The true information advantage has only one use: placing a bet before everyone else prices it in.

Over the past two years, everyone has been anxious, trying to find the answer to the same question, what will be the next big thing in AI?

Storage, optical modules, computing power stocks, energy stocks, and so on, a new narrative every few months, each time someone falls into the void, each time someone says it will definitely be next time.

Few have asked another question: what are the people who know AI the best betting on?

The group of people who left OpenAI now have a combined net worth approaching $1 trillion. And with their entrepreneurship and investments, they are leading the way into the next era of AI.

Dario Amodei founded Anthropic, potentially valued at $900 billion. Ilya Sutskever's SSI has no product, valued at $32 billion. Aravind Srinivas created Perplexity, valued at $21.2 billion. Mira Murati's Thinking Machines Lab, valued at $12 billion.

So, the most important output of OpenAI in recent years may not be GPT-4, but this group of departing employees who have been integrated into society.

Among them, the youngest OpenAI reject, Leopold Aschenbrenner, has become one of the most frequently mentioned names in the capital markets in the past two years.

A legendary record that the media has repeatedly chewed on: at 23, fired by OpenAI, he wrote a 165-page report "Situational Awareness," leveraged a hedge fund from $225 million to $5.5 billion within a year, heavily invested in nuclear power and fuel cells, hitting the mark on all.

The story is too complete, the contrast too strong, the result too successful. To this day, whenever discussing the investment logic of the AI era, he is almost an inevitable mention.

But Leopold was just the first of this group to be seen.

Those who left OpenAI have taken two paths.

One is the path of Ilya, Mira, Aravind: coming out to start a business, raising massive funding, rushing towards a disruptive product, following in the footsteps of every Silicon Valley genius exodus.

Another much quieter approach: a group of people chose to place bets, delegate execution to others, and focus solely on making judgments.

Leopold took the extreme form of the second approach.

He went to the public market, adopted an AI industry operator's perspective, identified mispriced assets in traditional energy stocks, and then made significant investments. He didn't understand energy, but he knew how much electricity AI would consume, and that was enough. This kind of insight cannot be replicated by reading reports or attending industry conferences; it can only be accumulated by being in that position.

Outside of this approach, there is another group of people following the same logic but in a different form: smaller-scale funds that complete months of due diligence in a few hours, where the exclusion list is more valuable than the investment list. They constitute the most easily overlooked and yet most worthy layer to delve into in this great exodus.

Most people leave a company with their resumes. Those coming out of OpenAI take with them a set of answers that others do not yet know 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 successful investments, he gradually reshuffled his portfolio by the end of 2025, liquidating Vistra and further concentrating funds on Bloom Energy and data center infrastructure.

Traditional energy analysts focus on these two stocks, extrapolating grid expansion plans, comparing carbon tax policies, and constructing demand growth models. Leopold's path is entirely different from this.

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

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

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

Leopold did something more precise: adopting an AI industry operator's perspective, he found a pricing lag loophole in the traditional energy public market.

This path is very difficult to replicate.

2. Zero Shot: The Most Valuable is the Veto List

Zero Shot Fund's founder Evan Morikawa, also coming from OpenAI, has a strong technical background and then moved into VC.

Despite being alumni, their paths diverged significantly.

Leopold's judgment comes from his firsthand experience in a core AI position, with insights into model training costs, data center planning, and energy requirements—accumulated only by sitting in that specific position, with no fast-forward button. The number of people truly qualified to tackle this challenge from within OpenAI's core positions is extremely limited.

In April of this year, a $100 million fund quietly emerged, named Zero Shot.

This is a term in AI training referring to a model providing an answer without seeing any examples.

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

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

$100 million might seem small compared to today's AI funds, which easily reach billions.

However, it is more telling to discuss the sectors they have chosen to avoid.

Mayne has publicly expressed skepticism towards most "ambient programming" tools, the kind of products that help you write code using natural language.

His rationale is quite straightforward—he knows what OpenAI has accumulated internally in the programming space and is aware of how quickly these tools' moats will be eroded by base models. Morikawa, on the other hand, is keeping his distance from numerous "human-centric video data companies" in the robotics field. These are companies that specialize in collecting human motion data to train robots. In his view, this technological path will hit a dead end.

