The only true information advantage has one usage: to place a bet before others set the price.

In the past two years, everyone has been anxious, trying to find the answer to the same question: what will AI's next sector rise?

Storage, optical modules, computing stocks, energy stocks, and so on, each narrative changes every few months, each time someone misses out, every time someone says they will definitely get it next time.

Very few people ask another question: what are those who know AI best betting on?

The group of people who left OpenAI has collectively reached a valuation nearing 1 trillion USD. Their entrepreneurship and investment is at the start of the next era of AI.

Dario Amodei founded Anthropic, with a potential valuation of 900 billion. Ilya Sutskever’s SSI has no product, valued at 32 billion. Aravind Srinivas made Perplexity, valued at 21.2 billion. Mira Murati’s Thinking Machines Lab, valued at 12 billion.

Thus, the most important output from OpenAI over the past few years may not be GPT-4, but this group of employees who left and reentered society.

Among them, the youngest, Leopold Aschenbrenner, who was fired by OpenAI, has become one of the most frequently cited names in the capital market in the past two years.

The legendary record has been repeatedly chewed over by the media: fired from OpenAI at 23, he wrote a 165-page report titled "Situational Awareness," leveraging a hedge fund from 225 million to 5.5 billion within a year, heavily investing in nuclear power and fuel cells, hitting everything right.

The story is too complete, the contrast too strong, and the result too successful. By now, whenever AI investment logic is discussed, he is nearly inescapable.

But Leopold is just the first among this group to be noticed.

The people who left OpenAI have gradually walked two paths.

One path is that of Ilya, Mira, and Aravind: launching startups, raising huge funds, rushing for the next disruptive product, echoing every time Silicon Valley geniuses have exited.

The other path is much quieter: a group of people chose to place bets, leaving execution to others while only making judgments themselves.

Leopold took the extreme form of the second path.

He went to the public market, using the operator's perspective of the AI industry, found mispriced assets in traditional energy stocks, and then heavily bought in. He doesn’t understand energy, but he knows how much electricity AI will consume, and that is enough. This kind of understanding cannot be replicated simply by reading reports or attending industry meetings; it can only be accumulated by having been in that position.

Outside of this path, there is another group of people doing the same logic but in different forms: smaller funds, completing due diligence in hours that take others months, and having a list of rejected prospects that is more valuable than the investment list. They form the layer that is easiest to overlook in this great exodus and is also the most worthy of in-depth study.

Most people leave a company with a resume. Those who come out of OpenAI carry a set of answers that others do not yet realize they need.

1. There is No Second Leopold

Leopold heavily invested in the nuclear power company Vistra and the fuel cell company Bloom Energy.

After betting on both, he gradually adjusted his portfolio by the end of 2025, clearing out Vistra and further concentrating funds on Bloom Energy and data center infrastructure.

Traditional energy analysts watch these two stocks, monitoring grid expansion plans, comparing carbon tax policies, and modeling demand growth. Leopold's path is completely different.

He has seen the scale of server rooms at OpenAI, seen the electricity bills for training a flagship model, and heard engineers discuss why the next generation of data centers must be sited next to nuclear power plants. These details are not in any financial reports or analyst reports, but they form a conclusion about energy demand that is more real than any model.

This approach is called "cross-industry cognitive arbitrage" in investment circles: translating internal information from one industry into undervalued assets in another industry.

In the past, this was the patent of top macro hedge funds, relying on global macroeconomic perspectives.

Leopold did something more precise: he found pricing lag issues in the traditional energy public market using the perspective of an AI industry operator.

This path is hard to replicate.

2. Zero Shot: The Most Valuable is That Rejection List

Evan Morikawa, founder of Zero Shot fund, also came from OpenAI, with a solid technical background, and he went into venture capital.

Also an alumnus, their paths are completely different.

Leopold's judgment comes from his specific experience in the most core positions of AI, which is first-hand perception of model training costs, data center planning, and energy demand; one can only accumulate this from being in that position—there’s no fast-forward button. Very few people have the qualifications to answer this question from core OpenAI positions.

In April of this year, a new fund with a scale of 100 million USD quietly surfaced, named Zero Shot.

This is a term from AI training, referring to a model answering directly without having seen any samples.

The three co-founders are from OpenAI: former DALL-E and ChatGPT application engineering lead Evan Morikawa, OpenAI’s original prompt engineer Andrew Mayne, and former researcher and engineer Shawn Jain.

They have already invested in three companies: AI enterprise workflow company Worktrace, AI-enhanced factory robotics company Foundry Robotics, and another project that is still operating in stealth mode.

At 100 million USD, in today’s AI funds often reaching tens of billions, it's a small figure.

But discussing which sectors they refuse to invest in better illustrates the problem.

Mayne openly stated he is bearish on most "ambient programming" tools, a category that helps you write code using natural language.

