In the latest chapter of the talent war, Meta has successfully recruited Trapit Bansal, an influential researcher from OpenAI and key contributor to the development of reasoning models, to bolster its new superintelligence team.
Article contents:
- Major catch: Meta recruits a key mind from OpenAI
- Who is Trapit Bansal and why does he matter?
- Zuckerberg’s hiring spree: Multi-million dollar offers
- What are reasoning models and why are they important?
- Strategic importance for Meta
- Sam Altman’s response: Attempting to downplay the loss
- The reasoning race heats up
Major catch: Meta recruits a key mind from OpenAI
In a fresh escalating move within Silicon Valley’s fierce talent war, Meta has successfully hired influential OpenAI researcher Trapit Bansal to work on artificial intelligence reasoning models within its new superintelligence unit. OpenAI confirmed the departure of Bansal, whose LinkedIn profile indicates he left the company in June.
This hiring represents a blow to OpenAI and a major gain for Meta, which is striving to strengthen its capabilities in one of today’s most crucial AI research areas.
Who is Trapit Bansal and why does he matter?
Bansal worked at OpenAI since 2022 and was a key player in initiating the company’s work on reinforcement learning alongside co-founder Ilya Sutskever. More importantly, he is listed as a primary contributor to OpenAI’s first reasoning model, o1.
This expertise makes him immensely valuable, as he can provide a major boost to Meta’s superintelligence lab and help the company develop a leading reasoning model capable of competing with industry-leading technologies such as OpenAI’s upcoming o3 or DeepSeek’s R1.
Zuckerberg’s hiring spree: Multi-million dollar offers
This hire comes in the context of a fierce campaign led by Mark Zuckerberg to build Meta’s new artificial intelligence team. In recent months, Zuckerberg has been willing to offer compensation packages reaching up to $100 million for top researchers to join his company. Although it remains unclear what was offered to Bansal to join, it is evident that Zuckerberg is succeeding in attracting top talent.
Three other former OpenAI researchers—Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai—joined Meta’s superintelligence team in recent weeks. Bansal will join them, alongside other prominent researchers from Google DeepMind and other startups.
What are reasoning models and why are they important?
Reasoning models represent the next frontier in artificial intelligence development. Instead of simply generating quick responses, these models are designed to “think” through problems before providing an answer. By utilizing extra computational time and resources to work through a problem step-by-step, AI labs have found success in significantly improving AI performance on standard benchmarks and complex real-world tasks.
This reasoning capability makes artificial intelligence more reliable and capable of solving difficult problems, which is critical for building truly advanced AI systems.
Strategic importance for Meta
Developing reasoning models is a key area where Meta’s superintelligence team must succeed. Currently, Meta does not publicly offer a reasoning model, which puts it behind its competitors. Meta’s superintelligence lab could become a major internal group supporting products across the company, much like Google’s DeepMind unit.
Meta has major ambitions to build business AI agents, and to build competitive agents, Meta needs to develop leading reasoning models to power them. Hiring Bansal is a crucial step in this direction.
Sam Altman’s response: Attempting to downplay the loss
In a recent podcast, OpenAI CEO Sam Altman said that Meta has been trying to poach top talent from his startup, but claimed that “none of our best people have decided to take his offer.” The departure of an influential researcher like Bansal, who played an essential role in key projects, appears to contradict this statement and highlights the intensity of the competition for rare talent.
This statement could be read as an attempt by Altman to minimize the scale of the loss and reassure his team and investors.
The reasoning race heats up
With the addition of Bansal and other key researchers, Meta hopes to pull ahead in the artificial intelligence race. This could prove difficult given that OpenAI plans to release an open-source reasoning model in the coming weeks—a move that may add further pressure to Meta’s open-source offerings. The battle for the top minds capable of building the next generation of artificial intelligence is growing fiercer, and Meta appears willing to pay any price to ensure it does not fall behind.
