As an introvert, I am accustomed to hiding behind the pseudonym 'Tibetan Fox' to analyze technology and industry trends. I prefer not to reveal too much about myself and consider my real life and personality quite ordinary—not worth mentioning.

However, in an era where AI can effortlessly generate countless words, I have grown increasingly fond of the emotions and neural impulses of carbon-based beings linked behind the text. No matter how ordinary, every life carries its unique weight. Thus, I have long wanted to find an opportunity to chat with readers and share some of my personal industry observations from this year.

Why summarize in October? The answer is, during an editorial meeting, our casual banter led us here...

At the routine weekly editorial meeting on September 28, we each shared recent work updates—or, in industry jargon, 'aligned our efforts.'

Someone mentioned the prevailing AI doomsday theories in Silicon Valley, saying Americans love to worry unnecessarily. Another recalled the hype around OpenClaw, noting that the installation files still lingered on their computer, and how the collective anxiety over 'raising virtual lobsters' seemed unnecessary in retrospect.

When it was my turn, I said, 'I’ve reached a point where I understand both sides.'

'Overall, my feelings are complex. I’m not as pessimistic about AI doomsday or bubble theories, but I’m not entirely anxiety-free either. I can clearly sense why such anxieties are prevalent among programmers or the general public. If AI is a lever, then it’s one everyone can access. AI is spreading and being adopted at an incredibly fast pace, but each person’s ability to leverage that lever varies.'

These were my internal reflections shared with the team. Indeed, two valid forms of anxiety exist. Sigh, I’ve become a contrarian again.

But why do I know this? Because this AI-related anxiety led to my most viral piece this year.

'Distillation,' published in April 2026, is the most explosive article I’ve written in recent years—by far.

It garnered nearly a million views across platforms, was reposted by multiple accounts, and even led other authors to add me on WeChat, asking how I conceived such an angle.

At the time, anxiety over lobsters, Agents, and skills was intense. I believed these things weren’t as frightening as they seemed. In the article, I argued that abilities that appeared magical now might, with advancements in large models, achieve similar effects through direct conversation without needing extra skill calls.

Facts proved me right: OpenClaw, Agents, and skills quickly became accessible, and now everyone can use them in AI applications without any command-line knowledge.

Back then, 'Distillation' went viral, and I felt proud for a day or two. But after the excitement faded, another part of me couldn’t help but ask: When everyone was anxious about skill distillation, were you truly not anxious? Were you just trying to prove your 'sober' perspective by differing from others?

Honestly, my rebuttals to the lobster craze and distillation fears weren’t deliberate.

First, seeking and exposing weaknesses in arguments, without halting inquiry just because a viewpoint becomes mainstream, is both my habit and a job requirement. After all, if I don’t provide incremental information, my writing loses value.

Moreover, 'Distillation' resonated not because I nailed the technical aspects, but because it addressed a collective panic over new technologies like Agents and skills. Feelings of 'falling behind' or 'exhaustion from learning' were widespread. Being closer to the AI industry, I saw things more clearly and could respond to these emotions sooner—though the views themselves weren’t groundbreaking.

This viral hit also made my colleagues and me realize we should engage more closely with the public. Over the next few months, my work underwent unprecedented changes: high-intensity field visits, trekking into remote mountainous areas, and discussing AI with ordinary people.

Getting closer to everyone had a practical reason too: survival.

With Agents, writing styles can easily be distilled. What advantage do we manual writers have? Perhaps our eyes, feet, mouths, and ears—the ability to venture where AI cannot, see people AI misses, and listen and converse. These have become some of our few remaining core competencies as human creators.

This is how AI changed me in 2026. Facing the latest tech trends, I’m not anxious, but as a creator, I must adapt.

My high-intensity travels revealed a more fragmented AI world.

Over the past ten months, we’ve engaged intensively offline and launched the 'Spark Plan,' soliciting paid AI stories from ordinary people. Every month, I traveled, meeting deaf developers, AI learners in Guizhou’s mountains, a sci-fi AI film director, AI short drama creators, those selling AI services on Xianyu, and individuals marketing AI courses to parents in county towns.

