industry, "Show me the Code" is its spiritual core. Most technical problems, ultimately, revolve around solving how humans can better control machines. When encountering something unknown, first check the technical documentation, then search online to see if anyone has encountered the same or similar problems, and if all else fails, leave a message or send an email to consult technical experts. The greatest value of technical blogs is solving the scenario of "encountering the same problem". For the same problem, predecessors step in pitfalls and leave records, and successors can learn from and refer to the predecessors' ideas and solutions or help improve them. It is precisely this open-source mutual aid spirit that has led the rapid development of the IT and Internet industry for decades.

But the arrival of AI changed all this. Now mainstream AI has almost learned all human machine language knowledge, and its understanding ability of code has long surpassed any single human. At this time, encountering problems and then going to "search" and "shake people" has almost no meaning. And as practitioners, writing this kind of technical blog seems to have lost its "soul". Writing back and forth is not as good as AI writing, so why write these.

In the end, it seems there is only one path left, writing about how to use AI, how to let AI solve problems better.

Especially, now that a large number of programmers have entered the state of VibeCoding, solving problems themselves also begins to rely on AI. Sometimes they are too lazy to open a directory or delete a character themselves, but directly command AI to do it. In this process, the vast majority of problems encountered are almost automatically learned and understood by AI. In the future, when others encounter similar problems, AI might automatically help them solve it, and successors may not even encounter similar problems. Under these circumstances, this further exacerbates the demise of technical blogs.

Problems AI Cannot Solve

In the process of using AI over the past two years, my overall feeling is that AI capabilities are significantly improving. At the beginning of last year, AI's persistent memory was just a string of simple json user tags. By this year, AI's persistent memory can already support long conversations for a whole day. And an Agent operating system like OpenClaw, which theoretically has unlimited memory capabilities with AI development, is even more of a cross-era landmark moment. But overall, AI is constrained by its own functional implementation model and still has a long way to go, such as data pollution problems, targeted "poisoning" problems, especially for problems with "no ready-made answers", AI is easily induced to output wrong results.

For example, I have discussed some dialect words multiple times in previous articles. I wanted to find the original characters for two high-frequency words in my hometown dialect. One is "hide", "conceal" "hide" which in our local area is expressed with another character, pronunciation can be roughly written as b'ɔn, hide-and-seek is also called 躲 b'ɔn; the other is "drive away", "kick out" this preceding verb, the character we use there is called p'ɔn. I used many AIs to help me infer what these two characters are, but until now, none have inferred the appropriate Chinese characters. However, in this process, it was not without gain. I forgot which AI recommended me to find the "Modern Chinese Dialect Dictionary" volume for our local area. I followed this thought and found the book, but after downloading and looking, oh ho, the expert who compiled the dictionary back then also couldn't verify what these two characters are. I can only give up. (The pronunciation in this dictionary is slightly different from my hometown. The pronunciation here all starts with p, but in my hometown it distinguishes between p and b.)

For issues with accurate corpora, or sources that are usually relatively "authoritative", even for some things without "standard answers", AI can still help infer and restore many meaningful details. For example, yesterday after watching a video, I casually asked ChatGPT what is the earliest traceable usage time for the term "Paper Tiger", and whether there are clear written records. Since the GPT database might not have ready-made information, it chose web search, and then combined with reasoning tools, believed that the source might be "Water Margin". On one hand, in the Wikipedia entry, it cited "Water Margin" Chapter 25 when Wu Dalang caught the adulterer, Ximen Qing was in a panic, Pan Jinlian couldn't help but angrily say: "Seeing a paper tiger, also scare a fall!" This example as the earliest source; on the other hand, it found that Luo Guanzhong and Feng Menglong also have a chapter "Paper Tiger Guards Jinshan" in "Ping Yao Zhuan". Furthermore, it inferred that Luo Guanzhong as Shi Nai'an's disciple, himself played a huge role in the writing and promotion of "Water Margin", and Luo Guanzhong's own "Ping Yao Zhuan" also has large sections about "Paper Tiger". Considering that the completion time of "Water Margin" is slightly earlier than Luo Guanzhong's works, therefore determining "Water Margin" is the source of the "Paper Tiger" concept is no problem. I subsequently searched this term on the "Shidian" Ancient Text Search Platform, and indeed found the earliest records are all in the Ming Dynasty, especially with the most usage in Luo Guanzhong's works. And before the Song Dynasty "Paper Tiger" was truly just a physical concept in paper cutting.

Writing to this point, I feel a bit dazed. It seems that recently, indeed I haven't found many problems that AI completely cannot solve, or cannot provide solution ideas for. Including, originally I was very annoyed by Gemini's deceitfulness, often "overly confident" "imagining" some weird conclusions. But looking at the latest CAIS AI Capability Assessment results, I found that although Gemini is still very pitful, Claude and ChatGPT have largely solved this problem. GPT 5.4 has even dropped to within 10 points (the lower the score the better). And in another test of the "Ultimate Human Questions" which is the most difficult for AI, a year ago GPT 4o only achieved a score of 2.7. Inside 2500 questions, it could hardly answer a few; and up to now, Gemini 3.1 Pro has achieved a score of 45.9, equivalent to being able to tackle close to half of the difficult problems at any time. And those questions, themselves 99.99% of humans cannot answer.

What is the Future of Technical Blogs

Based on the current trend of AI development, and the value contained in the blog form itself, thinking about it, I feel there may be several aspects that can still be done long-term.

Posing Questions

Looking back now, AI's strongest ability is actually not "solving problems", but that the user can state the problem clearly. But the problem is, many times, we ourselves may not describe the problem clearly. For example, last night when I was debugging Pocket Hugo Theme I discovered a very headache-inducing phenomenon. When the article page loads, there always seems to be a fleeting jitter. I sent this phenomenon to OpenCode, it helped me change four or five directions, including possibly scrollbar problems, main CSS layout loading problems, browser auto-loading problems when font uses rem, and some padding might be wrong problems, but finally I chose to use browser debugging tools to manually check, and found it was caused by an external JS loading too slowly. This is a typical performance of "only knowing how to state the phenomenon" failing to state the problem clearly.

In other words, many times, we think we are discussing a technical problem, but actually this problem is likely a problem that has been misunderstood two or three layers. Like the above, the exception is just the surface, the real problem might be architecture design, or even earlier decision-making mistakes. And these things, AI is very difficult to truly help users find answers for. It will only within the boundaries given by the user, as much as possible give a "looks reasonable" answer.

So the technical blogs that can still be written in the future, may no longer be "how to solve this problem", but "is this problem actually a problem". Whoever can explain the problem clearly, has already won half.

Writing to My Future Self

For many years, I have joked that I am running this 1ip blog. Although sometimes there are a few visitors, but from the heart, I have always treated it as a work with only myself as the reader. Of course, if it can help others that is even better. Thus, considering search engine optimization, considering theme design, considering more eye-catching titles and copywriting also became helpless choices.

But now slowly realizing, actually this problem has returned to the origin. That is ultimately, blog this kind of thing, like a diary, high probability is still for self-use. Especially now under this development method heavily relying on AI, many things are actually not "cannot", but "forgot", forgot how I thought about this thing last time I encountered it.

And blogs, happen to leave for oneself a most systematic knowledge base, memory bank. These contents, when in the future AI is more mature, very likely will become a part of a similar "Digital Life". One blog after another, is the most real, most vivid "Life Snapshots", can be called, can be inherited, can be preserved long-term.

Recording Thought Processes and Personal Insights

Current VibeCoding has reached a level previously unimaginable. Coding as a part of life, meaning is becoming more diluted. Ultimately, still have to return to how to live one's own life well, how to reconcile, coexist, develop with this society.

This forces us back to those "ultimate philosophy" phenomena familiar to the public. Why in human history the top brains, many people ultimately returned from the objective world to the subjective world, from science, technology backwards to do philosophy, theology, metaphysics.

Before always thought this was "going astray", now instead feel, this might be the inevitable path.

Because problems in the objective world, one day will be solved by tools. And problems in the subjective world, such as our daily choices, anxiety, happiness, pain, these things never have standard answers, and also cannot be outsourced to AI to complete. Even when you are in pain, want to let AI comfort you, that effect, is obviously not as good as finding a real person to confide in.

And technical blogs, if they want to continue to exist, may also only be able to go in this direction. No longer recording oneself "how to do", but more recording "why do this", and in this process, one's real thoughts. These contents, may look a bit "no value" "no meaning", but undoubtedly is a person's most real response to this world. As long as people are still here, these things cannot become obsolete.

Everyone is marveling that the "AI operating system" has finally arrived, while showing off their workflows, their Agents, and what amazing things they've made AI do again. Last month I actually went hard at it too — trying all kinds of tools, switching between all kinds of models. In just a short period, I burned through several hundred dollars worth of tokens alone.

At first I was very excited, feeling that a new era had truly arrived, but as I kept playing, I gradually calmed down instead.

It's not that these things aren't good — quite the opposite. Precisely because they are so good, they made me break out in a cold sweat.

Imagine: in the future, in a chat group of 500 people, the ones actually speaking every day might be 499 individual AI BOTs; on a short-video platform, you scroll through hundreds of videos and every single one is mass-produced garbage content made by AI; on public accounts, forums, and websites, page after page is densely packed with text and images automatically assembled by AI.

In other words, we will very likely be living in an environment flooded with AI-generated garbage information in the future.

And the further this environment develops, the more I feel there will definitely be a "counter-AI" trend — people will start emphasizing again "this was written by a human", "this was drawn by a human", "this was photographed by a human", just like today people emphasize organic vegetables and family farms.

A Bottom-Line Judgment About AI

Over the past two years I've written many articles about AI. Leaving aside large-scale application of physical AI robots, just looking at the IT and internet application level, I'm increasingly convinced that AI applications should at least have one minimum Bottom line.

I believe that for things ordinary people can directly read and feel — such as articles, images, videos, music, and human-to-human expression — these areas should as much as possible be left to real people. It's not that AI should never touch them at all, but that we shouldn't take it for granted to hand these things over to AI for mass production.

Because these contents are originally part of the real world; their value is not just "data information", but also includes the emotions of the person expressing them. When you read an article today, you're not just looking at the conclusions themselves — you're putting yourself in the writer's shoes, thinking: what exactly am I reading here, how does it help me, what would I do if I encountered this situation? In the end, everything leads back to the real world.

But for another category of things, I don't have such a strong conservative attitude.

For example, binary code, various machine languages, programming languages, or all the countless intermediate-layer languages that programmers around the world have created since the information revolution to make machines run. Strictly speaking, these things were never meant for ordinary people to read in the first place — they are for machines. Since that's the case, if AI can directly play a role in these areas, it might actually be a good thing.

The Two Recent VideCoding Projects

It was precisely with this mindset that, during this period, I used OpenCode to complete two open-source projects in a row: the first one is Pocket-Hugo, and the second is Pocket-Hugo-Theme. Both revolve around Hugo, the fastest static site generator in the world. The former is used for editing and publishing, and the latter is a site template. .

The main reason I didn't use "lobster" for VibeCoding is that I'm "poor"; the secondary reason is that the development needs of a real project are truly countless — many things need repeated modification and repeated testing to discover problems. Relying solely on lobster would be slower than just opening an IDE to debug.

Strictly speaking, these two projects are actually one integrated whole, because when I was developing Pocket-Hugo, I added a unique page-editing feature to it, and this feature must be implemented by adjusting the Hugo theme. Since that was the case, I simply made my own theme instead of doing secondary development on someone else's theme.

Because of this, while making PocketHugo, I had already planned ahead how the theme should be done later. Many things could be reused — buttons, information hierarchy, card rhythm, color scheme, etc. — all directly carried over to the second project, greatly reducing the development difficulty. During the process, while writing PocketHugo on one side and making some convenient UI or interaction, the pocket-hugo-theme on the other side could immediately reuse it; when thinking of a better card layout, multi-language switching, or mobile display method in the theme, the tool on this side could in turn absorb it. The whole process was actually researching, learning, and VibeCoding at the same time.

And in this process, I myself became the biggest beneficiary. Many things that I previously only vaguely knew "could be written this way" were truly forced into deep understanding through repeated back-and-forth tinkering. Whether it's Next.js, the Go language, front-end CSS, or Hugo's various template mechanisms, I now have a much more concrete and solid grasp than before.

And precisely because of these three prerequisites, later when doing pocket-hugo-theme, for me it was no longer simply "making another theme".

Why Make Another Theme

If I just wanted to find a theme to use, there are already plenty of choices on the official Hugo themes site. The ones I used long-term before — like Hugo-theme-stack, and later bear-cub, etc. — are all very good themes.

But what I really want is not just "usable", but "just right".

Anyone who has seen my heavily modified version of Hugo-theme-stack and bear-cub knows that I've always been fond of the "text + image bundled" mode — the default display style of WeChat public accounts. Unfortunately, because most Hugo users are geeks and most site content is related to coding, whether there are accompanying images or not basically doesn't matter to them. There are even many Small Web projects online where people compete to see how extremely small they can compress their sites — 100KB is already not impressive; the smallest ones are only a few KB.

Another important issue is that I've always wanted a stable solution for publishing articles on mobile phones. I already talked about this in the previous article about launching Pocket Hugo, so I won't repeat it here. Although I also like IDE tools very much, I can't deny that using an IDE on a phone is really painful — especially the "anti-human" interaction between the default iPhone keyboard and web input boxes. Anyone who has ever copied or edited code in the GitHub app on iPhone knows how miserable it is.

So ultimately, what I wanted to make this time is a graphic-heavy theme with more prominent cover images, stronger card feeling, and at the same time suitable for long-term personal writing.

How I Developed This Theme

Developing a Hugo theme sounds quite easy. At least in the years I've been hanging out in the Hugo forum, I've seen a bunch of Northern European and North American "old men" in their fifties and sixties casually knock out a page template with just a few lines of code every time.

But when I actually started, I found it's not that simple. Because first I needed it to run properly on my own site.

So I basically took all the various components I had used in Hugo-theme-stack and bear-cub before and moved them over one by one to test. I don't even know how many times hugo server failed to build during the process; only bit by bit did I manage to get the template set up.

But it was exactly this process that caused me a lot of trouble later in development, because the earlier phase was essentially built on the premise that it only needed to serve me alone. Various elements were hard-coded, and even some page structures and image habits were completely centered around my own writing style. For personal use this approach has no big problems — just do whatever is convenient — but the moment I wanted to open-source it, the problems immediately appeared.

For example, various elements in the site header were previously hard-coded in my site; various components, image effects, and features were also just "I know how it works". But to open-source it, I had to change them one by one into configurable options so others could understand and use them too. The most troublesome part here was actually not writing the templates, but repeatedly judging what should be kept as the theme's character and what must be opened up for users to configure themselves.

At the same time, to submit it to the Hugo themes site, I also had to do multi-language support, comments, bundling, cover images, demo site, etc. as much as possible. Later I added a bunch of example articles and pages. Only after reaching this point did I more clearly feel that making your own website and making an open-source theme are completely different things. In the former, content is the goal and the theme is just the carrier; in the latter, content becomes the documentation.

Moreover, during this process, Pocket Hugo was actually constantly forcing this theme to grow in return. Because originally when developing Pocket Hugo I built it completely according to personal needs — if someone happened to have the same needs they could use my solution. But later, in order to release it as open source, I had to refactor it from scratch. During the process, someone even opened an issue requesting a local-only version that doesn't require GitHub authentication, which cost me several more hours.

In the end, I basically turned every customizable part of this theme into specific configuration items. And this process, for me personally, was almost a complete reshaping of my understanding of Hugo — constantly questioning myself: why set it this way, why not that way, what kind of setting is actually better and more in line with real usage needs.

Of course, I certainly don't dare claim this theme is already very mature. A theme is something that naturally gets improved while being used and grows while being written. But at least today, I can say with certainty that I didn't do this to add another piece of AI-era template garbage to the internet — I wanted to add one more brick that can truly be used long-term to this old-school, clumsy, yet still very reliable Hugo world.