The reason is quite direct; he knows what OpenAI has internally accumulated in programming directions, and he knows how quickly the moat of such tools will be directly dissolved by foundational models. Morikawa distances himself from numerous "human-centered video data companies" in the robotics sector, thinking these companies that specifically collect human motion data to train robots will hit a wall.

These two judgments are something ordinary VCs cannot provide.

They haven't stayed at the source of information, haven’t seen those internal discussions, so they lack the means to judge which paths are dead ends.

Zero Shot's advantage lies in its rejection list. In a market where everyone is shouting about AI entrepreneurship, knowing where the pitfalls are is worth more than knowing who to bet on. Someone who has already mined will find a squat report more useful than a treasure map.

They deliberately kept the scale at 100 million USD for a specific reason.

They are clear about at which stage their advantage is most valuable: the early stage where technological routes have not yet converged. At that stage, insiders can distinguish at a glance which paths can succeed.

Once projects reach Series C or D, financial data and public information will cover up the information advantage, and that card will be done.

The larger the scale, the more one needs to chase "certainty's big track," and the more one fights using others' playbooks.

100 million is an honest judgment of their own advantage boundaries.

3. When Angel Investing is Another Business

Mira Murati and the Zero Shot fund both invested in former OpenAI colleague Angela Jiang's Worktrace, a company optimizing corporate workflows with AI.

But the investment logic is much more solid than just "good relations."

Mira has seen Angela's decision-making style in the high-pressure environment of OpenAI, has seen her judgments on the boundaries of AI products, and has seen her execution under real constraints. These things cannot be simulated in a two-hour founder pitch, and no matter how detailed the due diligence is, they cannot be restored.

Angela does not need to persuade Mira to believe in her because Mira has already formed a judgment. The information cost of angel investment approaches zero, but the quality of information far exceeds the market average.

A bigger flywheel is at Sam Altman.

Reportedly, Altman decides whether to co-invest within hours of hearing about former employees starting companies, adding the OpenAI Startup Fund's funds and abundant API resources.

He personally does not hold equity in OpenAI, but every alumnus’s success expands OpenAI’s data entry points, distribution channels, and policy influence. He is using capital to maintain an ecosystem that doesn’t belong to him but continuously rewards him. This is an invisible equity but genuinely compound.

This ecosystem has led many to mistakenly believe it is merely a warm gathering of old colleagues.

When compared with the PayPal mafia, the differences become very clear.

The cohesiveness of the PayPal mafia comes from shared suffering: enduring payment wars together, experiencing the eBay acquisition together, and forming trench friendships in those life-threatening years. This kind of trust is real, but their judgments about the future are individual. Thiel does venture capital, Musk builds rockets, Hoffman develops social networks; the paths diverge.

The OpenAI alumni’s unity is based on a shared bet on the future: AGI will come, the window of opportunity is limited, and now is a once-in-a-lifetime opportunity to position. The driving force of belief is more lasting than friendship because it directly connects to interests; once everyone's betting direction aligns, the entire network benefits.

This also makes the entry threshold of this circle very subtle.

If the product is good enough, securing these people’s money is not an issue. But if you hold a skeptical attitude about the future of AI, or if your entrepreneurial logic is rooted in a premise that "AGI is still far away," even the best products may find it hard to get these people's checks.

Differences in worldview may end conversations before the handshake.

4. From Builders to Investors

The paths of OpenAI alumni can be roughly categorized into three types.

Ilya, Aravind, and Mira all chose to start companies.

But even among entrepreneurs, they are doing completely different things. Aravind is in a fiercely competitive consumer business, Mira is in a tools platform, and Ilya's SSI has no product but has a valuation of 32 billion, betting on the very word "safety."

Leopold and Zero Shot chose investment.

Leopold went to the public market, and Zero Shot does early-stage VC; both paths externalize judgment into capital rather than executing personally. This is rare among OpenAI alumni, but this minority is worth a separate look: a person willing to place bets without doing it themselves usually means their judgment on outcomes is clear enough that they do not need to explore through action.

People usually believe that the highest expression of genius is creation. But this group of people offers another answer: when judgment is clear enough, diversifying understanding into multiple directions and letting capable people build is a more efficient choice.

Leopold's report is titled "Situational Awareness," a military term referring to a pilot's real-time perception ability of the battlefield as a whole.

A pilot's situational awareness dictates his actions two seconds later; losing it means death. What this group brought out of OpenAI is precisely this situational awareness of the AI battlefield. They know the direction of the battle, know where the high ground is, and understand which trench leads to a dead end.

What they are doing now is deploying based on this understanding.

The smartest people of the era starting to choose ALL IN indicates that the answer seems clear enough to them, clear enough not to need to rely on hands-on verification anymore.

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。