Phenomenon: Entrepreneurs who made billion-dollar exits are returning to work as employees
After billion-dollar exits, prominent tech entrepreneurs are stepping down from CEO roles to join AI giants like Anthropic and OpenAI as individual technical contributors.
After a billion-dollar exit, many entrepreneurs choose to rest on the beach or start their next startup. But recently, something strange is happening in the industry: the biggest names in high-tech, who have already proven their entrepreneurial skills and sold companies, are giving up the CEO chair and moving to work as regular employees at AI giants. Why are the founders of Instagram, Monzo, and even the fathers of LLM suddenly preferring a simple title of "Technical Staff Member"?
Sold one of the most powerful platforms in the world, and is now developing Claude
One of the most prominent names who took a step back for a seemingly "simple" role is Mike Krieger. The American entrepreneur managed to co-found Instagram with Kevin Systrom and sell it to Meta for over a billion dollars. After the two left Meta, they developed Artifact — a platform that tried to offer a new way to consume news. This experiment did not succeed, and Krieger found himself a new and interesting role: "Member of the Technical Staff".
The pathos-free role at a small company you might know by the name Anthropic allowed Krieger to simply build — without the commitment to targets before investors, without funding rounds, and without racking his brain with all the classic headaches of entrepreneurs. Today, Krieger leads Labs, the division of Anthropic from which developments like Claude Cowork emerged.
The new joiner
But Krieger is not alone. Just this week, one of the hottest names in the British ecosystem joined Anthropic — Tom Blomfield. The British entrepreneur founded the startup GoCardless and after it the successful digital bank Monzo. Since then, he managed to take a step back from the world of entrepreneurship, and for the last 4 years served as a mentor for entrepreneurs at the world's largest accelerator — Y Combinator. Recently, Blomfield announced that he would leave the role to join the Compute team at Anthropic, which manages the most critical resource the company needs to train models and run sub-agents in Claude Code.
The nomad and the godfather
Joining these two is Andrej Karpathy — the father of Vibe Coding and one of the 11 original founders of OpenAI. Karpathy already managed to work at Tesla (to which Elon Musk tempted him to move), return to OpenAI, and then go independent with a company called EurekaLabs that wanted to inject AI into the world of education. Despite this, last May he found a new role together with Krieger (and soon Blomfield) in the Pre-Training team at Anthropic. He explained the move by saying that "the coming years at the forefront of language model development are expected to be particularly significant and defining for the entire field" — and he wants to be there, with Anthropic.
Another interesting recruitment by Dario Amodei for the company is Peter Bailis, the entrepreneur who founded and sold a company to Snowflake for about half a billion dollars, and served as CTO of the American HR giant Workday. Last March, he left the C-Suite to become a "Member of the Technical Staff" at Anthropic, where he takes part in reinforcement learning or RL processes for Claude.
This trend has not skipped the big competitor, OpenAI. An intriguing reinforcement for Sam Altman's company is Jason Boehmig, who founded a startup in the legal-tech field called Ironclad — valued at over 3 billion dollars. Last month, Boehmig left the company to lead the connection between AI and the legal field at OpenAI. If relying on OpenAI's data, it seems that the efforts of Boehmig and the company's researchers have already borne fruit: a few weeks ago, the GPT-5.6 Sol model was launched, which officially surpassed Claude in benchmarks for tasks in the legal field.
The researcher, the entrepreneur, and the employee
Another intriguing entrepreneur who moved to be an "employee" is Noam Shazeer, who spent most of his professional life at Google and was part of the team that signed the historic paper Attention is all you need. When he left Google, he founded Character.ai and recorded a billion-dollar exit that brought him back to Google. After returning, Shazeer led the development of Gemini at Google, but last month he dropped a bombshell and announced a move to OpenAI. Sam Altman has already managed to welcome him, but Shazeer's job definition has not yet been officially revealed. However, a visit to his LinkedIn reveals that he defines himself as a "human" at OpenAI, while under his name appears the simple title "LLM Guy" — another wink illustrating the transition from entrepreneurship to a role that is seemingly below his level.
Will more come?
This influx of executives and entrepreneurs into focused roles at artificial intelligence companies — which are currently the global technological frontier — is a phenomenon that is likely only to intensify. The big players are fighting for leadership that is sometimes measured in small gaps in benchmarks, and the proven capabilities that the entrepreneurs developed in the field constitute a huge asset for these companies — even when they are not sitting in the senior management echelon.