I've never been a fan of this kind of freeloading behavior, but out of courtesy, I still took a look. Once I started reading, I was completely stunned.

To be honest, at first glance the contract left me a bit confused. It has all the elements, professional clauses, clear logic—clearly written by someone with industry experience. It's much stronger than ordinary “one-sheet agreements” or casually thrown-together one- or two-pagers. So why ask me to look at something like this? Even if you asked me to write one, I couldn't produce something this long. But after reading carefully… oh boy, from the very first sentence, it exudes an ominous feeling.

The further I read, the more my head hurt, and the worse it felt.

Throughout the process, only one question kept coming to mind: What kind of environment allows a contract that reeks of unfairness and illegality from beginning to end to appear so “righteous” and confident?

The Miserable Life of E-commerce Workers

After reading this contract, I feel that people working in e-commerce are truly suffering. This isn't “helping manage a store”—this is basically selling yourself into servitude.

The first article clearly states “Contract for Undertaking Services”. In plain language: the boss and employee are not in an employment relationship, but a cooperation/partnership. The boss doesn't pay social insurance, doesn't provide work injury protection—work one day, get paid for one day.

Yet at the same time, the contract requires working 8 hours every day, covering core operating hours (10 a.m. to 10 p.m.), with customer service online rate no lower than 80%, and product listing/delisting to be completed within the agreed cycle.

This is called a “cooperation relationship”?

If this doesn't count as an employment relationship, then there is no such thing as an employment relationship in the world.

Even worse, Article 7 also stipulates that Party B (the worker) must purchase commercial insurance on their own to reduce “cooperation risks”. I've seen shameless clauses, but never this shameless.

So the essence of this contract is very clear: it uses the skin of a cooperation agreement to cover up the reality of an employment relationship, transferring all costs and risks onto the worker.

What an E-commerce Worker Bears Every Day

Many people think e-commerce is just “selling stuff on a computer”—easy, free, quick money.

If you think that way, take a look at what an e-commerce operator actually does every day:

1. One person doing the work of three

   • Customer service: instant replies to buyer inquiries—even at 11 p.m.
   • Operations: listing/delisting products, creating detail pages, applying for promotions, analyzing data
   • After-sales: handling returns, dealing with bad reviews, getting cursed out by buyers

   These are three separate positions, but in small shops, one person often handles everything.

   Some might say: isn't this just what store clerks do? What's the big deal? But honestly, the biggest difference between e-commerce and physical stores is that labor is chained to data.

   In a traditional store, clerks know roughly how sales are going—no one chases you with a phone asking “Why did today's conversion rate drop 0.5% compared to yesterday?”

   E-commerce is different. Every sentence you send, every product title, every main image is monitored by data in real time.

   • Low inquiry-to-order conversion? → Customer service writes a self-criticism
   • DSR score drops? → Operations must find the reason
   • Visitor count falls? → Promotion person takes the blame

   But the problem is: do these metrics truly reflect an employee's real value? If buyers don't purchase, it could be because the product isn't good enough, the price is too high, or it's off-season. Yet the blame almost always falls on the execution layer.

2. On call 24 hours

   Traditional stores close at night and work stops. But e-commerce gets busier at night—because you never know how many night owls are out there. Many people lie in bed scrolling their shopping carts at 2–3 a.m. Not to mention endless man-made “shopping festivals”, ever-changing platform rules, and products that can suddenly explode or crash.

   The contract says “8 hours per day”, but anyone who's done e-commerce knows these 8 hours are just the “minimum spend”. Beyond the 10 a.m.–10 p.m. golden hours, there may be random “silver hours” or “diamond hours”. How long you actually work is something only you know.

3. Unstable income

   Many e-commerce operators work on “no base salary + commission”. When the store sells well, you get a share of the profit; when it sells poorly, you drink the northwest wind.

   And the commission rate can be “negotiated and adjusted based on store performance”? Once the contract is signed, the boss can change the split ratio at any time—you have zero recourse.

4. Zero protection

   No social insurance means no medical insurance, no pension, no work injury coverage.

   • Get into a traffic accident on the way to/from work? → Your problem
   • Get sick and need medicine? → Pay yourself
   • Retire one day with no pension? → Don't even dream about it

   But they work 9 a.m. to 9 p.m. shifts, under the same constraints as regular employees—why are they denied labor protections?

A Message to E-commerce Workers

Finally, I want to say: if you're currently working in e-commerce, or have friends doing it, take a careful look at the contract you (or they) signed. If it's this kind of contract, don't panic too much.

• When facing unreasonable demands → refuse them
• When wages are withheld → go for arbitration or file a lawsuit

If you're also an e-commerce boss, let me say this: the money you save on social insurance would be better spent hiring better people.

Those tricked in by this kind of contract won't stay. The ones who do stay are only there because they have no other choice. People who truly want to build a great store won't be retained by contracts like this; those retained by the contract are mostly already planning their exit.

Moreover, this kind of contract is almost certainly invalid under the law. Once the worker goes to arbitration—confirming the employment relationship, demanding back payment of social insurance, and claiming compensation—none of it can be escaped

Attachment: Original Contract Text (for reference if needed)

# E-commerce Store Operation Cooperation Agreement

## Article 1 Nature of Cooperation and Core Definitions
1. Both parties confirm that this Agreement is an equal civil cooperation / undertaking service agreement, not an employment contract or hiring relationship, and the Labor Law of the People's Republic of China, Labor Contract Law of the People's Republic of China, and work injury insurance-related provisions do not apply.
2. Core of cooperation: Party A provides all goods required for e-commerce store operation and bears operating expenses; Party B is only responsible for store operation-related services. Profits and responsibilities are allocated according to this Agreement.

## Article 2 Service Content and Scope
Party B provides the following operation services for the e-commerce store designated by Party A (Store name: ____, Store link: ____):
1. Product management: responsible for listing/delisting products in the store, syncing inventory information, basic optimization of product detail pages (only layout and text adjustments, not involving brand infringement, false advertising, or other violations).
2. Customer service: responsible for online customer service, including answering customer inquiries, handling order-related questions, and reporting reasonable customer demands (not bearing final compensation liability for after-sales disputes; special cases to be negotiated with Party A).
3. Basic operations: assist Party A in daily basic store maintenance (e.g., event registration coordination, order logistics information verification; not responsible for actual logistics handling).
4. Party B undertakes to complete the above work only within the agreed service hours and shall not arbitrarily leave the post and affect normal store operation.

## Article 3 Service Hours and Arrangement
1. Party B shall devote 8 hours of service per day. Specific start and end times are arranged autonomously by Party B; no clock-in attendance or reporting of specific periods to Party A is required, only ensuring coverage of core operating hours (e.g., 10:00 to 22:00; specific times to be supplemented by mutual agreement).
2. Party B may adjust the specific 8-hour service periods according to their own situation, but must ensure customer service online rate not lower than ____% (recommended 80% or above), and product listing/delisting operations completed within the agreed delivery cycle.

## Article 4 Cost Bearing and Profit Distribution
1. Party A's responsibilities: bear all costs of goods procurement and daily operating expenses required for store operation (platform service fees, promotion fees, store decoration fees, logistics shipping fees, etc.).
2. Party B's income: no base salary system. Income calculated based on monthly operating data of the store under Party B's responsibility. Specific allocation method: ____% of the store's actual monthly transaction total (after deducting refunds, returns, platform handling fees, logistics fees, promotion fees); settlement cycle is before the ____ day of each month. Party A shall complete payment within ____ working days after verifying last month's data is correct.
3. Both parties confirm: profit distribution ratio may be negotiated and adjusted based on store operating conditions; adjustments require a written supplementary agreement.

## Article 5 Social Insurance and Risk Bearing
1. Social insurance payment: since the parties are in a civil cooperation relationship, Party A shall not pay any social insurance for Party B (including pension, medical, unemployment, work injury, maternity insurance, and housing provident fund).
2. Party B undertakes: to handle its own social insurance payment matters and bear all related expenses, with no connection to Party A.
3. Risk confirmation: Party B confirms it is fully aware of social insurance regulations and consequences, voluntarily bears all responsibilities arising from non-payment of social insurance, and shall not claim any social insurance-related rights against Party A.

## Article 6 Commuting Safety and Liability Exemption (Core Clause)
1. Safety obligations: When traveling to/from the service location (or corresponding home office scenario), Party B must strictly comply with the Road Traffic Safety Law of the People's Republic of China and relevant laws and regulations, use transportation lawfully (wear helmet when riding non-motor vehicles, hold valid driver's license when driving motor vehicles, prohibit drunk driving, speeding, running red lights, etc.), and actively protect own safety.
2. Liability agreement: Any traffic accidents, personal injuries, property losses, or other unexpected incidents occurring during commuting or travel to/from service shall be fully borne by Party B and have no relation to Party A.
3. Party A exemption: Party B shall not claim any compensation, indemnity, work injury benefits, or other rights from Party A due to the above incidents. Party A bears no legal or economic responsibility. Party B confirms full awareness and voluntary assumption of all the above risks.

## Article 7 Rights and Obligations of Both Parties

### Party A's Rights and Obligations
1. Provide goods required for operations, ensure goods are qualified and supply stable, promptly provide product information and inventory data to Party B.
2. Bear all expenses related to store operations, promptly handle product quality issues, logistics disputes, and other matters outside Party B's service scope.
3. Pay Party B's income on time and in full as agreed, without unjustified delay.
4. Right to supervise Party B's service quality. If Party B seriously fails duties (e.g., long-term offline causing customer loss, product listing errors causing major losses), Party A may demand rectification; in serious cases, may terminate the agreement.

### Party B's Rights and Obligations
1. Complete store operation services as agreed, conscientiously handle online customer service, product management, etc., and maintain a good image of Party A's store.
2. Not disclose Party A's product costs, operating data, customer information, or other trade secrets; confidentiality obligation continues after termination.
3. Not engage in illegal operations using Party A's store (brushing orders, selling fakes, false advertising, etc.); otherwise, all penalties and losses shall be borne by Party B, and Party A has right to terminate and pursue liability.
4. Purchase necessary commercial insurance (e.g., personal accident insurance) on own initiative to reduce own and cooperation-related risks.

## Article 8 Term and Termination
1. Term of the agreement: from ____ ____, ____ to ____ ____, ____.
2. 30 days before expiration, parties may negotiate renewal; if no negotiation, agreement automatically terminates.
3. If either party wishes to terminate early, must give 15 days' written notice; after settling outstanding income, agreement may be terminated.
4. Party A may unilaterally terminate immediately without compensation if Party B: fails to perform customer service duties for 3 consecutive days causing serious bad reviews or increased complaints; causes major economic loss to Party A due to operational errors (e.g., listing errors leading to platform penalties exceeding ____ yuan); engages in illegal or违规 activities using the store and harming Party A's interests.

## Article 9 Dispute Resolution
1. Disputes arising during performance shall first be resolved through friendly negotiation.
2. If negotiation fails, either party may file a lawsuit with the people's court having jurisdiction at Party A's location.

## Article 10 Miscellaneous
1. This Agreement is made in duplicate, one copy each for Party A and Party B, effective upon signature (Party A affixes seal, Party B signs and provides fingerprint), with equal legal effect.
2. This Agreement constitutes the complete agreement between the parties, superseding all prior oral or written communications and agreements. Matters not covered may be supplemented by separate supplementary agreement(s), which shall have equal legal effect with this Agreement.

If you search my blog with "Hugo" as the keyword, you'll find four or five articles discussing the same topic: how to publish a blog from a mobile phone.

This wasn't my first attempt. Every time I had a writing inspiration while waiting for work, waiting for my child to finish school, or at home in the evening, I wanted to open my phone and update the blog. But then I'd think, "Forget it, I'll have to open the computer," or "I'll have to make edits after uploading anyway, might as well do it tomorrow," and just give up.

Anyone who uses Hugo knows the content itself isn't the problem—Markdown can be written anywhere. The problem is how to handle images and how to publish after writing.

Why Hugo?

Hugo's official tagline is "The world's fastest static site generator." This isn't just marketing; it's the real user experience. This blog currently has over 400 articles and more than 2,000 images, and each local build takes only 10-20 seconds. Compared to other static site generators, this speed is in a league of its own.

More importantly, Hugo's consistent build speed and clean file organization give me zero motivation to switch to any other platform.

The Obsession with Integrated Text and Images

As an old-timer in IT who came from the WordPress era, I have deep-seated trauma regarding the "separation of text and images."

Early WordPress didn't have a built-in image library. To add images to articles, you had to find your own image hosting and use external links. Back in the day, I used Weibo, Flickr, Photobucket, Renren... Later, these services either shut down or stopped allowing hotlinking, turning all the image links in my articles into red X's. Want to get them back? No chance.

So when I switched to Hugo later, what I appreciated most was its bundle mode: one folder per article, with the Markdown file and images placed in the same directory. When publishing, the entire folder is moved; images stay with the article forever—no loss, no mess.

This is, in my opinion, the ideal way to manage text and images. Bar none.

The Pain Point of Integrated Text and Images

Of course, I know that placing images in the article directory increases the GitHub repository size.

This issue is especially prominent on mobile—nowadays, a single phone photo can easily be 10MB. If you try to publish an article using the GitHub App on your phone, just uploading the images can bloat your repository.

In 2024, while optimizing my blog, I made a firm decision: convert all images to WebP format. Now my blog has about 2,000 images across 400+ articles, and the entire repository is only around 100MB. For most individual users, this solution is entirely practical—it retains the convenience of integrated text and images without causing repository explosion.

The Methods I've Tried Over the Years That Weren't Great

Back to the issue of mobile publishing. Since Hugo is so good, are there ready-made solutions for writing and publishing articles from a phone?

The Official GitHub App

The most straightforward idea. The result? The app is mainly for code collaboration; the Markdown writing experience is at a "barely usable" level. What's even more troublesome is adding images—photos taken with a phone are often 10MB+. Uploading them directly to the repository is simply not feasible. You have to compress them first, then upload, then paste the links into the article. Going through this entire process is more troublesome than doing it on a computer.

GitHub Issue Workflow

I once researched a specific solution: using GitHub Issues as a publishing entry point, with Actions automatically converting Issues into Hugo articles. I wrote an article detailing the implementation: Publishing a Static Blog via GitHub Issue (in Chinese). Feel free to check it out if interested.

At that time, I felt I had reached the "endgame" for "publishing Hugo from a phone." I used GitHub Actions to automatically fetch issue content and custom label formats to extract Frontmatter information. But in practice, it felt "not formal enough." I had to remember various settings in the workflow, and I actually used it very little—only twice since that article.

Headless CMS

I tried solutions like Netlify CMS, Pages CMS, and others. The experience was indeed better than the GitHub web interface, with editors and image management.

But they all had a fundamental problem: separation of text and images. The design logic of these CMS platforms is to manage "articles" and "images" separately—articles stored in one repository, images in another asset library, linked together via URLs in the end.

Even if you managed to configure it successfully, there were usage issues. For example, images wouldn't display in the editor.

This inherently conflicts with Hugo's bundle mode. You publish an article in the CMS, images get uploaded to a remote image library, and Hugo's rendering gets all messed up with the paths. It's not that the CMS is bad; it's just incompatible with Hugo's design philosophy.

Temporary Solutions

I've also used several temporary solutions, like CodeServer (a browser-based VSCode) and online editors like StackEdit. In fact, this time I initially wanted to fork and modify StackEdit to adapt to Hugo's default upload paths. But during the modification process, I found the project was too old with many outdated dependencies. Moreover, the project itself was quite large, making modifications less efficient than building something new from scratch.

And thus, Pocket Hugo was born!

What Pocket Hugo Solves

1. Browser-first, open and use

No need to download an app or configure complex environments. Phone, tablet, desktop—open a browser, log into GitHub, and start writing.

2. Respects Hugo's bundle mode

Supports three content structures:

• Flat Markdown: Traditional single file
• Single-language Bundle: One folder per article, containing index.md + images in the same directory
• Multilingual Bundle: Multi-language versions

Maintains Hugo's original directory structure. No need to change writing habits to accommodate the tool.

