Meta Loses, Thinking Machines Wins: How the Talent Migration Signaled a Power Shift 

In the AI market, major shifts are not always measured by official announcements or the launch of new models, but sometimes by a less obvious movement: who is moving where.

This is precisely what the recent movements between Meta and Thinking Machines Lab reveal, where the matter is no longer just about hiring, but an indicator of a deeper change in the power centers within the sector.

The story, as tracked by a TechCrunch report, does not begin with Thinking Machines, but with Meta itself, which during the past period was one of the most prominent companies attracting talent from competitors, in an attempt to bolster its capabilities in the field of artificial intelligence. However, this trend did not continue as it was.

Over time, the picture began to reverse gradually. Instead of Meta being the party attracting researchers, Thinking Machines Lab emerged as an entity attracting an increasing number of Meta employees, to the extent that the startup now hires from Meta more than from any other source.

This shift does not appear in a single announcement, but in tracking new appointments within Thinking Machines, where the list includes prominent names who worked for years at Meta and recently moved to the startup, which the report considered a direct signal that the scales are no longer tipping entirely in favor of large companies.

At the heart of this story stands a newly established company. Thinking Machines Lab, launched in 2025 under the leadership of Mira Murati, did not take long to become a point of attraction in the market. Within a short period, it managed to build a team comprising researchers coming from companies like Meta and OpenAI, which made it appear as a real player despite its newness.

However, what is striking in the report is not just the movement of individuals, but the pattern. The flow is no longer unidirectional; it has become a continuous exchange of talent, reflecting an open market environment where competencies move quickly between large companies and startups.

This change, as the report suggests, is not only related to salaries or incentives, but to the nature of the work itself. Startups like Thinking Machines provide researchers with more space to work on advanced models more quickly, away from the organizational complexity that might slow down work within large companies.

In contrast, Meta is going through a phase of internal reorganization, which makes its internal environment different from what it was, and opens the door for some talent to move to other entities.

Between these two paths, a clear result emerges: the market no longer moves only by the size of companies, but by their ability to retain or attract talent.

What the TechCrunch report presents here is not just news about employee transfers, but a reading of a broader trend, where the movement of individuals has become an indicator of the redistribution of power within the AI sector.

In this context, the question is no longer just who develops the best model, but who possesses the team capable of developing it, and who can maintain it in a market whose features are rapidly changing.