My prediction of AI-driven technological empowerment has been validated countless times in reality.

Earlier this year, I interviewed Wang Jingjing, a senior AI developer and hardware engineer who used OpenClaw to control a robotic arm—a feat then achievable only by a few experts. By September, I saw integrated AI development boards at the HIC Conference, purchasable online by beginners to build their own AI hardware.

The same goes for Agents. What once required complex setups and paid installations is now as simple as office software, thanks to AI application companies.

Later, I visited Leishan, Guizhou, a former national-level poverty-stricken county. Without searching, I found numerous AI users. AI had become a daily tool for mountain residents; a shop selling honey and wild mushrooms displayed many AI guidebooks.

The once sci-fi future is now rapidly spreading and democratizing. AI’s social foundation is far broader than imagined. So, while debates rage over AI industry bubbles or overhyped computing power, I still believe current infrastructure and even computing power remain insufficient.

Then, another voice in me asked: Does this mean you’re optimistic about AI this year?

I’d say no.

I’m optimistic because I see more, but also anxious for the same reason.

Are the people I see the same?

They differ from the general public—all use AI because someone helps them. It could be an AI club, a company’s public welfare activity (public welfare activity), or a foundation’s free courses...

They also differ among themselves, using not the same AI or in the same way.

I met three women with zero AI background who first encountered it in 2026:

One was Xu Mengjiao, a hearing-impaired girl. Untrained in programming, she described her street dance needs to AI, which generated programming prompts. She and her peers created a street dance assistance app demo for the deaf, participating in a tech company’s hackathon. Since then, she’s led her deaf dance troupe in competitions and become a star developer at AI events.

Another is a real-life friend, also without programming knowledge. She often discussed AI with me, realizing no existing programs met her needs for watching dramas or practicing dance. When AI coding became accessible, she immediately bought a top-tier model package, supervised AI in writing code daily, quickly launched a product, registered a company, and planned to apply for Beijing’s OPC policies—all within months.

The third is Hou Changju from Guizhou’s mountains. She uses AI, asks Doubao questions, and generates Miao embroidery designs. As a Miao woman, her Mandarin wasn’t fluent, making learning initially difficult.

If you fear being left behind by AI, take heart—it’s not hard. These zero-basis women were all integrated into the tech world.

But AI’s reach is uneven.

The gap between people hasn’t disappeared with AI; it may have widened. If AI is a lever, someone originally at '1' might become '10,' but someone at '10' could reach '100.' The gap expands from '9' to '90.'""Last year, the programmer community discussed K-shaped differentiation—some rising rapidly, others declining. This year, alongside AI doomsday theories, we see I-shaped differentiation: a small group of AI experts rises above the masses. This might reflect the sentiment among programmers closest to AI’s frontier—the 'water level' of AI’s core group.

In 'The Model Crisis,' I wrote about a model factory struggling not with customer leads or enterprise demand, but budgets. AI investments concentrate in tech companies and industry leaders, becoming the biggest bottleneck for AI commercialization—I-shaped differentiation among AI firms.

Meanwhile, Beijing’s white-collar workers use premium packages to turn their favorite dramas into products, with Claude becoming the latest urban fashion accessory and status symbol. Yet, teachers in mountainous primary schools worry about exhausting their AI class tokens. If schools can’t meet students’ AI demands, will they secretly use their parents’ phones to pay for generation?

AI’s widespread adoption is inevitable, but the future will be uneven. Will it lead to I-shaped societal differentiation? Optimism or pessimism depends on who you are and whom you see.

Reflecting on these ten months, the anxious me is more pessimistic than ordinary people embracing AI; the me walking through fields is more optimistic than foreigners believing in AI doomsday.

This self-anxiety and self-rebuttal, technological speculation and real-world verification, constantly swirl in my mind without resolution.

I can only keep questioning what others hesitate to ask, keep traveling to reach distant places, and bring back authentic firsthand information and industry clues. Or, at least, keep arguing—not pretending AI creates no problems.