3. Automatic image compression before publishing

Take a photo with your phone and upload directly. The system automatically compresses and converts to WebP; no manual image processing required. This retains the convenience of integrated text and images without bloating the repository.

4. Drafts saved in the browser

Didn't finish writing halfway through? Open it next time and continue. No content loss due to interruptions.

5. Zero server-side storage

All data stays in the browser. The server doesn't store tokens, articles, or any user data. Use it and leave; no traces remain.

Ready-to-Use Online Version:

• Entry: https://leftn.com
• Guide: https://leftn.com/guide

Self-deployment is also simple, with documentation available for Vercel or Cloudflare Workers.

Open-source project address: https://github.com/h2dcc/pocket-hugo

If you're also struggling with Hugo mobile publishing, feel free to give it a try. Any issues are welcome as GitHub Issues, and stars are appreciated too.

Back then, I wasn’t sitting at a computer typing seriously. Instead, I used voice-to-text, dumping whatever came to mind straight into the notes. The original recordings have long since vanished; all that remains are the transcribed texts from that time. The speech recognition accuracy back then was far lower than today, with a high error rate. Many passages are fragmented, disjointed, or even look like gibberish at first glance. But precisely because they weren’t “rewritten” later, but were spoken and recorded on the fly, they preserve something of the raw shock of that first reading.

Looking back now, the fact that these words can see the light of day again feels somewhat miraculous: if iCloud sync hadn’t preserved the old notes, and if I hadn’t been able to use AI today to help restore some of the broken sentences, these pieces would probably have been lost forever.

So this article is not just a reading reflection on The Three-Body Problem; it is also a reOrganizing reading traces from many years ago. It is neither purely “today’s me” writing a commentary for the present, nor a verbatim copy of those error-filled voice transcripts. Instead, it tries to strike a balance between the two: staying faithful to the original thoughts while making them readable and fluent today.

Why I Started Writing Again Back Then

During the period when I was reading The Three-Body Problem, I had actually gone a long time without seriously writing blog posts. It wasn’t that I never thought about “writing something again,” but I never stuck with it. In the end, it was partly a matter of time, partly plain laziness, and more importantly, because I was typing so much at work that I developed an instinctive aversion to “writing anything more.”

This also explains why my blog updates became fewer and fewer, and why I eventually filled the gaps by posting pictures or scattered fragments. It wasn’t that I had nothing to say; typing itself had started to disgust me.

Later, one day, I slowly discovered that voice input suited me perfectly. Especially on the iPhone and iPad, speaking directly was far easier than sitting down and typing word by word. At the time, I also compared several tools. Youdao Cloud Notes’ speech recognition, to be honest, was not as good as iFlyTek’s—more errors—but iFlyTek had strict limits: only thirty seconds without holding the button, and just one minute even if you held it. It always felt like you were being interrupted, and your train of thought broke easily. In comparison, although Youdao Cloud’s recognition was a bit worse, it was better for continuous recording. Its biggest problem was that it segmented too aggressively—almost every two or three seconds it would cut a new section—making later cleanup extremely troublesome.

But regardless, it was from that time onward that I slowly picked up the habit of “recording ideas” again. And the reading reflection on The Three-Body Problem was exactly what came out of that background.

Why I Decided to Read The Three-Body Problem

I had actually heard about The Three-Body Problem very early on and was never completely uninterested. But that interest was more like “I know this thing is famous and powerful, I should read it when I have the chance,” without ever actually acting on it.

My impression of science fiction novels had stayed stuck in childhood—Twenty Thousand Leagues Under the Sea, Gulliver’s Travels, that sort of feeling. As for Chinese science fiction, before reading The Three-Body Problem, I had almost no clear impression, let alone any anticipation. If I exaggerate a bit, my sense of Chinese sci-fi was even less defined than my sense of traditional imaginative literature like Journey to the West.

What really made me decide to read it was a long video about The Three-Body Problem I saw on Bilibili at the time. The video was over eighty minutes long and used montages and commentary to lay out the general story arc of the entire trilogy. Combined with the frequent discussions I saw on Zhihu, the book had developed a very special aura in my mind: it was hugely famous, highly praised, yet seemed extremely complicated—complicated enough that I always felt I couldn’t keep up right away, so I kept putting it off.

That night I originally planned to buy the physical book—after all, it was only a few dozen yuan, not expensive. But on a whim, I first found a PDF version someone had shared online, thinking that since I wanted to read it now, I might as well start immediately. As a result, I was almost instantly sucked in.

The Three-Body Problem Burst My Original “Worldview” Wide Open

The first shock I received while reading the first book wasn’t any specific plot point, but a very direct feeling: it suddenly pushed my previous understanding of “how big the world can be imagined” outward.

We Chinese are not lacking in “grand” world imaginations. From Pangu creating the heavens and Nüwa patching the sky, to the cosmic order, reincarnation of all beings, and the systems of gods and Buddhas in Buddhism and Taoism, traditional Chinese culture has always had a complete set of ways to explain the universe and life. Journey to the West may look like a gods-and-demons novel on the surface, but the hierarchy of the Heavenly Court, the Buddhist realm, the demon realm, and the mortal world—how ranks are arranged, who stands where—is essentially a quite complete world structure.

But The Three-Body Problem is different. It doesn’t expand the world in a mythological sense; instead, on the foundation of modern cosmology, it rebuilds a cosmic panorama that far exceeds everyday experience. It doesn’t tell you how many gods are in the sky, nor does it speak of another mysterious kingdom beyond the world. It says: the universe we live in may itself be far more complex than the world we are accustomed to understanding. Multiple universes, small universes and large universes, dimension changes, cosmic reincarnation, the rise and fall and restart of civilizations… Once these things are strung together, you suddenly feel that much of what you once thought was already grand imagination was still, in many cases, revolving around “humans.”

One of the most powerful things about The Three-Body Problem is that it does not place humanity at the center of the universe to tell its story. Instead, it first acknowledges the universe’s vastness, indifference, and complexity, then turns around to examine what humans really amount to within it.

Why the Universe Is the Way It Is

Some Western science fiction I had encountered before—especially films—often felt quite limited to me. The most common trope was: an alien civilization arrives, either invades Earth or threatens humanity, then both sides fight fiercely, and the story basically ends after the battle. No matter how large the background, it essentially revolves around a war, a crisis, or a conflict.

Even some works with rich settings, in my view, rarely reached the scale of The Three-Body Problem. For example, Harry Potter has a world that is certainly not small, but ultimately it’s still just a few races, a few powers, and a few main storylines; its narrative grandeur is not on the same level as The Three-Body Problem. Or take many typical Hollywood sci-fi blockbusters—they may look like the sky is falling and the universe is at war, but they’re still fundamentally about “the enemy has come, how do we fight back,” rarely touching on “the truth of the universe” itself.

Of course, that doesn’t mean the West has no higher-level science fiction. Works like 2001: A Space Odyssey also discuss human origins and civilizational leaps. But overall, they lean more toward imagery and philosophical atmosphere. In contrast, The Three-Body Problem feels to me like it doesn’t just touch these questions occasionally; from beginning to end it is trying to explain why the universe is the way it is, why civilizations develop this way, and exactly where humanity stands within it.

It doesn’t use the universe as a backdrop to tell a story; instead, through the process of telling the story, it gradually pushes certain fundamental laws of the universe in front of you. This is the biggest difference between it and many ordinary science-fiction works.

Chinese People Can Also Create a “Modern Cosmic Myth”

Later, when I thought back, one very important reason I found The Three-Body Problem especially shocking was that it made me realize for the first time, very strongly: Chinese people can completely write a grand, self-consistent cosmic narrative that belongs to the modern world.

The grand narratives we were familiar with in the past mostly started from history, mythology, religion, or ethical order. Buddhism, Taoism, Journey to the West, the Fengshen system—these things are certainly huge and complete, but they remain grand narratives within the language of traditional culture. The Three-Body Problem is different. On the foundation of modern science, cosmology, and civilizational theory, it builds an entirely new “mythic structure.”

I use the term “modern cosmic myth” not to say it departs from science, but because it possesses the same kind of power that only myths have: it explains the world to you, explains order, explains life and death, explains reincarnation, explains fate—except it uses civilizations, technology, dimensions, and cosmic laws instead of gods and Buddhas.

This surprised me greatly. In the past, if someone asked who in Chinese literature could make the world feel especially vast, the first things that would come to my mind were still the classical works. But The Three-Body Problem made me realize that in today’s context, Chinese people can also retell “how big the world is” without using mythic language—using science fiction, cosmology, and the scale of civilization—and do it with great persuasiveness.

It’s Not Just Because the Sci-Fi Setting Is Brilliant

I’ve always felt that the reason The Three-Body Problem caused such a huge stir abroad is, first and foremost, because the book itself is genuinely strong: strong setting, grand scope, astonishing imagination, and internal logic that largely holds together. But beyond that, there are other elements inside it that probably caught the attention of overseas readers—especially Western readers—particularly easily.

One very obvious point is that it uses the background of the Cultural Revolution to open the entire story. This choice is, of course, first and foremost literary—providing sources for character development and historical trauma—but it also naturally carries a layer of “letting the outside world see how China reflects on its own history.” Western readers are often very willing to see Chinese works present and reflect on humanity and political trauma under extreme historical conditions; I think this is an objective fact.

Of course, this does not mean that The Three-Body Problem’s value relies on this to stand. Quite the opposite—I believe the book truly holds up because it is solid enough in itself. But if we’re talking about why it gained extra attention in its overseas spread, this layer of displaying China’s historical shadows and extreme states of humanity is indeed a very real reason.

The Real Core of the First Book Is Ye Wenjie

If we look only at narrative structure, Wang Miao is naturally the main viewpoint character in the first book; many mysteries and settings unfold gradually through him. But if I had to say who the true core figure of the first book is, I would still say it is Ye Wenjie.

Wang Miao is more like the person who brings the reader into this world—a thread, an entrance—while Ye Wenjie is the real spiritual center of the entire first book. Because the crisis begins with her, and the deepest ethical questions are concentrated on her. She is not an ordinary “villain,” but a person who, in an extreme historical environment, is gradually pushed to the point of completely denying humanity.

The novel uses the Cultural Revolution to shape her, not as a simple historical smear, but by placing the first half of her life in a context where humanity is extremely torn apart. In that environment, family bonds, knowledge, dignity, and order can all be easily destroyed. It is not entirely incomprehensible that a person would lose faith in humanity and even reach the point of “since humans are so unworthy, let’s destroy them all together.”

What I find most powerful about The Three-Body Problem in Ye Wenjie’s character is this: it does not rush to judge whether she is right or wrong. Instead, it first lets you see how her thinking grew. She did not go mad for no reason; she reached complete despair toward the entire human civilization only after experiencing extreme human evil.

Ye Wenjie “Makes Decisions for the Entire Earth”

Of course, understanding Ye Wenjie does not equal agreeing with her.

The extremity of her thinking lies in this: after losing faith in humanity, what she ultimately chooses is not “to distance herself from humanity” or “to change humanity,” but to expose Earth to a higher-ranking alien civilization. The problem with this choice is that it is no longer merely judging humanity, but deciding for the entire Earth.

Because Earth is not only humans. Even if humans are indeed evil and have done many ugly, cruel, even anti-human things, Earth still has other life forms and an entire ecosystem. If an alien civilization truly arrives, what it changes and destroys may not be only humans, but possibly the entire Earth’s biosphere. At this level, the question is no longer “whether humanity deserves destruction,” but “by what right do you decide the fate of the entire planet?”

This is also where I think The Three-Body Problem is very clever: it does not simply stay at the level of “human nature is evil, humanity deserves destruction,” but quickly pushes the issue to a higher ethical dilemma. Even if you are disappointed in humanity to the extreme, that does not mean you have the right to let all species go down with you.

In this sense, Ye Wenjie’s thinking is not even entirely the same as ordinary terrorism. Traditional terrorism often still has the goal of “preserving oneself or one’s own group,” whereas Ye Wenjie is more like a total negation that drags everyone down together. She is not destroying others to save herself; she simply no longer cares about herself either. This idea is extreme and terrifying, but precisely because of that, it feels complex rather than something that can be summed up as “she is just a bad person.”

Simulating How a Civilization Can Restart in Desperation

Another important thread in the first book is Wang Miao entering the “Three-Body” game. The novel spends quite a few chapters here, using several sections to unfold the game. On the surface, it looks a bit like a monster-slaying leveling-up routine: enter a world, understand the rules, fail, restart, level up bit by bit, and finally approach the real problem of that world.

But what makes this truly interesting is not the “game feel,” but that it actually uses a very intuitive way to simulate how a civilization repeatedly destroys, restarts, and evolves in an extreme environment.

The core dilemma of the Three-Body world is the disorder and disaster caused by its three suns. The planet has no stable orbit; civilization cannot slowly accumulate like Earth civilization in a relatively stable natural environment. Instead, it is often just starting to develop when it is thrown back by extreme climate and cosmic order. This means the evolution of Three-Body civilization is not linear, but cyclical. It does not steadily progress from the Stone Age all the way to the Information Age; instead, it is destroyed again and again, and each time it must somehow preserve the spark of civilization.

This setting is actually very powerful. Because it is not just to show off “I came up with such a strange planet,” but to use this planet to discuss a bigger question: what exactly does a civilization depend on to continue stably? If you throw a civilization into a completely unstable cosmic environment, what does it need to survive?

Many of The Three-Body Problem’s Imaginations Are Not Isolated

The specific setting in the Three-Body game that impressed me most was “dehydration.”

Compressing an originally three-dimensional person into a flat “human skin” in an extreme environment, then rehydrating and restoring them when the environment recovers—this setting was truly shocking the first time I saw it. On one hand it carries a certain absurdity; on the other, it perfectly fits the extreme survival conditions of the Three-Body world: if normal life forms cannot withstand the disaster, temporarily change form and survive first.

On the surface this looks like nothing more than a wild sci-fi idea, but the further I read, the more I felt it was not an isolated invention. Because in the third book, the dimensional strike, the two-dimensional foil, and the two-dimensionalization of the Solar System are all essentially related to this “dimensional change.” In other words, many of the most stunning settings in The Three-Body Problem are not random thoughts thrown in; they have underlying echoes with one another.

This point is very important. Because the problem with many sci-fi works is that they have lots of settings, but they are scattered; after reading, you just feel “the author has a big imagination.” The Three-Body Problem is different—many of its imaginations come back later and explain one another. Precisely because of this, it makes you feel you are not simply watching a few spectacles, but watching an increasingly complete cosmic structure.

How Could Communication Between Different Civilizations Be So Easy?

Of course, the first book is not without places that made me skeptical.

One issue I cared about from the beginning was communication between Earth and the Trisolaran civilization. This question may be very common in sci-fi novels; many works simply assume that once a signal is sent and received, the two sides can gradually start talking. But when I read it, I always felt this was actually extremely difficult—far more difficult than imaginable.

Forget alien civilizations—even on Earth, communication between different languages and cultures often requires long periods of learning, contact, and trial and error. Even between humans and animals, after so many years of coexistence, true “communication” is still very limited. So why should a civilization from a completely different planet, with a completely different perceptual system and evolutionary path, be able to establish a relatively smooth information-understanding relationship with Earth?

To go deeper, language itself is a problem. How do you send Chinese content to aliens so they can understand it? Their way of perceiving, expressing, and structuring logic may not be the same as humans’. Of course the novel cannot spend too many pages on this; otherwise the single issue of “how to understand each other” could support another whole book. Even so, this doubt remained with me.

However, I don’t think this undermines The Three-Body Problem at all. On the contrary, it is more like a reservation that naturally arises when I, as a reader, bring real-life experience to the text. Great works do not have to seal every question airtight; they mainly build an overall persuasiveness strong enough that even if you keep a few doubts in your heart, you are still pulled forward by the story.

Half-Believing, Half-Doubting—Yet Eyes Opened Wide

In the second half of the first book, several settings left very different impressions on me.