These two judgments are not something your average VC can provide.

They have not been at the source of that information, have not been part of those internal discussions, so they cannot determine which path is a dead end.

The advantage of Zero Shot lies in its blacklist. In a market where everyone is shouting about AI startups, knowing where the pitfalls are is more valuable than knowing who to bet on. For those who have already mined, having a minefield report is more useful than having a treasure map.

They deliberately kept their scale at $100 million, for a very specific reason.

They are aware of when their advantage is most valuable: in the early stage where the technological roadmap has not yet converged. At that stage, those in the know can easily distinguish which path is viable.

As projects move on to Series C and D rounds, financial data and public information will overshadow the informational advantage, and that card will be played out.

The larger the scale, the more they need to pursue a "high certainty runway," essentially fighting battles using others' tactics.

$100 million is their honest assessment of the boundaries of their advantage.

3. When Angel Investing Is Another Business

Mira Murati and the Zero Shot Fund both invested in Worktrace, a company that optimizes enterprise workflows using AI, founded by former OpenAI colleague Angela Jiang.

But the investment logic is much more solid than just having a good relationship.

Mira has seen how Angela made decisions in a high-pressure environment at OpenAI, witnessed her judgment of AI product boundaries, and observed her execution under real constraints. These things cannot be faked in a two-hour founder pitch or unearthed through even the most thorough due diligence.

Angela doesn't need to convince Mira to believe in her because Mira had already formed her judgment. The information cost of angel investing approaches zero, but the quality of information far exceeds the market average.

A larger flywheel is in Sam Altman's domain.

Reportedly, Altman decides whether to follow-on investment within hours of hearing about a former employee starting a company, leveraging the OpenAI Startup Fund's capital and extensive API resources.

Although he doesn't hold any OpenAI equity himself, each successful alum expands OpenAI's data funnel, distribution channels, and policy influence. He is using capital to sustain an ecosystem that doesn't belong to him but continues to yield returns. It's an invisible equity that significantly compounds.

Many mistakenly see this ecosystem as a cozy arrangement among former colleagues.

Comparing it with the PayPal Mafia makes the difference very clear.

The cohesion of the PayPal Mafia stems from shared adversity: they fought together in the payment wars, went through the eBay acquisition together, and formed trench camaraderie during those almost-dead years. This trust is genuine, but their visions of the future are individual. Thiel does VC, Musk builds rockets, Hoffman networks, and their paths diverge.

OpenAI alumni are brought together by a common bet on the future: AGI will come, the window of opportunity is limited, and the present is a rare moment for strategic positioning. The driving force of belief is more enduring than camaraderie because it directly aligns with self-interest. Once everyone's bet is in the right direction, the entire network will benefit.

This also makes the barrier to entry into this circle very subtle.

If the product is good enough, raising money from this group of people is not a problem. But if you are skeptical about AI's future, or if your entrepreneurial logic is based on the premise that "AGI is still far away," even if the product is excellent, it is very difficult to get a check from this group of people.

A difference in worldview will end the conversation before the handshake.

IV. From Builders to Investors

The paths taken by OpenAI alumni can be grouped into three categories.

Ilya, Aravind, and Mira have all chosen entrepreneurship.

However, even though they are all entrepreneurs, they are doing completely different things. Aravind is in a fiercely competitive consumer business, Mira is building a tool platform, and Ilya's SSI doesn't even have a product, yet it has achieved a $32 billion valuation, betting on the word "security" itself.

Leopold and Zero Shot have chosen to become investors.

Leopold has entered the public market, while Zero Shot is in early-stage VC. Both have externalized their judgments into capital instead of executing them personally. This is a minority within OpenAI alumni, but this minority is worth taking a closer look at: when someone is willing to bet without doing it themselves, it usually means their judgment of the outcome is so clear that they don't need to explore it through action.

People usually think that the highest expression of genius is creation. But this group of people provides another answer: when judgment is clear enough, dispersing cognitive efforts across multiple directions and letting those with execution power build is a more efficient choice.

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

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

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

The smartest people of our time are starting to choose ALL IN, indicating that the answer seems clear enough to them, clear enough that they don’t need to rely on action to verify it further.

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia