Leaving OpenAI, how many times has their net worth multiplied?

There is only one way to use a true information advantage: bet before anyone else can price it in.

For the past two years, everyone has been anxious, trying to find the answer to the same question: What is the next sector for AI to rally?

Storage, optical modules, AI computing stocks, energy stocks, etc., the narrative changes every few months. Someone always misses the boat, and someone always says "next time for sure."

Few people ask the other question: What are the people who understand AI the most betting on?

The group that left OpenAI now has a combined net worth approaching $1 trillion. Their ventures and investments are walking at the dawn of the next era of AI.

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

So, the most important output of OpenAI in recent years might not be GPT-4, but the alumni it has exported to society.

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

The legendary track record has been chewed over by the media repeatedly: Fired from OpenAI at 23, he wrote a 165-page report titled "Situational Awareness," grew a hedge fund from $225 million to $5.5 billion within a year, and went all-in on nuclear power and fuel cells, betting correctly on every single one.

The story is too complete, the contrast too strong, the result too successful. Now, whenever the investment logic of the AI era is discussed, he is almost unavoidable.

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

Those who left OpenAI have gradually forged two paths.

One path is the one taken by Ilya, Mira, and Aravind: start a company, raise massive funds, pursue the next disruptive product – a classic repeat of Silicon Valley geniuses leaving to found startups.

The other path is much quieter: a group of people chose to place bets, leaving the execution to others while specializing in making judgment calls.

Leopold's path is the extreme form of the second route.

He went into the public markets, using the perspective of an AI industry operator to find mispriced assets in traditional energy stocks, and then bought heavily. He doesn't understand energy, but he knows how much electricity AI will consume, and that's enough. This kind of insight cannot be replicated by reading reports or attending industry conferences; it can only be accumulated by having held that position.

Beyond this path, there is another group of people doing similar things with different structures: smaller funds closing due diligence in hours that takes others months, where the list of rejections is more valuable than the list of investments. They constitute the most easily overlooked yet most worthy layer of this great exodus.

Most people, when leaving a company, take away a resume. What those leaving OpenAI take away is a set of answers that others don't yet know they need.

1. There Is No Second Leopold

Leopold's heavy bets were on nuclear power company Vistra and fuel cell company Bloom Energy.

After both bets paid off, he gradually adjusted his positions around the end of 2025, clearing out Vistra and further concentrating capital in Bloom Energy and data center infrastructure.

Traditional energy analysts looking at these two stocks would pull up grid expansion plans, compare carbon tax policies, and build demand growth models. Leopold's approach was entirely different.

He saw the scale of server rooms at OpenAI, witnessed the electricity bills for training a flagship model, and heard engineers discussing why the next generation of data centers must be located next to nuclear plants. These details aren't in any financial report or analyst report, but they formed a conclusion about energy demand more real than any model.

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

In the past, this was the domain of top-tier macro hedge funds, relying on a global perspective of the macro economy.

Leopold did something more precise: using the operator's perspective from the AI industry, he found a loophole in the lagging pricing within the traditional energy public markets.

This path is very difficult to replicate.

2. Zero Shot: The Most Valuable Thing Is the Rejection List

Evan Morikawa, founder of the Zero Shot fund, also came from OpenAI with a solid technical background, but he chose to go into VC.

Same alumni network, completely different path.

Leopold's judgment stems from his hands-on experience in a core AI role – first-hand perception of model training costs, data center planning, and energy demand. It can only be accumulated by being in that seat; there's no fast-forward button. Very few people in OpenAI's core roles are truly qualified to make these calls.

In April this year, a new fund with a $100 million size was quietly exposed, named Zero Shot.

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

The three co-founders come from OpenAI: Evan Morikawa, former head of application engineering for DALL-E and ChatGPT; Andrew Mayne, OpenAI's first prompt engineer; and Shawn Jain, former researcher and engineer.

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

$100 million is a small figure among today's AI funds, which often run into the tens of billions.

But explaining which sectors they refuse to invest in is more illuminating.

Mayne has publicly stated he is bearish on most "vibe coding" tools – products that help you write code using natural language.

The reason is relatively direct: he knows what OpenAI is developing internally in the coding space and how quickly the moat of such tools will be directly dissolved by foundation models. Morikawa, on the other hand, keeps his distance from the large number of "human-centric video data companies" in the robotics sector – enterprises that specifically collect human action data to train robots. In his view, this technical path will hit a dead end.