First, the Red Coast Base and Ye Wenjie’s storyline. This part is very important because it grounds the entire novel’s historical starting point, letting you know how the connection between Earth and Trisolaris actually happened. But here I instinctively felt that some specific technologies and communication methods were a bit too smooth. In other words, they work literarily and are necessary for the narrative to advance, but if you think about them carefully from a real technological perspective, they still feel a little “too fantastical.”

In contrast, the Sophon setting was the part that truly made my eyes light up. The novel says the Trisolarans unfold a proton into a higher dimension, process and etch circuits in higher-dimensional space, then fold it back, ultimately obtaining a Sophon that can move at high speed and interfere with Earth’s scientific experiments. The first time I saw this setting, I genuinely had a “mind blown” feeling.

It reminded me of the “big thing compressed into a tiny entity” imagination in some sci-fi movies, such as the AllSpark in Transformers—a very huge thing shrunk into a very small particle through some method. But the explanation The Three-Body Problem gives here feels more systematic: it doesn’t simply say “technology is advanced enough to shrink things,” but links it to dimensions and spatial structure. A three-dimensional object enters four-dimensional space, is unfolded, processed, then folded back into low dimensions—this way of explaining makes it feel not just a prop, but part of an entire cosmic view.

As for the later scene where humans use nano-wires to slice an entire cruise ship in half, I was half-believing at the time. Because once it involved real Earth technology, I would naturally measure it against “can this really be done in reality?” That passage was indeed very visual and impactful, but I still felt it was a bit fantastical. Even so, it had a very strong novelty. That is to say, I may not have fully believed how realistic it was, but I acknowledged that as literary and sci-fi imagination, it worked—and it was fresh enough.

The Logical Relationship of the Trilogy

After finishing the first book, I had a very strong feeling: it is certainly a complete novel in itself, but it does not feel like a fully closed, self-contained ending. It feels more like laying out a huge net, gradually planting the things that the second and third books will truly unfold.

This is actually quite interesting. You discover that many parts of the first book seem to still be at the starting point of the crisis, the unfolding of mysteries, and the throwing out of settings, yet many truly powerful elements—such as the later cosmic laws, the fate of civilizations, dimensional disasters, technological explosion, etc.—have already been subtly buried.

So at the time, while I felt that some parts of the first book had not fully connected to the grander scope of what followed, I also felt that the author had already planted many things in the first half. You may not be able to say clearly whether he had the entire sequel completely planned out at the time, but at least you can see that the groundwork is extremely solid. Whether it grew while being serialized or had a rough framework from the start, being able to write the first book this way and then connect it to such a huge scope is already very impressive.

The “Wallfacer Project” Idea Itself Is Admirable

By the second book, The Three-Body Problem’s scope suddenly becomes much larger than the first. The first book mainly establishes the crisis, builds the world, and sets up that unsettling connection between human and Trisolaran civilizations, while the second book immediately pushes the question to a deeper level: when Earth civilization already knows that it will almost certainly face a destructive threat in the future, what can humanity do?

The Wallfacer Project is proposed in exactly this situation.

When I first saw this setting, the first feeling I had was not “this method must be very effective,” but “the thinking behind it is already very impressive.” Because it seizes a particularly key point: when an alien civilization is highly advanced, Sophons have locked Earth’s basic science, and they can observe human society in various ways, what advantage does humanity still have? The answer turns out to be not technology or weapons, but the opacity of thought itself.

In other words, the last thing humanity can rely on may not be greater power, but “you don’t know what I’m really thinking.”

This point is very clever. It instantly elevates war from the physical level to the cognitive and psychological level. The Wallfacers do not save the world by building a bigger weapon, but by keeping a plan in their own minds that others cannot fully decipher for the time being. This setting made me feel for the first time that The Three-Body Problem is not simply escalating enemy confrontation, but constantly delving into the essential question of “how civilizations fight each other.”

The Aliens Are Not Fighting Humanity Alone

Of course, the reason the Wallfacer Project feels both brilliant and cruel from the beginning is that it buries a huge loophole: the Trisolaran civilization is not a purely external force fighting humanity in isolation. It has organizations, supporters, sympathizers, and rebels on Earth—and those people are themselves human.

This point is very crucial. Because if the enemy were purely alien observers, then “the opacity of human thought” could indeed be a final barrier. But the problem is that what can truly crack human behavior, emotions, habits, and psychological patterns is often not the aliens, but precisely other humans living in human society.

In other words, the greatest enemy of the Wallfacer Project is not necessarily the Trisolaran civilization itself, but “humans helping to see through humans.”

So looking back, many of the Wallfacers’ failures were not because their ideas had no value, but because they ultimately could not escape being part of human society. Every move you make, every word you speak, every gesture, every long-term behavioral pattern you display will be repeatedly interpreted and dissected by others. The Trisolarans may not see through all the plans in your brain, but those people standing beside you, who are familiar with the rules of human society, may very well gradually force them out.

In this sense, the Wallfacer Project actually carries a strong tragic quality: the smartest solution humanity finally came up with may very well first be destroyed by humanity itself.

Why the Dark Forest Law Is So Shocking

The real core of the second book is, of course, Luo Ji and the Dark Forest Law that is ultimately fully revealed.

Many people feel that once this law appears, it is as if the entire book is suddenly split open by a bolt of lightning—everything instantly makes sense. I certainly had that strong feeling the first time I encountered it: the chill of “so the universe may not operate on a logic of cooperation at all, but on a logic of hiding and hunting” is truly unforgettable.

But looking back, I also feel it is not an entirely out-of-the-blue conclusion. Because in the first book, Ye Wenjie had already given some very important hints in advance; they just hadn’t been fully unfolded at the time. Concepts such as the chain of suspicion and technological explosion are not completely unimaginable either. Once you accept that it is very difficult for civilizations in the universe to establish real trust, and that technological development can explode in a very short time, then the conclusion of “hide first, stay vigilant, and strike first when necessary” is already lurking in the shadows.

So the most powerful thing about the Dark Forest Law is not that it is completely unexpected, but that it suddenly turns those scattered anxieties and scattered hints into an extremely cold yet almost complete logical chain.

What truly makes this law chilling is: it makes too much sense. You may not want to accept it, but it is very hard to casually refute it. Because once you look at it from the cosmic scale, the relationship between civilizations may not naturally tend toward understanding and cooperation as humans imagine. On the contrary, in a situation of scarce resources, opaque information, and extremely rapid technological leaps, hiding and eliminating may be more in line with “rationality.”

Are Humans in the Doomsday Still “Human”?

Besides the Dark Forest Law, another part of the second book that struck me hard was its depiction of doomsday society.

What impressed me deeply here is that when civilization truly faces the risk of extinction, the very concept of “human” changes. Or rather, a person who was originally constrained by morality, law, and ethics within an Earth community, once truly detached from Earth and from the original community, may no longer be the same kind of “human” in the original sense.

This may sound a bit extreme, but that is exactly the feeling the novel gave me. As long as you still maintain direct contact with Earth—like a kite that has flown far but the string is still tied to the ground—you still belong to Earth society and are still bound by those rules. But if one day Earth is gone, or your connection to Earth is fundamentally severed, you become a separate small world, a new kingdom. At that point, the morality, law, and ethics originally built on the Earth community begin to fail.

In other words, human morality does not exist in a vacuum; it has a strong material and environmental foundation. This also makes me think of ideas like “matter determines consciousness.” Under what survival conditions you live, you are more likely to form what kind of consciousness structure. When Earth still exists and humanity is still a whole, morality can hold; but once you reach the stage of civilizational doomsday, extreme scarcity of survival, and having to seize the last chances, people may quickly regress to a more primitive, colder state.

I feel that on this point, The Three-Body Problem goes deeper than many ordinary doomsday works. It does not simply say “doomsday comes, human nature turns bad,” but reminds you: much of the civilization we think exists naturally is actually attached to a specific community and survival environment. Once that foundation disappears, no one dares to guarantee what people will become.

The Third Book’s Information Density Is “Too Packed”

The third book, Death’s End, gave me the overall feeling that its information volume far exceeds the first two. The first book mainly establishes the Trisolaran crisis and the world framework; the second pulls up key structures such as the Dark Forest, the Wallfacer Project, and the droplet attack; but by the third book, the entire work suddenly leaps from the “Earth–Trisolaris” bilateral relationship directly to a larger cosmic-civilization scale.

Dimensional strikes, two-dimensional foil, curvature propulsion, lightspeed ships, pocket universes, cosmic laws, cosmic restart… these things pour out almost in heaps. The reading experience it brings is, on one hand, extremely shocking—because you clearly feel that this work is no longer satisfied with discussing a particular civilizational conflict, but wants to discuss the basic operating rules of the entire universe; on the other hand, it does make you feel that some parts are overly packed.

I felt very directly at the time that Yun Tianming’s three fairy tales were too long and could have been more concise. Some parts about civilizational backup, preserving the spark, and later arrangements, I actually skimmed. Not that they are unimportant, but the overall information density of the third book is so high that you feel your brain is constantly being forced to expand capacity. In some places, when it is written too meticulously, the reading rhythm even drags a bit.

Even so, it still stands. Because the truly powerful thing about the third book is not that every part is extremely tight, but that it finally fully opens the truth the entire series has been approaching: Earth civilization and Trisolaran civilization, which already seem grand enough, may still be only a very elementary layer when placed in the larger scale of cosmic civilizations.

Cheng Xin Is More Like an Ordinary Person in a Cosmic Predicament

In online discussions about The Three-Body Problem, there are a lot of people scolding Cheng Xin. Many people mention her and immediately say “saint,” “everything bad happened because of her,” “she missed two opportunities,” “destroyed Earth and then destroyed the universe,” and so on. But if you really follow the novel itself, I think it is not that simple.

The first time Cheng Xin faces a truly unsolvable predicament is when she takes over from Luo Ji as Swordholder. The position itself means: you hold a deterrence power in your hands; once you press the button, you expose the locations of both Earth and Trisolaris, almost equivalent to sending both civilizations into the Dark Forest together. But if you don’t press it, it may mean Earth loses its last deterrence capability and is suppressed or invaded by Trisolaris. In other words, this is not a simple “press and you’re right, don’t press and you’re wrong” choice, but a situation where no matter which way you choose, disaster may result.

In this situation, if Cheng Xin presses the button, she can immediately be regarded as the great demon who destroyed Earth; if she doesn’t press it, the consequences may also be extremely serious. Moreover, the reason she stood in that position is itself the overall choice of Earth civilization at the time. You cannot on one hand appoint a person who represents all of humanity’s ethical expectations to take this position, and on the other hand demand that at the critical moment she must act like someone who has completely abandoned human ethics. This is inherently contradictory.

The second time is similar. When it comes to issues such as curvature propulsion, lightspeed flight, civilizational escape, and technological disclosure, what she faces is not a clear game, but a problem with insufficient information, extremely severe consequences, and involvement in the entire internal order of humanity. At that time, it was widely believed that if curvature propulsion were rashly developed, it would very likely expose Earth’s location and invite higher-level Dark Forest strikes; on the other hand, if a minority grasped this technology first, it could also cause new tears within civilization—conflict between those who could fly away and those who could not. No matter who you give such a problem to, it is not something that “being a bit smarter can solve.”

Overall, if I have to say Cheng Xin has a saintly heart and deliberately put Earth in danger, I think that is unrealistic, because in her environment, based on the judgment at the time, it cannot be said that she was wrong. As for the later discovery that simply having a few thousand ships with curvature propulsion start at the same time could form a black domain around the Solar System and hide it—this was something that could only be discovered after later technological development; you cannot blame her. In fact, Cheng Xin already did a lot in the whole story. For example, in the communication with Yun Tianming, two of the three fairy tales were directly deciphered by Cheng Xin’s side. As for the hardest one—the one about the two-dimensional foil—in that situation at the time, it was hard to expect an ordinary person to figure it out. This kind of dimensional strike was something no one could have anticipated; only after humanity had actually seen something similar happen would they know that such a powerful thing existed in the universe.

So I prefer to see Cheng Xin as a sample: she is not simply “the wrong person,” but the key character The Three-Body Problem uses to show “normal human ethics in society become ineffective under extreme cosmic conditions.” What she fails is not her conscience, but humanity’s original scale.

The Ruthless Gap Between Cosmic Civilization Levels

Throughout the entire Three-Body Problem, the plot that made the chill finally settle deep in my bones was still the two-dimensional foil strike on the Solar System.

The Dark Forest Law is more of a logical coldness, while the two-dimensional foil is an almost concrete coldness that makes you tremble. Because at that moment you suddenly understand: Earth civilization spent so long, experienced so many technological leaps, and had even begun to approach lightspeed flight, curvature propulsion, and dimensional experience—abilities that once seemed like myths—but in front of a higher-level civilization, it still could not even count as a proper confrontation.

The harshest point in the book is how lightly it treats the character who carries out this act. The other side is not some awe-inspiring, arrogant cosmic overlord, but merely an extremely humble, ordinary cleaner in Singer civilization. With a casual flick, a very small two-dimensional foil is thrown out, and the entire Solar System is dimensionally destroyed. Although Earth, after the sun was struck, prepared to continue hiding, and although Earth civilization had already developed to a very high level—having gone through several technological explosions and achieved very great results, mastering lightspeed flight, curvature propulsion, and even truly experiencing the transition from three dimensions to four dimensions—it was still casually destroyed by a cleaning worker in Singer civilization with an almost random action. Just as the book says: “To annihilate you has nothing to do with you.”

This dramatic quality is very strong, but precisely because it is strong, it feels even more terrifying. The civilization you painstakingly built, your history, your wars, your ideals, your scientific breakthroughs—in the eyes of a higher-level civilization, may not even be worth being taken seriously. It is not “you can’t beat me,” but “you are not even qualified for me to treat this as a formal conflict.”

The feeling of “To annihilate you has nothing to do with you” is almost pushed to the extreme here.

The Handling of “Technological Explosion” Speed Is Shocking

Looking back, another place in The Three-Body Problem that left a very deep impression on me is its description of “technological explosion,” which almost reshaped my understanding of civilizational time scales.

If we look at the entire species history of humanity, we have already existed for a very long time; if we count from complete civilization history, it has been several thousand years. But The Three-Body Problem made me feel for the first time, very intuitively: for true cosmic civilizations, a few thousand years may not count for anything at all, and once entering the technological explosion stage, change can become frighteningly fast.

The Trisolaran civilization in the novel is a typical example. At the very beginning, we know that the first fleet that flew out actually had very low speed—less than even a small fraction of lightspeed. But looking further, in just one or two hundred years, they had already achieved completely different levels of technological leaps. Earth civilization is the same: it originally seemed to be struggling bitterly, yet in the following short few hundred years, it also rapidly came into contact with abilities that were once unimaginable.

In other words, the development of civilization is not necessarily uniform forward progress. It may accumulate slowly for a long time, then suddenly leap upward at a certain stage. This also exactly explains why the Dark Forest Law is so effective: because on the cosmic scale, you cannot use today to judge whether a civilization will still be weak a few hundred years later. A few hundred years is very long for humans, but for the universe it may be an almost negligible instant.

This is also why The Three-Body Problem increasingly makes you feel that many of humanity’s judgments about the universe today are actually very elementary. Not because we are not working hard enough, but because we may not even have truly understood “how civilization will actually change.”

A Warning for Humanity’s Future Space Exploration

I have always felt that the most valuable thing about The Three-Body Problem is not that it simply gave people many spectacles, but that it forces people to rethink: if the universe is really like what it says, then how should humanity face the universe in the future?

There is a very important point in the book, mentioned in Yun Tianming’s final recollection: if Earth had no life, then it would have no essential difference from other dead planets. What truly changes the face of a planet is life itself. But the problem is, once you acknowledge that life is not a particularly accidental or unique existence in the universe, you must also acknowledge: there may be far more than Earth that has life in the universe, and there may even exist civilizations far more advanced than us that we simply cannot detect.

