Former OpenAI Staff Quietly Launch Zero Shot, a Venture Fund Targeting $100M

Former OpenAI employees have launched a new venture fund, Zero Shot, with a target size of $100M. The fund has already started investing quietly, with no public announcements and no details about its investments or its management team.

Zero Shot is not just a catchy name. In machine learning, "zero-shot" refers to a model's ability to solve tasks without specific training on examples of them. It is a foundational concept built into most modern language models, including the GPT series. The choice of name is a deliberate signal: the founders understand the technology from the inside and want that to be obvious right away to potential LPs and to the startup founders they are evaluating.

The trend is not new, but it becomes more visible every year. OpenAI has become a true breeding ground for venture investors and a new generation of founders. Among those who left the company to build their own ventures are the Anthropic team led by Dario Amodei, Ilya Sutskever with Safe Superintelligence, and dozens of other researchers and product managers who have scattered across the industry. Some created rival labs, some moved into the enterprise sector, and now more and more people are choosing the venture path, to shape the next wave rather than merely take part in it.

People who worked at OpenAI in its early stages hold a rare advantage. They saw from the inside how AI products that later transform entire industries are built. They can tell real technological progress from hype, and they know exactly what AI startup founders need, because they were in that role themselves.

Zero Shot has operated in stealth mode: no press releases, no conferences, no media activity. According to TechCrunch, the fund has already written several investment checks, although the names of portfolio companies have not been disclosed. A $100M target is typical for a debut early-stage fund. That amount allows 20-40 bets in the $1-5M range at seed and pre-seed rounds, enough to build a diversified portfolio without chasing mega-rounds.

The fund's secrecy is most likely a deliberate strategy. In venture capital, reputation is built slowly, and early bets are better made without public pressure. Stealth mode buys time to refine internal processes, build a steady deal flow and show first results to a limited group of partners, before going to market with a large-scale fundraise or public comments.

For the AI ecosystem as a whole, the emergence of specialized funds with deep technical expertise is a good sign. Generative AI is now going through a difficult transition from experimentation to commercial deployment. Startups need not just money but investors who can help with product decisions, technical architecture and landing their first enterprise customers. That is exactly what alumni of the world's leading AI lab can offer.

If Zero Shot successfully closes its first $100M fund, it will join the ranks of specialized AI funds such as AI Grant or Conviction Capital, but with the most direct operational link to OpenAI. The first portfolio companies will show the team's real priorities: vertical AI agents, model infrastructure, multimodal applications or something fundamentally different.

For now, the managers' names have not been made public and the portfolio has not been disclosed. But the very existence of Zero Shot says one thing: the people who built the AI revolution from the inside now want a financial stake in its next wave, and this time from the other side of the table.