These two judgments are ones that ordinary VCs cannot make.

They haven't sat at the source of information, haven't seen those internal discussions, and therefore cannot judge which path is a dead end.

Zero Shot's advantage lies hidden in its rejection list. In a market where everyone is shouting about AI startups, knowing where the traps are is more valuable than knowing who to bet on. Someone who has already mined the field finds a landmine report more useful than a treasure map.

They deliberately kept the fund size to $100 million, and the reason is quite specific.

They understand clearly at which stage their advantage is most valuable: the early stage where the technical roadmap hasn't converged. At that stage, someone with insider knowledge can instantly distinguish which paths are viable.

Once a project reaches Series C or D, financial data and public information will cover their information advantage, and this card is played out.

The larger the scale, the more one needs to chase "certain high-certainty tracks," and the more one fights using others' playbooks.

$100 million is an honest assessment of the boundary of their own advantage.

3. Angel Investing Is a Different Business

Mira Murati and the Zero Shot fund both invested in Worktrace, founded by former OpenAI colleague Angela Jiang, a company that uses AI to optimize enterprise workflows.

But the investment logic is far more robust than "good relationships."

Mira has seen Angela's decision-making process under OpenAI's high-pressure environment, witnessed her judgment regarding the boundaries of AI products, and observed her execution capabilities under real-world constraints. These things cannot be faked in a two-hour founder pitch and cannot be reconstructed through the most meticulous due diligence.

Angela didn't need to convince Mira to believe in her, because Mira had already formed a judgment. The information cost for this angel investment approaches zero, yet the information quality far exceeds the market average.

An even larger flywheel exists with Sam Altman.

Reports indicate that Altman decides within hours of hearing about a former employee starting a company whether to co-invest, often layering on capital from the OpenAI Startup Fund and substantial API credits.

He himself holds no equity in OpenAI, but every success of an alumnus 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 continues to yield returns for him. It's an invisible form of equity, yet it compounds in real terms.

This ecosystem leads many to mistakenly believe it’s just about old colleagues sticking together.

The difference becomes clear when comparing it to the PayPal Mafia.

The PayPal Mafia's cohesion came from shared hardship: fighting the payments war together, going through the eBay acquisition together, forming a trench-bound camaraderie in those years when they nearly died. That trust is genuine, but their judgments about the future were divergent. Thiel went into venture capital, Musk into rockets, Hoffman into social networks – their paths scattered.

What binds the OpenAI alumni together is a shared bet on the future: AGI is coming, the window of opportunity is limited, and the present is a once-in-a-lifetime moment for deployment. The driving force of faith is more enduring than camaraderie, because it directly connects to interests; once everyone's bet direction is correct, the entire network benefits.

This also makes the entry barrier to this circle quite subtle.

If your product is good enough, getting funding from this group isn't a problem. But if you are skeptical about the future of AI, or if your entrepreneurial logic is based on the premise that "AGI is still far away," even with an excellent product, getting a check from this group will be very difficult.

A disagreement on worldview ends the conversation before the handshake.

4. From Builders to Investors

The destinations of OpenAI alumni can be summarized into three categories.

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

But even within entrepreneurship, they are doing completely different things. Aravind is in a highly competitive consumer business, Mira is building a tool platform, and Ilya's SSI has no product yet but secured a $32 billion valuation, betting on the concept of "safety" itself.

Leopold and Zero Shot chose investing.

Leopold went into the public markets, Zero Shot does early-stage VC. Both are externalizing judgment as capital rather than executing personally. This is a minority among OpenAI alumni, but this minority deserves a closer look: a person willing to bet rather than build usually means their judgment about the outcome is already so clear that it doesn't require action to explore.

People often think the highest form of genius is creation. But this group offers another answer: when judgment is clear enough, distributing cognition across multiple directions and letting people with execution skills do the building is a more efficient choice.

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

A pilot's situational awareness determines their actions two seconds later; losing it means death. What this group took away from OpenAI is precisely this situational awareness of the AI battlefield. They know the direction of the battle, know where the high ground is, and know which trenches lead to dead ends.

What they are doing now is deploying their forces accordingly.

When the smartest people of an era start choosing to go ALL IN, it means the answer is already clear enough in their eyes – clear enough that it no longer requires hands-on tinkering to verify.