At this point, the kind of vigilance The Three-Body Problem proposes becomes very real. The reason we cannot see them today is not necessarily because they do not exist, but possibly because they have long mastered the means to hide themselves. Just like stealth fighters on Earth—you cannot conclude that nothing is there just because conventional radar cannot see them. Once civilization develops to a higher stage, it can completely hide and disguise itself in ways we fundamentally cannot recognize.

From this angle, The Three-Body Problem does affect people’s imagination of future space exploration. It is not necessarily persuading people not to explore the universe, but reminding them: do not too lightly assume that the universe is safe, transparent, and empty. And do not too lightly regard “speaking to the universe” as a naturally romantic act. Perhaps on a larger scale, this behavior itself may carry enormous risk.

A Cornerstone-Level Science-Fiction Novel

If I must say, The Three-Body Problem is of course not without problems. Some communication settings in the first book make people skeptical, the climax at the end of the second book comes too abruptly, and some parts of the third book do feel overly packed and lengthy. Even certain technical descriptions, if you are particularly nitpicky and question them, are not completely free of flaws.

But these problems do not prevent me from regarding it as a truly important work.

Because what makes it truly powerful is not whether the details are 100% perfect, but that it establishes an entire grand, self-consistent, and sufficiently thought-provoking ideological structure. It is not simply stacking a few very cool ideas; it lets these ideas support and explain one another, ultimately forming a very complete sense of cosmic oppression.

This is also why I feel it fully deserves evaluations such as “classic” or even “cornerstone-level science-fiction novel.” It is not just telling a story about an alien civilization and human civilization, but constantly forcing you to reunderstand more fundamental things: civilization, human nature, morality, technology, cosmic rules.

The Lingering Shock Years Later

Now, reorganizing these voice-to-text remnants left many years ago, I can clearly feel: the me at that time actually did not have as many clear, mature expressions as I do now. Many judgments jumped around, many places were thought and spoken at the same time, and many sentences were not even smooth. But precisely because of that, they preserved something that is very hard to replicate later—the feeling of having your sense of the world suddenly and completely expanded by a book for the first time.

That feeling is not simply “thinking the book is good,” but suddenly realizing that humanity can be placed in such a large scale for viewing, that civilization can be understood within such a cold logic, and that the universe can be not a romantic background, but a structure that shows no mercy to anything.

Many years later, when I reorganize these remnants again, I am of course no longer in the same reading state as back then. Yet I can still see the shock of that moment from these drafts. For me, the truly precious thing about The Three-Body Problem is exactly here: it did not merely provide one reading experience, but left a very clear coordinate in a person’s thinking.

Many years later you may forget the specific plot, forget some names, even forget which device you first read it on, but you will remember that the first time you truly realized “the universe may not operate according to human imagination at all” was very likely when you read The Three-Body Problem.

The Three-Body Problem Forces Humanity to Reunderstand Itself

Overall, The Three-Body Problem has already become more than just a particularly good science-fiction novel in my heart.

What makes it truly powerful is that it does not use the universe as a backdrop to tell a legend, but constantly forces the reader to rethink: what exactly do humans count as? What exactly does civilization depend on? Under what conditions does morality hold? Is technology salvation, or exposure? If the universe fundamentally does not operate according to human scale, then how much of the judgments we are familiar with can still remain?

It does not give a standard answer; it gives you a new scale.

And once you accept this scale, many things become different. When you look again at human history, civilizational conflicts, technological development, and space exploration, you will subconsciously think one more step: if you put all this against a larger cosmic background, what exactly does it mean?

If some novels merely finish telling a story, then The Three-Body Problem is more like opening a crack in a person’s brain.
Looking out from this crack, the world does not necessarily become gentler, but it certainly becomes vaster—and more unsettling.

And that is exactly its most fascinating part.

So, when I left iPhone back then, it wasn't because I suddenly found Android to be comprehensively better. It was because iPhone consistently failed to meet some very practical needs.

Features like call recording, dual SIM support, long screenshots, and SMS export might not seem like high-end functionalities to many—some might even hesitate to bring them up in discussion. But for me, they were incredibly practical. Especially in work scenarios, it's often not about whether something can be done, but whether it can be done more conveniently. Add to that the fact that the battery life of the iPhone 7 Plus back then was far from satisfactory—it drained rapidly under heavy use. Playing a single round of PUBG could consume 25% of the battery, and using it a bit more meant scrambling for a charger. The Lightning port at the time also had to handle both charging and headphones, making daily use feel somewhat chaotic.

In that sense, my switch to Android back then wasn't because Android was already advanced; it was because iPhone simply wasn't convenient enough for some very specific, real-world needs.

But nine years have passed, and I recently bought an iPhone again, initially with a sense of anticipation for a long-awaited reunion. After all, Apple's progress in performance, imaging, display, and system smoothness over the years is well-known. Theoretically, today's iPhone 17 is nothing like the iPhone 7 from back then. Logically, it should be easier to use than before and shouldn't cause any significant discomfort.

I Don't Know How to Use iPhone Anymore

After getting the phone, I immediately logged into my Apple account, which I hadn't used in nine years. Looking at the account, my previously used iPhone 4, 5S, SE, 7 Plus, as well as long-idle iPad mini, iPad Pro, and Apple Watch 3 were all clearly listed. Some photos and messages left in iCloud automatically downloaded, bringing back a familiar feeling.

Then I opened the App Store, which recorded all my app downloads from 2011 to 2017. However, the unclickable download buttons next to them reminded me that these were all things of the past.

It was amidst this familiar sense of "reunion" that I started using it again.

But after actually using it for two days, I realized things weren't as smooth as I'd expected. I even felt a sense of helplessness, as if I didn't know how to use an iPhone anymore.

This feeling was strange. It wasn't that the phone had too many features I couldn't figure out; rather, many of the most basic, everyday operations—things I'd do dozens of times a day—felt awkward to me instinctively. Ultimately, it's not that I don't know how to use a smartphone; it's that my usage habits over these years have been completely reshaped by another system.

This feeling was strongest when it came to text copying.

iPhone's Basic Text Processing Capabilities Are Concerning

As I recall, the ability to long-press the screen on Android to copy text from apps or images likely started with Luo Yonghao's Smartisan OS, which he famously named the "Big Bang" feature. Since then, almost all domestic smartphones have built in this convenient text processing functionality.

Unfortunately, such an efficient method is still missing on iPhone to this day.

On iPhone, if you want to copy text from an app, perform a word search, or translate, you need to rely on a program called "Shortcuts" and edit commands in a pattern reminiscent of elementary school programming. For example, for translation, you first need to set up a command to take a screenshot, then extract text from the screenshot. If the text in the image isn't empty, you use the phone's built-in translation app to convert the text to Chinese, and finally use the built-in "Show Text" function to display the translated text.

This kind of operation significantly increases user difficulty, and I encountered issues multiple times while using it. For instance, when I bound this shortcut to a double-tap on the back of the phone, sometimes nothing happened. To solve this problem, I spent hours testing. This included situations where the double-tap on the back didn't work, but running it via Siri was fine; replacing "Show Text" with "Show Notification" allowed it to run normally; and after using "Show Notification" multiple times, error code prompts would pop up. In the end, just for this one "Screenshot Translation" shortcut, seamless collaboration between five apps within the phone is required: the screenshot program, text extraction program, translation program, text display program, and the Shortcuts app itself. If any one program's permissions or the program itself malfunctions, the entire process fails.

Seeing code error prompts pop up directly on the screen like this on iPhone is something I haven't encountered on a phone in many years. At least, during my use of Huawei, VIVO, and Smartisan phones, I never had such an experience.

iPhone's Input Method Is Hard to Describe

I don't remember what it was like typing on iPhone before. But after switching from Android to iPhone this time, I found that the current iPhone input method is incredibly difficult to use.

Before using iPhone, in recent years, the most difficult input method I'd experienced was Google's native Gboard. However, Gboard's difficulty mainly lies in hiding some special symbols behind conventional punctuation. For example, to input Chinese book title marks《》, you need to long-press the quotation marks to reveal them; to input symbols for RMB, Euro, or Pound, you need to long-press the dollar sign $ and then select. But this input method's difficulty is limited to these possibly less frequently used symbols.

On iPhone, similar shortcomings to Gboard's can be considered "niche experiences." The native iPhone input method requires clicking the "123" key in the lower left corner even for the most basic commas and periods. When chatting on WeChat and typing a simple sentence, I constantly have to switch between the Chinese and punctuation input interfaces. Especially when writing articles like this one, besides Chinese, I also need to input English and other symbols, including commonly used URL prefixes. This requires constantly clicking between the "Input Method Switch Icon" and the "123 Icon" in the lower left corner. The Chinese and English interfaces of the iPhone input method look almost identical, as do the Chinese and English punctuation interfaces. This often has me "pumping the bellows," clicking back and forth on those switch keys, unsure if I've even selected the correct interface.

Even more bizarre is the Chinese double-pinyin input method. I've long used the Natural Code double-pinyin scheme, so as soon as I got the phone, I instinctively looked for the corresponding option. After searching through the settings, I found that the iPhone's native double-pinyin schemes included Pinyin Jiajia, Sogou Double Pinyin, and Microsoft Double Pinyin, butJust not"Natural Code." My first thought was, "Looks like I'll have to install a third-party input method." Later, with a "nothing to lose" attitude, I tried a scheme called "Common Double Pinyin." After typing a bit, it felt strangely familiar. Upon careful testing, I finally realized that this so-called "Common Double Pinyin" was actually Natural Code.

This issue itself isn't a big deal; it's even a bit of dark humor. But it conversely shows that Apple hasn't delved as deeply into the details of Chinese input as domestic systems have. Many things that seem理所当然 to Chinese users still carries a sense of being taken for granted."good enough" detachment at Apple.

iPhone's Home Screen Icon Management Is Difficult to Use

Besides text processing and the input method, another point I found hard to adapt to is the iPhone's home screen layout and functionality.

First is that eternally unchanged 4-icon-per-row layout. After getting the phone, IOnce thought about where to change this layout. I searched around but couldn't find it. Then I searched on short video platforms and discovered that iPhone hasn't changed this layout in 20 years. From the first-generation iPhone's 3.5-inch screen to today's 6.9-inch large screen, it has never considered expanding the 4-column icon layout to 5 columns. For modern phones that often have 512GB/1TB storage and install hundreds of apps, a 4-column layout is too conservative and wasteful of space.

This feeling is like moving from a 70-square-meter old apartment to a 140-square-meter large house, only to find the living room still furnished with the small coffee table and side table from the rental days. It's not that they can't be used, but it just feels off.

Another issue is the inefficient performance of desktop icon organization.

Unlike Android, where you can typically select multiple icons simultaneously and merge them into a folder, iPhone still uses the most primitive method of operating on icons one by one. Maybe I'm getting old, but what's particularly uncomfortable during this process is that when I try to drag an app icon into another folder, that folderalways thought I'm trying to displace its position—it either moves forward or backward. A simple drag-and-drop operation sometimes takes over a dozen attempts to succeed. In the end, I'm not even sure how I managed to stuff them all in.

Some Operations Look Refined but Are Cumbersome to Use

Besides the interface layout, another aspect of iPhone that increasingly bothers me is that many operations still require users to "tap accurately."

The most typical example is browser tab operations. For instance, when using the Edge browser, if I want to close a tab, I have to precisely tap the small "x" in the upper right corner. This action itself isn't complicated, but its high frequency adds up. Opening and closing dozens of tabs a day, over time, reveals that this "must precisely hit a small area" interaction method is actually quite draining.

In contrast, many similar scenarios on Android have long moved towards "rough but efficient interaction." Tabs can be closed with a horizontal swipe; some background tasks can be cleared with a flick; even many notifications and floating cards can be operated with broader gestures. It might not be as neat as Apple's "tap a small icon to execute a command," but in the era of large screens and high-frequency operations, this interaction logic is clearly more convenient.

Operations like swiping from any edge of the screen towards the center on Android areonly limited to swiping up from the bottom on iPhone.

These issues, like the input method and text copying, aren'tfatal. But oncestacked up, they make it very apparent that iPhone's system still retains a relatively "antiquated" operational inertia. Since there's "Power Button + Volume Up" for screenshots, is it necessary to keep a "Screenshot" button in the Control Center? Since a single "HOME" button solved everything before, must all operations now still originate from the bottom of the phone? Since many practical functions can be set up via "Shortcuts," is it necessary to add them directly?

It certainly has its own elegance and sense of order. But sometimes, this elegance isn't free—it's paid for by users through more precise taps, online searches, and more restrained adaptation to its ways.

Apple's Charging Is Really Too Slow

Another issue falls into the category of experiences that were tolerable in the past but are increasingly hard to bear now: charging speed.

Honestly, if we go back to 2016 or 2017, while iPhone's slow charging was annoying, it didn't create a huge psychologicaldrop. Back then, many phones didn't have the tens or even hundreds of watts of fast charging capability common today. People's understanding of "charging" was more about plugging it in overnight or slowly topping up during the day.

But now it's different.

Over the past years, the progress of domestic Android in fast charging isn't just about improved technical specs; it has genuinely changed user habits. For me now, phone charging mainly happens in two scenarios: while driving and while at work. I don't always manage to fully charge the phone overnight. Sometimes it goes through the night with low battery, and I genuinely rely on those 10-20 minutes in the morning—while taking kids to school, commuting, or after arriving at work—to quickly boost the battery.

With this usage rhythm, fast charging is no longer some "enthusiast selling point"; it's a very practical, fundamental capability. With Android phones now, 10-20 minutes is often enough to restore a sense of security, at least knowing you don't have to constantly worry about battery life. But iPhone, to this day, still feelsquite slow in this regard. It's not that it doesn't charge, but that recharge speed, in today's environment, clearly lags behind the expectations I've developed over the past few years.

Ultimately, this might not entirely be due to Apple's lack of technology; it's more a difference in philosophy. Apple has never been keen on pushing fast charging to the extreme, seemingly preferring a conservative balance between battery lifespan, safety, and overall pace. The problem is, Chinese users'The pace of life is inherently more fragmented.. Often, phone charging doesn't happen during long, uninterrupted windows but is squeezed into variousspare moments. In such real-world scenarios, fast charging is no longer just a spec issue; it's a difference in usage patterns.

It's Not That iPhone Has Regressed, But That Domestic Android Has Advanced Too Much

ThinkingCalm down, today's iPhone is certainly not a bad phone. It's still exquisitely crafted, has a complete ecosystem, and maintains Apple's own advantages in imaging, performance, animations, and detail consistency. The problem isn't that it has regressed; it's that over these nine years, domestic Android has gone too far down another path.

Nine years ago, domestic Android was far from as mature as it is today. Back then, many systems were still catching up on smoothness,texture, and stability. Some seemingly cool features weren't necessarily truly usable. But over the past years, in the high-intensity competition of the Chinese market, domestic Android has been honed into developing another set of capabilities. They haven't just strengthened hardware or smoothed out the system; they've gradually turned manyWhat once seemed fragmented, peripheral, and even"not worth mentioning" high-frequency needs into fundamental capabilities.

During my use of this current iPhone, there are many issues I haven'tdiscussed in detail. For example, portrait photography, multi-functional screenshots, regional screen recording, data plan management, package delivery reminders, spam message blocking, real-time video translation, and the verythe hassle of downloading apps not available on the Chinese App Store.

Looking at these capabilities individually, none seemenough to decide on a phone's "high-end" status. But when they allstacked together, they create a very distinct difference in user experience. Domestic Android increasingly feels like a set of "out-of-the-box ready-to-use" high-density tools tailored for real life in China. iPhone, on the other hand, still feels more like plain water—if you want any flavor, you have to spend time figuring it out yourself, and you might still fail.

So, what surprised me most about switching back to iPhone this time wasn't that it's bad, but that Ifeeling a sense of unfamiliarity again in many of the most basic operations.

On the surface, it seems like I don't know how to use iPhone anymore. But upon careful thought, perhaps what has truly changed isn't me, nor is it just Apple. It's that over these nine years, both domestic Android and the usage environment for Chinese users have changed.

