According to several sources familiar with the matter, Open researcher Jason Wei Meta is joining the new sprintlene lab.
According to his personal website, V worked on Openi’s O3 and deep research models. He joined Openi after a stent in Google in 2023, where he worked on China -off -thought research, which included AI model training to implement complex questions. In the open, V -reinforcement became a “dye hard” to learn, a way to train or improve the AI model with positive or negative views. It has become a passionate area of AI research.
A source tells the Wired that another Openi researcher, Hong wins Ching, will also be included in the meta. Many sources confirm that both V and Chung’s internal open AI slack profiles are currently inactive. Open, Meta, V and Chung did not immediately respond to the wired comments requests.
According to Chung’s personal website, Chung worked on some similar projects in Openi, including deep research and Openi’s O1 models. The website says its research is primarily focused on reasoning and agents. Chung also overlap with V in Google, and joined the open like V -LinkedIn profiles.
Wired from several sources, Wei and Chung have a close relationship. Meta has previously defeated groups of researchers who have experience working together for their new Superintendeal Lab, including a three -in -three Switzerland office office that joined Google’s Chat GPT macker.
Meta has been going on an illegal hunt last month, offering up to $ 300 million in four years. The Wired reported last month that Mark Zuckerberg, CEO of Meta, sent an internal memo to the staff that introduced a new plan for the company’s AI effort. It included a list of new staff for the Sprinteline team, most of which was recruited from the open.
There are no signs of slowdown in rented frenzy, and openness is fighting. Just last week, Wired reported that Openi had recruited four high -ranking engineers from Tesla, Z and Meta.
On Tuesday, V shared a post on social media that reflected that he called a “important lesson” that learned to learn that he taught him about “living his life”.
We wrote, in life, (and when you make an AI model), imitation is good and you have to do it first. But “the teachers need to walk their path and take risks and rewards from the environment to defeat.”
