A large number of articles about AI emerge on the internet almost every day. In my opinion, most of these articles have very little reference value aside from grabbing attention, driving traffic, and deliberately creating anxiety. So I almost just glance through these types of articles.

However, the article I’m recommending to you today, although it inevitably causes some anxiety, reflects conditions very close to what I and my surrounding friends have personally experienced in several aspects. The overall judgment and viewpoints presented are also relatively objective and forward-looking, so today I’d like to share a few points from the article "Full Token-Maxxing, an Arms Race No One Dares to Stop" (see the reference link at the end of the article) that I think are worth contemplating.

- “As I sat in the audience of YC W26 batch Demo Day, I decided to stop taking notes when the fifth company went on stage for a roadshow... I realized that the things I wrote down might be outdated by next month... Among the over a hundred companies in this batch, what they are doing is actually highly concentrated: about 80% are vertical agents, such as helping lawyers organize documents, helping customer service distribute tickets, helping HR screen resumes... These vertical agents, before establishing business barriers, a regular engineer today, even I, could produce them over a weekend; they have lost their investment value... Silicon Valley itself is starting to fall behind its own pace.”

In my previous articles, I believed that for the general public to enjoy the dividends of the AI era, there are three ways: start a business in the AI field, find a job in an AI company, and invest in the AI sector.

Regarding starting a business in the AI field, I know I am not good at it, so I write very little, but I am very pleased with those who have the courage to venture into it.

But after reading this description, I feel that entrepreneurship in AI (at least at this stage) is likely far more difficult than I originally imagined.

YC is a leading incubation organization in Silicon Valley, and even the projects they incubate are constantly facing the danger of being disrupted by large model companies, making it even harder for other projects to break through.

“Silicon Valley itself is starting to fall behind”------- When I read this sentence, I feel that if one hasn’t personally experienced such brutal torture, it’s probably hard to say something like that.

What does this indicate?

This indicates that entrepreneurship in the AI era, at least at this stage, carries exceptionally high risks, so my inclination towards the option of AI entrepreneurship has now turned conservative and cautious.

- “This sense of anxiety is spreading to the researcher group. Researchers are the pinnacle talents; it does not generically refer to ‘research personnel,’ but specifically to those responsible for model training and algorithm innovation in large model companies (like OpenAI, Anthropic, DeepMind, etc.)... And now, even the work of researchers is being automated. This is what the people at DeepMind are doing—using models to train models, which is also the hot trend of AI self-evolution in Silicon Valley this year. Those who are eliminated this year are engineers, and by the end of the year, researchers will also begin to be replaced.”

The replacement of ordinary engineers and large-scale layoffs is something I’ve heard quite a bit about already.

But researchers are academics. Generally speaking, many researchers in large companies can meet the level of university professors or associate professors. If even such talents are facing the risk of unemployment and replacement by the end of this year, it’s hard for me to imagine what other professions that were previously technology-gated can remain irreplaceable in AI companies.

Here, I mean those professions that were previously technology-gated, referring to a large number of mid-level and even senior technical personnel. If their jobs are not secure, what jobs can ordinary people hope to find in AI enterprises?

Of course, I still believe that top talents are hard to be replaced, but those positions do not belong to the general public, nor do they belong to ordinary high-level talents.

There’s also a broader assertion that AI will create more new positions while replacing human jobs; I believe this assertion, but in the short term, I see no new positions being created that can accommodate ordinary people in large numbers. Even if such positions arise in the future, it might be that for our generation, once we pass a certain age threshold, we may not have the opportunity to participate.

Now, among the three options I mentioned earlier, the only one I can barely still see clearly is investment.

- “The entire Meta is using Claude Code. This is not a startup, nor is it some experimental team, but a company worth trillions... Moreover, the token consumption rate is very likely not to grow linearly... Once the reliability of the agent improves to another level, token consumption will no longer be an issue of increasing by 50% each year, but rather increasing by an order of magnitude overnight. There is a prediction that has reached a consensus among friends: by the end of this year, many companies (including large tech giants) will actually only need 20% of their current workforce... The future scenario may be that 10 people do the work of 100, earning 20 salaries, while 90 will be unemployed.”

This passage mentions the widespread use of Claude at Meta. In fact, not only Meta, but various major companies in Silicon Valley (except for some companies like Google that explicitly prohibit the use of Claude) are almost all deeply using Claude.

According to the current situation, American large model companies have basically formed a monopoly pattern. I believe that no more than three will ultimately come out on top, with a high probability being OpenAI, Anthropic, and Google; even XAI is likely to be out of the picture now.

Among these three, I am most optimistic about Anthropic.

Regardless of which of these three ultimately emerges as the top contender, it is likely that their future market value will reach or even exceed that of NVIDIA.

Additionally, what has inspired me greatly in this passage is the usage of tokens. According to this statement, once the stability of the AI agents improves, then the consumption of tokens will experience non-linear growth, or even exponential growth?

If the usage of tokens can increase exponentially, then all industries surrounding tokens, especially those with hard constraints, will have a vast room for growth in the future, such as graphics cards, electricity, data centers...

Here, NVIDIA is worth pondering.

- “This seemingly distributed innovative world is, at its core, extremely centralized. This center is NVIDIA... I originally thought the scarcity of cards had eased over the past year. But this time I found that scarcity has returned, and is more outrageous than the last time... The power structure behind is very clear: whoever has the cards is powerful, and who has the cards is decided by NVIDIA.”

This passage emphasizes NVIDIA's monopolistic position.

For a long time, I have been observing the solidity of NVIDIA's monopoly status.

Although Google and AMD are also fiercely competing, I have a gut feeling that Google seems to lack the genetic makeup for hardware; its TPUs might find it hard to dominate the market like NVIDIA. As for AMD? Successful cases of the second-place shaking up the first-place in history seem extremely rare.

So if the development of AI and the technology route continues as it is now, NVIDIA's monopoly position might not be so easily shaken.

However, if a different AI development path emerges in the future, leading to a complete overhaul of the chip design needed for computing power, that might be when NVIDIA's position becomes vulnerable.

But no one knows if or when that situation will occur.

So NVIDIA's monopolistic advantage may be greater than imagined.

Though NVIDIA is a good stock, its price is quite awkward.

I remember not long ago when NVIDIA was about 180 dollars, there were readers commenting that it wasn’t appropriate to buy NVIDIA at that price point, and I wouldn’t buy at that price either.

But it hasn’t been long, and its stock price has surprisingly exceeded 210 dollars.

I still don’t dare to buy now.

Compared to NVIDIA, I am now more inclined to buy Anthropic.

- “Silicon Valley is undergoing a deep crisis of security. During this trip to Silicon Valley, I repeatedly heard friends seriously discussing the same issue: buying Bitcoin, building bunkers, installing bulletproof glass at home; they weren’t joking.”

This passage makes me curious about their logic for buying Bitcoin.

From this description, I guess they are considering from safety and extreme conditions, so in the physical world, they choose to insure themselves with bunkers and bulletproof glass, and in the virtual world, they choose to insure themselves with Bitcoin.

Reference link:

https://mp.weixin.qq.com/s/kwErGjX231e2efVWhERzTw

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。