Many aspects that required compromise, adaptation, andPatience has long been smoothed out bit by bit on domestic phones.. When you turn back andreconfront the logic that the iPhone has always adhered to, it naturally feels out of sync.

These names are not necessarily neutral, but they generally fall within the conventional rhetorical framework used by modern nations. They emphasize order, freedom, and determination, serving both domestic public opinion mobilization and the construction of international narratives.

However, "Operation Epic Fury" is entirely different.

"Epic" is a term of historical scale, suggesting events at the level of civilization, while "fury" is an emotional expression. The combination of the two makes the naming of this operation more akin to religious or mythological narratives rather than a purely military action.

When a nation begins to define a real war as "epic," it is emphasizing not just geopolitics but historical significance.

This shift in naming directly reflects a profound change in the American mindset.

The Modern U.S. Military Is Leaning Toward Religious Influence

Modern Chinese perceptions of the U.S. military have two key reference points: the "cowardly but well-supplied" military during the Korean War, and the "high-tech, modernized force" seen in the Gulf War and the subsequent Iraq War.

But by 2026, it may be necessary to add another layer to this perception: the U.S. military may be evolving toward religious influence.

According to a report by The Guardian, within just a few days of the U.S. declaring war on Iran, numerous grassroots soldiers from over 50 U.S. military bases filed complaints with religious organizations. They claimed that their commanders told them the war against Iran was a "holy war" aimed at bringing about the "end of the world and the return of Jesus."

For instance, one complaint stated:

"This morning, our commander urged us during the combat readiness briefing not to 'fear' the ongoing operations in Iran." "He urged us to tell the troops that this is 'all part of God's divine plan,' specifically citing passages from the Book of Revelation about the end of the world and the imminent return of Jesus Christ. He said, 'President Trump has been anointed by Jesus (anointing is an ancient Hebrew religious ritual involving the application of holy oil to establish the divine authority and legitimacy of priests, kings, or prophets) to ignite the signal fire in Iran, trigger the end of the world, and mark his return to Earth.'"

But the issue goes beyond this.

If it were just a few officers using religious language in briefings, it could be dismissed as individual behavior. However, when similar complaints emerge from dozens of bases and religious organizations receive a large volume of feedback in a short period, it no longer appears as an isolated incident but rather a shift in atmosphere.

The U.S. military has long been seen as a model of a modern military not only because of its equipment and technological capabilities but also due to its institutionalized, professional, and de-ideologized self-image. Individuals may have religious beliefs, but the rationale for war must be based on national interests. This distinction is a fundamental principle of modern national militaries.

Now, when American grassroots soldiers hear terms like "end of the world," "return of Jesus," and "divine plan" in combat readiness briefings, these boundaries are beginning to blur.

The Pentagon's decision to refrain from public comment, citing the Uniform Code of Military Justice (UCMJ), underscores the sensitivity of the issue. A direct response would require addressing a fundamental question: Is the military becoming religiously influenced?

And as recent public reports indicate, this trend is not without precedent.

Last June, there were reports that U.S. Secretary of Defense Heggeses held weekly Bible study sessions at the White House, openly promoting the theological logic that "God blesses Israel's allies and curses Israel's enemies." At the same time, Heggeses integrated Christianity into the highest levels of the U.S. military, broadcasting monthly prayer meetings throughout the Pentagon, with content similarly emphasizing "God's command for America to support Israel." This phenomenon clearly goes beyond personal faith and enters the language system of state power structures.

From "Father-Son Relationship" to Theological History

In the Chinese internet context, people often jokingly refer to Israel as America's "father," with the U.S. "obeying Israel's every command." While this phrasing is clearly tongue-in-cheek, it is not entirely baseless.

The primary reason lies in Christianity's origins in Judaism, which holds a unique significance in theology.

Chinese society, with its long-standing atheistic tradition, has a relatively superficial understanding of religious history. But the analogy of a "father-son relationship" makes it easier to grasp.

In Christian tradition, Israel is not merely a modern nation. Jerusalem, Israel, and the Middle East hold dual symbolic significance in religious narratives: historical and eschatological. For the Western world, supporting Israel is not just a foreign policy but a religious duty.

Within this framework, conflicts in the Middle East are no longer seen as mere geopolitical struggles but as "part of the historical process."

This is why, when the Chinese-speaking world uses the "father-son relationship" to explain U.S.-Israel relations, despite its crude expression, it inadvertently touches upon a certain truth. Behind U.S.-Israel relations, there indeed exist cultural and religious bonds that transcend practical interest calculations.

Epic Fury or Impotent Rage?

Setting aside religious narratives, even without discussing the Book of Revelation, "divine plans," or the "holy wars" of both sides, the nature of this war is not complicated when viewed through the lens of international law and the principle of sovereignty.

When one sovereign state launches a military strike against another sovereign state, it is, in essence, an act of aggression.

No matter the codename, no matter the historical significance attached, no matter the moral veneer applied, this fact remains unchanged.

The foundational logic of the modern international order is the equality of sovereignty. Any unilateral use of force in the name of "justice," "civilization," or "historical mission" fundamentally violates this principle. The only differences lie in the grandiosity of the narrative and the strength of the allies.

Whether "Operation Epic Fury" will become an epic ultimately depends on history itself, not the will of those who named it.

But one thing is certain: When a nation begins to frequently use theological language to justify war, and when "end of the world" and "return" rhetoric emerges within the military for mobilization, it reflects more than just a military operation.

It reflects a shift in mindset.

This shift is more dangerous than the war itself.

Because once war is imbued with theological meaning, it is no longer constrained by practical boundaries. Once "fury" is rationalized, it becomes a tool of policy.

In the Chinese context, there is a term: "impotent rage."

If grand historical narratives ultimately fail to deliver tangible results, if the so-called "epic" turns out to be nothing more than an emotional display, then what this war leaves behind may not be an epic.

Instead, it may raise the question of whether, under the influence of impotent rage, even more dangerous actions will be taken to preserve the integrity of the narrative.

What is truly thought-provoking is the direction the comments took. A large number of comments focused on the issue of "slow-moving vehicles," almost forming a consensus: Low speed is the root cause of congestion and accidents; excessively low speed limits create risks; the key to easing congestion lies in increasing speed. Highways are called highways precisely because they shouldn't be slow.

This isn't just about highways; similar situations occur on national and provincial roads in China, especially on urban roads and expressways in urban-rural fringe areas, where the situation is even more complex. Large freight trucks often travel in clusters. These vehicles are heavy, carry substantial loads, and have long braking distances. If slow-moving vehicles appear ahead, it can easily lead to rear-end collisions and wave-like congestion. On these roads, slow-moving vehicles are perceived as "hazards"—not without reason but as an intuitive response to the actual driving environment. On highways, national roads, urban expressways, and urban-rural fringe areas, a tense relationship exists between low speeds and high-density transportation, leading people to instinctively react by calling for "speeding up."

However, this sentiment starkly contrasts with recent policy trends in some developed cities. For example, a few years ago, South Korea launched an "action to reduce traffic accident fatalities to zero," with one of the core measures being to lower vehicle speed limits. Last year, Helsinki, the capital of Finland, claimed to have achieved zero traffic deaths for a full year, also through systematic speed reductions and spatial redesign. Recently, Singapore proposed following Helsinki's example by comprehensively strengthening traffic governance, including measures such as lowering the legal blood alcohol limit and reducing speed limits.

On one side, developed countries are actively slowing down; on the other, there are calls in China to go faster. The difference is not merely a matter of differing perspectives but a practical conflict rooted in developmental stages and governance capacity.

I. Why Are Slow Vehicles Perceived as Hazards in China?

Physically, higher speeds mean greater collision kinetic energy—this is common sense. However, in the actual operating environment of Chinese roads, people have derived a different set of experiences.

In high-density traffic flows, any localized slow driving can trigger chain-reaction braking. Wave-like congestion can further evolve into high-risk areas for rear-end collisions, leading to slow vehicles being labeled as "creating danger."

A deeper reason lies in China's high dependence on traffic flow efficiency. As is well known, China has one of the world's largest holiday population migrations, and its logistics system supports nearly half of the global industrial capacity. Transportation authorities have consistently emphasized improving transportation efficiency and reducing transportation costs per unit of GDP.

Within this structure, "slowing down" vehicles means extending time in the industrial chain and amplifying pressure on the entire social system. Therefore, low speed is not merely seen as an individual behavioral issue but as a potential systemic risk source. Those "slow-moving vehicles" on highways, national roads, provincial roads, and expressways in urban-rural fringe areas are naturally regarded as genuine hazards in transportation.

Yet, paradoxically, in the everyday language of ordinary Chinese people—whether from elders' reminders, friends' advice, or company management—the parting words when seeing someone off are almost always: "Drive slowly!" or "Take it easy on the road!" This opposition between logic and system, the intertwining of perception and experience, is indeed highly complex.

II. Why Are Western Cities Choosing to "Slow Down" Instead?

In the West, reducing speed is not an emotional choice but a governance strategy based on economic development and the carrying capacity of social structures.

South Korea is the most typical case. I previously wrote an article briefly introducing South Korea's experience in traffic accident governance . Over a decade ago, South Korea still had the highest traffic fatality rate among OECD countries. However, over the past ten years, South Korea completed its economic transformation, entering the ranks of high-income countries. Society now has sufficient resources to invest in road governance, education, and infrastructure construction. This has made reducing urban speeds a feasible and systematic policy without causing economic efficiency to collapse. In April 2021, South Korea's revised Road Traffic Act implemented the "Safe Speed 5030 Policy," lowering speed limits on urban arterial roads to 50 km/h and on roads near residential areas and schools to 30 km/h.

Helsinki, the capital of Finland, widely praised in recent years, follows a similar path. Helsinki's Vision Zero strategy originates from a concept proposed by Sweden in the 1990s: "No one should die or be seriously injured in the traffic system due to foreseeable errors." The core logic of this concept is not simply to punish drivers or demand "zero mistakes" but to shift responsibility from "requiring perfect driving behavior from everyone" to urban planning, road engineering, and traffic management. It acknowledges that humans make mistakes (such as distraction, fatigue, or drunk driving) and thus must "forgive" these errors through multiple layers of protection (Safe System) to prevent fatal consequences. Under this philosophy, Helsinki has repeatedly lowered urban speed limits since the 1990s. After 2019, almost all urban and residential streets have speed limits of 30 km/h, with only a few arterial roads retaining limits of 40–50 km/h. Of course, besides speed limits, Finland also updated a series of supporting measures, including road design, vehicle technical standards, and citizen training, ultimately achieving the goal of "zero traffic deaths."

From this, we can see that the "slowness" pursued by developed countries through speed limits starkly contrasts with the reality on Chinese streets. In high-income countries like Western Europe, Northern Europe, Japan, and South Korea, slowing down means risks are systematically compressed; in China, slowing down may trigger new chain reactions of risk. Overall, economic foundations, population density, and logistics pressure determine the feasibility of slow-traffic policies.

The Outliers in the West

However, a closer look at OECD traffic accident data quickly reveals two puzzling countries: Germany and the United States.

Germany has long had one of the lowest traffic fatality rates in the OECD, at about 3.3 per 100,000 people. Yet, when people think of Germany, many immediately recall that it is a country with no speed limits on highways—cars can go as fast as they want on German autobahns, which has become an important symbol of German industrial culture. Besides highways, German urban roads, national roads, and rural roads have not adopted the "slow down for safety" approach seen in Northern Europe.

The United States presents another contrast. The U.S. has long ranked among the top in GDP per capita globally, with technological innovation and economic growth rates standing out among developed countries. It has a large population, strong consumption power, and medical resources far exceeding most nations. Yet, its traffic fatality rate is much higher than countries like Germany, Canada, and Australia—arguably the worst among developed countries.

One country is safe without speed limits; the other is highly developed yet dangerous.

If we simplistically equate "slowing down" with "safety," this difference cannot be explained.

Germany's low fatality rate is built upon a highly mature traffic system: strict driver's license exams, high-standard vehicle inspection systems, finely designed separation of freight and passenger traffic, well-maintained road surfaces and signage, and a stable culture of law-abiding behavior. Speed is enveloped within a highly standardized system.

The United States' problems, however, lie precisely at the structural level.

First, American cities are highly car-dependent. Many cities adopt low-density, sprawling development patterns with weak public transportation, long travel distances for residents, and driving as a rigid necessity. More driving naturally leads to higher exposure to risk.

Second, road structures vary greatly. While highway standards are high, many roads in urban-rural fringe areas, interstate highways, and rural roads have inconsistent design standards, complex intersections, and insufficient pedestrian protection facilities.

Third, large-displacement pickup trucks and SUVs dominate the American market. These vehicles are larger and heavier, posing higher fatality risks to pedestrians and smaller vehicles in collisions.

Fourth, complex social structures. Issues like wealth inequality, disparities in healthcare access, drunk driving, and substance abuse in the U.S. collectively impact traffic safety.

In other words, in the United States, whether speed limits exist may not matter much. What truly determines traffic safety levels is never speed limits themselves but the overall maturity of the traffic system. When a society's traffic system is sufficiently stable, speed is merely a variable; when the system has structural shortcomings, speed amplifies all risks.

III. Learning from Successes and Failures

From the experiences and lessons of developed countries, traffic safety is never the result of a single policy but the product of a combination of institutions.

The practices of South Korea and Helsinki show us that when a country completes industrial upgrading, has sufficient fiscal capacity, and stable social consensus, speed limits can become part of systemic optimization. It is not simply about "reducing numbers" but is implemented alongside road engineering improvements, driver education systems, and enhanced enforcement capabilities. Slowing down is an active choice within a mature system.

However, the contrast between Germany and the United States reminds us that speed limits are not a universal solution. Germany's low fatality rate stems from long-accumulated institutional capacity, while the United States' high fatality rate shows that even with GDP per capita ranking among the world's highest, if urban structures are highly car-dependent, road standards vary greatly between urban and rural areas, and enforcement and cultural consensus are difficult to unify, fatality rates can remain high.

This means we cannot simply replicate the "comprehensive 30 km/h" model of Northern Europe nor fantasize that "greater economic development naturally leads to more safety." What we can truly learn from are those universal institutional details.

For example: strengthening 30 km/h controls around schools and aging communities; reducing vehicle-pedestrian conflicts through physical separation measures; implementing high-intensity, sustained crackdowns on drunk and dangerous driving; mandating seatbelt use for all seats, not just the front; establishing巡回教育机制针对老年驾驶员, even setting up specialized training centers; optimizing pedestrian crossing facilities and signal timing in areas with dense elderly populations; implementing centralized, unified management of school buses for children to reduce risks from decentralized, socialized transportation.

These measures are not grand "speed reduction revolutions" but refined repairs targeting risk structures. Their common feature is that they do not rely on slowing down the overall societal pace nor sacrifice efficiency as a prerequisite, yet they can significantly reduce traffic safety risks.

Simultaneously, China has unique advantages in the adoption rate of new energy vehicles, penetration of assisted driving technologies, and the construction of intelligent transportation infrastructure. If institutional development and technological application can synergize, certain risks might even be preemptively resolved through technological pathways.

IV. E-bikes: A Chinese-Style Dilemma

If car speed limits are a topic the world can discuss together, then e-bikes might be a uniquely Chinese problem.

For instance, my family once bought a certain brand of e-bike. Whether set to gear one, two, or three, the speedometer always showed a maximum speed of 25 km/h. However, in reality, the actual speed at full throttle in first gear versus third gear was completely different. On e-commerce platforms, the same model of e-bike is blatantly advertised with a maximum speed of 55 km/h and a weight of 104 kg, clearly not aligning with the national standard for "electric bicycles." Of course, when selling, they "technically" label the product as an "electric motorcycle." Yet, walking into any offline store reveals that these vehicles are sold as "electric bicycles" and registered with white license plates.

This is not merely a matter of marketing rhetoric but a microcosm of practical compromise.

Ordinary families need vehicles that can carry children, climb slopes, and commute; e-bike manufacturers need sales; and regulators need to maintain order.

Here, speed is not just a safety parameter but also a basic livelihood issue, a guarantee of fundamental life efficiency.

I have been following traffic governance in recent years and have written several related articles. Last year, during training at a police academy, I had in-depth discussions with professors from public security and traffic management colleges about e-bike governance. During the process, I proposed an improvement idea: for example, in e-bike management, we could refer to Taiwan Province's experience with light motorcycles and design a "two-stage turning" rule for e-bikes to reduce turning accidents, which have the highest accident rates in reality.

The logic of "two-stage turning" is quite simple: avoid interweaving conflicts with motor vehicles during the same turning moment. Instead, first go straight to the opposite side of the intersection, wait for the next signal to proceed, and then complete the turn. It sacrifices a few seconds but significantly reduces risk. In areas with high motorcycle density, this is a long-verified method of risk diversion.

However, the professor offered a more realistic assessment: the biggest obstacle currently is not the technical solution itself but the lack of regulatory consensus.

The primary contradiction lies in legal classification. According to current laws, e-bikes are not allowed on motor vehicle lanes; they should travel on non-motorized lanes or pedestrian systems. Legally, this definition is very clear. But reality is far more complex than theprovision. A large number of e-bikes routinely travel on motor vehicle lanes, parallel with cars, weaving through traffic, and even rushing at intersections. In other words, the actual operating state itself exists in a gray area of "widespread violation."

And this gray area is not entirely due to riders' subjective choices.

I once commuted by e-bike for a period. On a so-called "non-motorized lane" of just one kilometer, more than half was marked as roadside parking spaces; either occupied by motor vehicles or with illegal parking at intersections. The remaining parts were narrow and discontinuous, even interwoven with bus stops and temporary unloading zones. If I strictly followed the rule of "not entering motor vehicle lanes," I often couldn't move at all.

Thus, a dilemma arises: the law requires e-bikes to use non-motorized lanes, but non-motorized lanes themselves are compressed, encroached upon, and marginalized. When e-bikes enter motor vehicle lanes, they are deemedViolation.

The core of this issue still lies in: are e-bikes considered "motor vehicles"?

If they are motor vehicles, that would mean requiring driver's license exams, mandatory insurance, stricter production standards, and legal responsibilities matching those of motor vehicles.

But if they are not motor vehicles, then they must operate within the non-motorized system.

However, the reality is that the weight, acceleration, and maximum speed of many current e-bikes are already close to those of light motorcycles; their daily use scenarios for families also more closely resemble "low-cost car substitutes." Functionally, they are close to motor vehicles; legally, they are classified as non-motorized; and in terms of infrastructure, they lack completesafeguard.

Thus, there is a misalignment between institutional classification, road design, and actualDemand.

This also explains why technical optimization solutions like "two-stage turning" are difficult to implement. Because their premise is acknowledging e-bikes as a stable category of traffic participants and designing clear, enforceable rules for them. The current institutional state is more like a temporary compromise: tacitly acknowledging their existence without fully creating institutional space for them.

The professor's point was straightforward: before legal classification is clarified, non-motorized lanecoverage is insufficient, and enforcement capacity and social consensus are unified,Merely discussing technical optimization often remains theoretical..

This is not a lack of solutions but a lack of a social environment capable ofcarrying those solutions.

It was this rural journey that gave me a new perspective on Hengyang and the southern Hunan region.

The Mountain Roads Weren't Difficult

For over a decade, I’ve mostly traveled between Hunan and Guangdong via expressways, with only occasional trips on national highways. For instance, during the 2019 Spring Festival, to avoid congestion on the Erguang Expressway, I took G107, G358, and S260 between Linwu and Sihui. What left the deepest impression then was that the national and provincial roads in Guangdong were generally in good condition, with many stretches feeling almost like "no-man's-land"—few cars, few people, making risks more manageable. The only issue was the many mountain curves, which made nighttime driving stressful.

The 200-plus kilometers of mountain roads I took after exiting the expressway at Changning were similar. Most were undivided cement roads, but driving during the day wasn’t particularly challenging. After all, the last few hundred meters of cement road to my hometown is only a little over two meters wide.

In a sense, the mountain roads weren’t the problem, nor was the road condition. What really caught my attention was the appearance of the towns along the way.

The Complicated Face of Rural Hengyang

Around the year 2000, when I had just started middle school, I visited my third aunt for New Year’s greetings. Her home was in a Hengyang township bordering our county. It seemed more mountainous, remote, and poorer than my hometown, yet the roads were noticeably better.

At the time, my third aunt said something I’ve always remembered. She said Hengyang was an "industrial county," while our Shuangfeng was an agricultural county, so Hengyang had better roads and a stronger economy, something Shuangfeng couldn’t match for a long time.

Back then, we almost always went to Hengyang for everything: taking buses to Hengyang stations, seeing doctors at Hengyang hospitals, getting ultrasounds during pregnancy in Hengyang, shopping in Hengyang, attending university in either Changsha or Hengyang, and even hiring monks for religious ceremonies—those from Nanyue Mountain were considered more "authentic."

In my mind, Hengyang was the default "central city" for our surrounding area.

In contrast, Loudi, the prefecture-level city that Shuangfeng nominally belongs to, has barely registered in my life. In over thirty years, I’ve only been to downtown Loudi once.

But this time, traveling south from Changning through the town centers of Pengtang Township, Xiling Town, and Baisha Town, I felt a clear sense ofdrop(disparity).

The roads were worn, the streets messy, and the overall appearance seemed stuck in a certain era.

Most storefronts were traditional variety stores, small restaurants, and pool halls. At a glance, it was hard to find any recently renovated facades. Shops with uniform signage, floor-to-ceiling glass, LED-lit signs, and minimalist styles common in cities were almost nowhere to be seen.

Most were just ground floors of self-built houses converted into shops—rolling shutters pulled up, a few shelves arranged, and business began. There was no sense of design or branding.

It felt strange.

As if time around 2010 had been paused.

A Puzzling Comparison

Since childhood, I’ve heard that places like Changning and Leiyang were more developed than Shuangfeng, being established county-level cities. Logically, living standards should be higher, and commerce more vibrant.

But the reality didn’t match this perception.

Even though my hometown is in one of Shuangfeng’s most remote townships, the town center still has milk tea shops, snack stores, bakeries, and fast-food outlets. Brands found in cities, like Snack You, Mixue Ice City, and burger and fried chicken chains, are basically all there. During the Spring Festival, I even waited over an hour with my child to buy milk tea.

Barbecue stalls are even more common. Over a dozen operate year-round, scattered around the town center and surrounding settlements. At night, you can see "smoky haze" filling the streets, with young people gathered around chatting.

Yet on the streets of Changning, I hardly saw any barbecue stalls. It wasn’t just quiet; it lacked that bustlingthe smell of cooking(lively atmosphere).

I initially thought it might be a population issue.

But checking the data made this seem unlikely. From 2010 to 2020, Changning’s population decreased by only a little over 20,000, while Shuangfeng’s dropped by over 100,000. By the numbers, Shuangfeng should appear more "declined."

But the town streets presented a completely different picture.

The Matter of Rural Housing

Another detail.

Before the Spring Festival, a friend from Chenzhou remarked to me withsigh with emotion(emotion) that while passing through Loudi, they saw densely packed villas lining the rural roads, thinking our rural areas were doing quite well.

I was skeptical at the time, feeling it might be "mutual flattery." How could Loudi, a hilly, mountainous area in central Hunan that seems "unloved by both province and prefecture," compare favorably to theirthe backyard garden of the Pearl River Delta(Pearl River Delta backyard)?

But after traveling through Hengyang and Chenzhou this time, I realized they might have been right.

In many villages I passed through in Changning, Guiyang, and Jiahe, I noticed very few newly built modern villas. In some villages with dozens of houses, you couldn’t find a single decent three-story Western-style house.

In Shuangfeng, it’s hard to drive a few hundred meters without seeing a villa with a distinctly modern-style gatehouse. Some villages have almost entire rows rebuilt.

More importantly, this renewal isn’t just in the villages but also in the towns. Town streets are being updated, and village houses are being renewed, simultaneously.

This leaves me somewhat perplexed.

Shuangfeng has more obvious population outflow yet appears more "modern" in town appearance and rural housing. Changning has relatively stable population numbers, but its towns and villages seem more "stagnant."

Where has the money gone? Where have the people gone? Where is the drive for renewal coming from?

I don’t really have answers.

Another Rural Scene in Chenzhou

Heading south from Changning into Guiyang, my expectations had already lowered, so the situation didn’t seem as bad.

This stretch passed through Qiaoshi Township, Heping Town, Chunlingjiang Town, and Haotang Town in Guiyang County; Puman Township and Xinglang Town in Jiahe County; and Maishi Town in Linwu County. Among these, Xinglang Town, being close to Jiahe County’s urban area, had noticeable industrial zones and standard urban roads. More importantly, I could clearly sense a trend toward "clustered settlement."

Chenzhou’s rural areas aren’t as severely scattered as those in Loudi and Hengyang. The roadsides aren’t lined with sporadically placed self-built houses but rather more sizable villages, much like in Guangdong. Some villages even have a unified style, resembling small communities.

Agricultural production also appears more systematic. Cash crops like tobacco, tea oil, and citrus are grown in concentrated areas. Even on Guiyang’s "Lotus Avenue," I saw many peculiar "family farms" by the roadside, giving a sense of large-scale agricultural production reminiscent of Northeast China. The feeling was that Chenzhou’s rural land seems to be utilized in an organized manner rather than fragmented farming.

This might reflect two different development paths. The rural areas in Hengyang seem to suffer from the aftermath of industrialUpdate(renewal) following early industrialization driven by mineral resources. In contrast, Chenzhou is pursuing agricultural scale, county-level industrial support, andconnecting with the Pearl River Delta market(connecting with the Pearl River Delta market).

Multiple Facets of Regional Differences

The biggest takeaway from this detour isn’t a simple judgment of which place is better, but the clear differentiation in developmentrhythm(pace) and appearance across different parts of Hunan.

Against the same backdrop of outbound migration and population decline, some townships appear active in renewing housing and streets, while others seem more like they’ve been paused.

Behind this could be differences in local preferences, cultural psychology,inertia(inertia) of industrial paths, or evensubtle difference(subtle differences) in local governance and planning.

I also tried to find an answer I could accept andwent out of my way to look it up(specifically looked up) the 2024 statistical bulletins for several counties. Comparing the data actually clarified manydoubt(doubts).

For example, Changning, as an established county-level city, hassignificantly higher than(significantly higher) registered population,permanent resident population(permanent resident population),urbanization rate(urbanization rate),total retail sales of consumer goods(total retail sales of consumer goods), andbuilt-up urban area(urban built-up area) compared to Shuangfeng. Its industry is supported byleading enterprise(leading enterprises) like Wukuang Copper and Zhuye Nonferrous Metals, giving it a larger economic scale and a more complete urban framework.

Shuangfeng’s characteristics, however, lie in morefine structure(subtle structures). It has higher numbers and coverage rates forUrban Employee Pension Insurance(urban employee pension insurance),Employee health insurance(employee medical insurance), andwork-related injury insurance(work injury insurance). Although its county town’s built-up area isn’t large, theprivate sector(private economy) is active. Nearly one-third of the county town’s area is occupied by private manufacturing enterprises, quietly changing the traditionalimpression(impression) of an "agricultural county." Particularly noteworthy are thedeposit and loan balances(deposit and loan balances). Shuangfeng’s total deposits andhousehold deposits(household deposits) are higher than Changning’s,perfectly aligned(precisely matching) the real rural ecosystem: money stays in people’s hands, used more for building houses and township consumption, hence livelier streets and more active rural renewal.

The real difference between the two places might not be in strength or weakness, but inpath(path). Changning follows a path of big cities, big industry, andmass concentration(high concentration). Shuangfeng, despite population outflow, hasvibrant civil society(vibrant private sector),township self-motivation(township self-drive), andsmall-scale, dense, and vibrant community life(small, dense, livelihood-oriented prosperity).

Traveling through the landscapes of southern Hunan, I’ve become increasingly convinced: judging whether a place hasvitality(vitality) shouldn’t rely solely on GDP orsize of the urban area(urban area size). One must also look at whether towns havethe smell of cooking(bustling life), whether streets are being updated, whether ordinary people are willing to spend money locally and build their homes. Some places may not look "new" but are operating steadily. Some places may have declining populations but are modernizing quietly in their own way.

The rural areas of southern Hunan are never a single template. They are simply walking different paths toward the same destination called "Life is getting better" (life getting better).

Undoubtedly, the main reason for this severe congestion was the higher traffic density. According to Hunan Daily, on the sixth day of the Lunar New Year, the traffic volume on Hunan's expressways increased by 18.6% compared to the historical peak, with the single-day cross-province peak reaching 1.85 million vehicles.

There are only four expressways from Hunan to Guangdong, from west to east: Erguang, Xuguang, Jinggang'ao, and Wushen. Combined, these four expressways have eight lanes, with a maximum daily capacity of around 400,000 vehicles. This capacity has been insufficient in previous years, and with an additional 18% increase in traffic this year, congestion was inevitable. As a result, I had been driving for 10.5 hours and still hadn't left Hunan.

Fortunately, after leaving Hunan, the situation seemed to ease slightly. The four expressways from Hunan to Guangdong connect to more expressways in Guangdong, mainly adding Guanglian and Leguang expressways. However, the Erguang Expressway in the west and the Wushen Expressway in the east also need to handle traffic from Guangxi and Jiangxi, which is why I haven't dared to take the Erguang Expressway for many years. The section from Huaiji to Sihui on the Erguang Expressway consistently ranks among the most congested expressway sections nationwide during holidays.

The most unusual aspect of this congestion was the so-called "tidal lane" on the Qingyuan section of the Xuguang Expressway. While heading south to Guangdong, I could borrow the northbound lane of the opposite direction. However, I didn't have a great impression of this experience because, during peak traffic, this "tidal lane" clearly became "one of the sources of congestion."

The "tidal lane" I used was as shown in the image above: a two-lane expressway in one direction, where the fast lane could borrow the fast lane of the opposite direction. However, the actual experience revealed three obvious issues:

1. The Source of Congestion Is at the Diversion Point
   It wasn't until I reached the diversion point of the "tidal lane" that I realized the several-kilometer-long congestion earlier was caused by vehicles slowing down to around 20 km/h when the southbound fast lane was diverted to the opposite lane using traffic cones. Some vehicles, upon seeing the need to turn left into the opposite lane, stopped abruptly and signaled to change to the slow lane on the right, further worsening the congestion.

2. Slow Overall Traffic in the Tidal Lane
   After turning left into the "tidal lane," I realized that the fast lane I was originally in had effectively become a slow lane, while the slow lane on the right, after passing the diversion point, turned into the fast lane. The reason was that the "tidal lane" I used was originally the fast lane of the opposite direction, with traffic cones on the left and guardrails on the right, making it feel like driving on a narrow road just over 3 meters wide. It was naturally difficult to speed up, and with the added concern of vehicles in the original northbound slow lane separated by cones—especially at night with potential high beams—driving became even slower. After driving for a while, I encountered another standstill ahead, only to realize later that it was because vehicles had to turn right back into the original southbound lane. This secondary merging point caused another round of slow lane changes.

3. Equivalent to an Extra Parking Lot During Heavy Traffic
   Reflecting on this "tidal lane" experience, it felt like a two-lane road temporarily expanded to three lanes, only to revert to two lanes later. If there was congestion ahead at the point where the road narrowed back to two lanes, this section essentially functioned as an "expressway parking lot," doing little to alleviate overall congestion. As long as traffic density didn't ease, the number of lanes ahead didn't increase, and vehicles in the tidal lane didn't exit the expressway, congestion was inevitable.

Overall, though, such changes are undoubtedly a step in the right direction. At the very least, they show that the authorities are trying to find solutions within existing constraints.

Personally, I believe that for expressway "tidal lanes" to be truly effective, it's crucial to avoid pitfalls like "long-distance lane borrowing, cone isolation, and concentrated merging."

First, focus on bottleneck sections. For example, address "phantom traffic jams" near service areas or tunnels by allowing vehicles entering service areas to queue while letting others bypass the congested section.

Second, implement solid barriers. Replace traffic cones with movable steel guardrails, and split merging points into 2–3 staged entrances with sufficient acceleration lanes. This would eliminate slowdowns at diversion points and secondary merging congestion, making tidal lanes truly increase capacity rather than add to the problem.

Third, improve signage. During my experience, I entered the tidal lane almost blindly, only realizing the situation at the last moment when it was too late to change lanes. As mentioned earlier, the tidal lane itself was slow, while vehicles in the original slow lane, after our fast lane diverted to the tidal lane, sped ahead like wild horses, briefly enjoying two lanes. Unfortunately, they still encountered congestion at the secondary merging point ahead, as everyone had to merge slowly for safety, especially at night.

When a language model demonstrates "zero errors" in a formal reasoning experiment, it is no longer just a technical metric but a signal at the institutional level.

Even more intriguing is that the release of this paper coincided almost exactly with the one-year anniversary of DeepSeek's launch of its reasoning AI, DeepSeek R1. That moment could almost be seen as the "inflection point" when various industries in China began to seriously acknowledge AI's reasoning capabilities. A year later, the technological curve has proven far steeper than most anticipated. Even legal professionals who were once highly skeptical of AI now have to face reality.

A widely circulated saying may not be an exaggeration: AI will not replace lawyers, but lawyers who use AI are replacing those who do not.

Silicon Formalism

The title of this article, "Silicon Formalism," is a direct translation of "Silicon-based formalism." However, the term "formalism" here is not the negative connotation of "going through the motions" as in Chinese colloquial usage. Instead, it refers to the jurisprudential sense of prioritizing rules and structured reasoning, akin to "rule-based jurisprudence" or Max Weber's "formal rationality."

What truly deserves repeated reflection in this paper is not that "GPT-5 is smarter," but that it is more consistent. The experiment itself is not overly complex: switching between rules and standards, enhancing or diminishing the sympathy for the parties in the narrative, and creating differences in legal consequences under different state laws. Human judges, influenced by these variables, are swayed by contextual factors, while the model strictly derives conclusions based on the rule structure.

What does this mean? It means that in scenarios with clear rules, machines are closer to the "ideal formalist judge" than humans. They do not deviate due to sympathy, alter their application path due to narrative techniques, or emotionally stretch standard clauses. We have always said that law is not about 0s and 1s. Conflicts in evidence, value trade-offs, and situational judgments inherently introduce shades of gray into legal practice. But when a model demonstrates zero errors at the level of rule application, we must reconsider whether, in a significant proportion of cases, the law can indeed be highly formalized.

If the answer is "partially yes," then so-called "silicon formalism" is not merely a technological phenomenon but an institutional issue. Are the "errors" of human judges a necessary flexibility of the system or a manifestation of human cognitive limitations? When machines eliminate errors, are we losing human touch or approaching the rules themselves?

Where Is the Real Inflection Point?

The inflection point for AI in the legal industry does not lie in whether models possess astonishing capabilities, but in whether they have begun to alter the actual structure of legal production.

1. Document and Research Level: The Inflection Point Has Already Occurred

In scenarios such as contract review, clause comparison, due diligence summaries, and case summarization, large language models have already entered the actual production workflow. Legal tech companies like Harvey AI, which provide customized systems for law firms based on the GPT series of models, have been formally deployed and are being paid for by numerous top law firms globally. Harvey AI's valuation has reached $11 billion. Lawyers use AI within internal systems for contract analysis, legal research, and drafting initial versions of documents, establishing corresponding usage norms and risk control processes. This is no longer an experimental "toy" but a production tool operating within frameworks of billing, compliance, and internal controls.

Changes in corporate legal departments are equally evident. An increasing number of companies are embedding AI models into their internal knowledge bases for quickly retrieving compliance clauses, summarizing regulatory changes, and generating risk alerts. The starting point for legal text production is shifting from "manual drafting" to "machine-generated draft + human revision." The inflection point at this level has, in fact, already occurred.

2. Legal Reasoning Level: The Inflection Point Is Approaching

The zero-error results demonstrated in the paper indicate that in scenarios with clear rules and disputes focused on legal application, models can achieve consistency that matches or even exceeds average human judgment. If this capability proves consistently stable across different types of cases, then the handling logic for some cases will be reshaped. The value of lawyers and judges will increasingly lie in fact construction, evidence examination, and value judgment, rather than mere rule matching.

Of course, this does not mean AI will directly replace legal professionals. Instead, it signifies a shift in the professional structure. When rule application capabilities can be highly standardized, the hierarchy of skills within the legal profession will change. Those who are better at leveraging models to improve efficiency will gain a competitive edge.

3. Institutional and Power Level: The Inflection Point Has Not Yet Arrived

Admittedly, law is not merely about rule application. How issues are framed, how cases are narrated, and how standards are defined still remain in human hands. Models can only reason within given frameworks; they cannot determine the frameworks themselves. The legitimacy of judicial authority does not stem solely from reasoning accuracy but from procedure, credibility, and accountability mechanisms.

Currently, whether in the United States or China, models have not entered the formal structure of adjudicative power. They influence production processes, not power allocation. A true institutional inflection point must involve shifts in responsibility attribution and decision-making authority, and this has not yet happened.

From Hallucination Anxiety to Capability Stratification

A year ago, our concerns centered on AI hallucinations, erroneous citations of legal provisions, and model confusion in the face of complex evidence. Today, the direction of anxiety is shifting. We are beginning to discuss whether AI is more stable, less influenced by emotions, and more consistent than humans in certain domains.

The legal system has long oscillated between rules and standards. Rules provide certainty; standards provide flexibility. The "errors" of human judges and lawyers are sometimes the very source of discretionary space. The advantage of models lies precisely in eliminating errors. When rules are clear, they do not get fatigued, distracted, or swayed by narratives.

Therefore, the so-called inflection point is not about whether AI will replace judges or lawyers. In the short term, this scenario is highly unlikely. What is truly changing is the capability structure. When rule application, research summarization, and initial document drafting can be handled by models, legal professionals' time and value will increasingly shift toward fact organization, strategy design, and value judgment. Those who still spend time on automatable tasks will gradually lose their competitive advantage.

Perhaps the inflection point has already arrived. But what it changes is not the number of positions but the boundaries of efficiency and the stratification of capabilities. We have moved from "can it do it?" to "does it do it more consistently?" This, in itself, is a sign of the times.

Regardless of whether this term is exaggerated, the phenomenon it points to is all too familiar: in many Hunan official families, "losing control and failing to educate" their children has become a frequent issue. Many officials are not lacking in personal cultivation, work ability, or even their advocacy of family ethos, yet they repeatedly fail when it comes to their children.

As a Hunanese, I have many relatives and friends working within the system, most of whom are not in high-ranking positions. These families spare no effort in educating their children, investing time, money, and energy, but few truly succeed in keeping their children under control and within boundaries. This phenomenon is not an isolated issue but a recurring structural dilemma.

What puzzles me even more is that after leaving Hunan nearly two decades ago and observing official families in different regions, I found that the phenomenon of "failing to control children" is particularly concentrated and typical in Hunan. This is clearly difficult to explain away as "a few individuals with poor character."

The Traditional Family Ethos Culture of Hunan

When discussing Hunan's family ethos, one cannot avoid mentioning Zeng Guofan.

As a prominent official in the late Qing Dynasty, Zeng Guofan is not only an important symbol of Hunan culture but also a representative figure in Chinese family history for his family precepts. He repeatedly emphasized "diligence, frugality, self-discipline, and avoiding extravagance," stating that "leading troops breeds arrogance, and leading greed breeds luxury." He strictly forbade his descendants from interfering in local affairs or leveraging influence for personal gain. In his family letters, his demands for his children to "be cautious in solitude," "fear power," and "know their place" were almost harsh.

This family ethos was not empty talk. Zeng's descendants maintained a low profile and restraint for generations, rarely engaging in reckless behavior, and the family's reputation endured for multiple generations. Behind this was not Zeng Guofan's "perfect" personal morality but his clear understanding that for a family, power is never a talisman but a high-risk variable.

In the context of the late Qing Dynasty, power was relatively decentralized, and local society exerted strong public opinion constraints on official families. Once a family "lost its virtue," the entire clan often paid the price. In such an environment, family ethos was not just a moral issue but a survival strategy.

The Phenomenon of "Familial Corruption" in Contemporary Hunan

However, in contemporary times, Hunan's official circles present a different picture.

In recent years, several high-ranking Hunan officials have been explicitly cited in their disciplinary notices for "failing to manage and educate family members," "allowing spouses or children to leverage authority or influence for personal gain," and "engaging in large-scale familial corruption." Examples include Peng Guofu, former Vice Chairman of the Hunan Provincial Committee of the Chinese People's Political Consultative Conference (CPPCC); Li Weiwei, former Chairwoman of the Hunan Provincial Committee of the CPPCC; and Yi Pengfei, former Vice Chairman of the Hunan Provincial Committee of the CPPCC.

In places like Chenzhou, multiple consecutive municipal party secretaries have been implicated in cases where their spouses or children were deeply involved in engineering, personnel, and project operations. These cases are not isolated incidents but exhibit highly similar patterns: the father exercises power openly, while the children cash in behind the scenes.

Over time, a "semi-public" unwritten rule has gradually formed: to smoothly access certain resources, one must first "grease the wheels" through these privileged channels. This is the real-world soil in which rumors of the "Seven Princes of Hunan" thrive.

Why Has Traditional Family Ethos "Failed" in Contemporary Times?

This is not a sign of Hunan culture's "decline" but rather a reflection of fundamental changes in the power environment on which culture depends.

First, power is highly concentrated, and resources have become immensely profitable.
Since the reform and opening-up, Hunan has seen rapid expansion in capital volumes in areas such as infrastructure, real estate, mining, and finance, with administrative power playing an extremely high role in resource allocation. For children, as long as they can "leverage influence," there is almost no real market risk. This kind of "risk-free business" did not exist in Zeng Guofan's era.

Second, the logic of official careers has shifted from "self-cultivation" to "circles."
The "Xiang Army" spirit in traditional Huxiang culture emphasized responsibility, accountability, and discipline. However, in modern officialdom, it has gradually evolved into a more pragmatic network of "fellow townspeople—circles—resources." Immersed in the machinations of official careers, parents expose their children to "how power translates into opportunities" rather than "why power should be feared."

Third, intergenerational disconnection and demonstration effects compound the problem.
Many officials themselves were not without principles in their early years but gradually compromised in the power environment, turning a blind eye to their children's actions. Once someone around them "succeeds by relying on their father," this path is quickly replicated, creating a herd mentality until systemic control is lost.

Fourth, institutional oversight has long lagged, with family ties serving as a buffer for corruption.
For a considerable period, restrictions on relatives engaging in business, property declarations, and recusal systems were loosely enforced, making the family the safest "haven" for power. The so-called "family bond coercion" is not a problem with emotions themselves but rather the failure of institutions to sever the transmission path between family ties and power.

Conclusion

Therefore, the issue has never been that "Hunanese do not value family ethos." On the contrary, it is precisely because Hunan places great emphasis on family ethos but overly moralizes and de-institutionalizes it that it fails completely in the face of real power.

The lesson from Zeng Guofan is not merely "to emphasize family education" but to maintain a structural reverence for power. He understood that if family ethos cannot be supported by institutions, it will ultimately become nothing more than self-consoling rhetoric.

Today, when revisiting family ethos, what truly needs rebuilding is not slogans but boundaries:
the boundary between family ties and power, the boundary between family and public resources, and the boundary between moral self-discipline and institutional constraints.

If there is anything worth inheriting from Huxiang culture, it might be Zeng Guofan's oft-ignored statement: "Once you enter officialdom, you are a person awaiting judgment."

This statement should have been said to one's sons first.

In Hunan, this has almost become an unwritten rule: On New Year's Eve at noon, sons-in-law go to their mother-in-law's home for a meal; then, they rush back to their own home for the New Year's Eve dinner in the evening. It's not something any family explicitly emphasizes, but rather "everyone does it this way." Not following suit would make you seem inconsiderate.

The question is, is this seemingly "traditional" arrangement truly an old custom?

A New Tradition That Looks Very "Traditional"

If judged solely by real-life observations, it's easy to assume this is a long-standing local custom. However, when I browsed through various local chronicles in the Hunan Digital Local Chronicles Archive, I discovered a counterintuitive phenomenon. In Hunan's local chronicles from the 1970s and 1980s to around the 2000s, the descriptions of New Year's Eve customs are highly consistent, yet they hardly leave any room for sons-in-law.

Whether in the chronicles of Shuangfeng, Xiangxiang, Hengshan, Changning, Yiyang, You County, or Shaoshan, the core of New Year's Eve descriptions remains "returning to one's own home," "family reunion," and "worshipping ancestors and staying up late." New Year's Eve is regarded as the most important family gathering of the year, where those away from home must return to their own families. The reunion dinner, staying up late, and giving red envelopes all revolve around the same family system.

For this reason, if the custom of sons-in-law visiting their mother-in-law's home for New Year were a stable and widely recognized practice, it would be unlikely to be entirely absent from so many local chronicles.

The "Missing" Sons-in-Law in Local Chronicles

However, this does not mean that traditional society ignored the mother-in-law's family. On the contrary, local chronicles provide very clear and consistent records of "the second day of the New Year."

Shuangfeng, Shaoshan, and Xiangxiang all have the saying, "On the first day, sons; on the second day, sons-in-law; on the third and fourth days, visiting neighbors." The Yiyang County Chronicle explicitly states, "On the first day, sons; on the second day, sons-in-law." Hengshan has, "On the first day, sons; on the second day, sons-in-law; on the third and fourth days, visiting maternal relatives." Leiyang is even more direct: "On the first day, sons; on the second day, sons-in-law." In other words, in the traditional festival sequence, the "main stage" for sons-in-law is not New Year's Eve but the second day of the New Year.

The logic behind this arrangement is actually quite clear. New Year's Eve is the most solemn and core ritual time within the clan, emphasizing returning to one's own family and worshipping ancestors. Relationships with the maternal family are systematically arranged for the post-festival visiting phase. Sons-in-law are not unimportant but are placed in a "deferred handling" position.

For this reason, sons-in-law in traditional society did not need to engage in complex time-switching on New Year's Eve. The real "hardship" began in modern times.

So, Where Did This "New Rule" Come From?

Personally, I feel the answer lies not in traditional rituals but in reality.

In recent decades, clan structures have gradually dissolved, shrinking from "clans" to "couples and parents." New Year's Eve is no longer a mandatory festival requiring a return to one's hometown or ancestral hall. At the same time, transportation conditions have undergone profound changes. While seemingly convenient, they have brought the issue to the forefront: Same-day round trips have become technically feasible and psychologically "expected."

Thus, a highly modern compromise has solidified. Sons-in-law visit their mother-in-law's home at noon and rush back to their own home in the evening. It is fair, dignified, considerate of emotions, and not too costly—unless the geographical distance is particularly large, making the round trip genuinely difficult.

However, what is perplexing is that this arrangement has not replaced "the second day for sons-in-law" but has been superimposed on it. The result is that sons-in-law must first visit their mother-in-law's home on New Year's Eve and then formally visit again two days later for New Year greetings. The traditional "one-time arrangement" has been split into "two obligations" in modern times.

As more and more families adopt this practice, it is no longer seen as a negotiated outcome but quickly moralized into a "rule." The cost of refusing it far exceeds the cost of following it.

In summary, the expectation for sons-in-law to visit their mother-in-law's home for New Year is a "new tradition," but it is often packaged as an "old custom," making it difficult to refuse. After all, traditional customs won't turn against you, but reality might. Of course, in this process, it's not just the sons-in-law who face inconvenience; their wives and children must also join in the hustle, repeatedly embarking on this unavoidable journey.