Silicon Valley AI Talent War Intensifies, Meta Launches Aggressive "Talent Acquisition", OpenAI Exclaims: Our Home Was Stolen!

According to media reports, OpenAI is currently experiencing an unprecedented talent loss crisis. In the past week, eight top researchers have left one by one to join Meta, including four Chinese core researchers, who are all leaders of OpenAI's core projects, including o3, GPT-4 series, and other key models.

According to a previous podcast by OpenAI CEO Sam Altman, Meta has offered some researchers up to $100 million in signing bonuses. However, Meta executives refuted this figure internally.

OpenAI is facing a challenge in retaining its talent. In an internal memo sent to employees on Saturday, OpenAI Chief Research Officer Mark Chen stated that the company is "realigning compensation" and promised to take "creative ways to recognize and reward top talent."

According to Wired, Chen expressed extreme shock and dissatisfaction in the memo, writing, "I now have a deep feeling that someone has broken into our home and stolen something. Please believe we are not standing idly by."

OpenAI Work Pressure Too High, Meta Seizes the Opportunity to "Recruit from the Wall"

According to media reports, Meta is significantly increasing its research recruitment efforts, especially focusing on talent from OpenAI and Google. Zuckerberg has always adopted a particularly proactive recruitment strategy, even personally contacting potential candidates.

It is reported that OpenAI employees face intense work pressure of 80 hours per week, which has created opportunities for competitors to recruit. Chen also acknowledged in the memo that the company has been overly focused on regular product releases and short-term comparisons with competitors, and now needs to refocus on achieving general artificial intelligence as the "main task."

To stabilize morale, Chen promised to confront this talent battle head-on with CEO Sam Altman and other leaders. He clearly stated in the memo that the leadership "is not standing idly by," but is taking action with an unprecedented proactive attitude. The core measures include re-evaluating and adjusting the compensation system and designing more creative incentive schemes. However, Chen also emphasized that he has "high personal standards of fairness" and will not undermine fairness to other employees to retain certain individuals.

The memo also included messages from seven other research leaders, urging employees to stay calm and communicate with management when receiving offers from Meta.

The loss of personnel has forced OpenAI to reflect on its internal management and strategic direction. To alleviate employee stress, OpenAI has decided to arrange a week of collective vacation, allowing employees to "recharge." However, a leader in the memo warned that Meta knows OpenAI will take a collective vacation "to recharge" this week and may use this opportunity to exert pressure and make "ridiculous, time-limited explosive offers."

Four Chinese Top Executives, a Major Loss for OpenAI

The four Chinese researchers who were lured away by Meta are all recognized technical backbone within OpenAI, including Jiahui Yu, the leader of the o3, o4-mini, and GPT-4.1 projects, Hongyu Ren, the creator of the o3-mini and o1-mini models, Shuchao Bi, the head of the post-training team for multi-modal models, and Shengjia Zhao, a key contributor to the GPT-4 and o1 projects.

Jiahui Yu is the Head of Perception Research at OpenAI. He graduated from the University of Science and Technology of China (USTC) and earned his Ph.D. from the University of Illinois Urbana-Champaign. He joined Google Brain in 2019, then moved to Google DeepMind in 2020, and joined OpenAI in October 2023. During his time at OpenAI, Yu led the o3, o4-mini, and GPT-4.1 projects, served as a consultant for GPT-4o image generation research, and was responsible for developing the "Thinking with Images" project for image reasoning.

Hongyu Ren is the creator of the o3-mini and o1-mini models. He graduated from Peking University and earned his Ph.D. from Stanford University, joining OpenAI in July 2023. He is also a core contributor to the o1 project and the head of the GPT-4o mini, mainly responsible for the post-training team, dedicated to training models to think faster, deeper, and more keenly.

Shuchao Bi is the head of the post-training team for multi-modal models at OpenAI. He graduated from Zhejiang University and earned his Ph.D. from the University of California, Berkeley. He previously served as a director at Google and co-founded YouTube Shorts. His research areas at OpenAI cover cutting-edge technologies such as new paradigms for post-training, multi-modal reasoning, and development of multi-modal agents.

Shengjia Zhao is a key contributor to the GPT-4 and o1 projects. He graduated from Tsinghua University and earned his Ph.D. from Stanford. He joined OpenAI in June 2022 and focused on the development of ChatGPT and the GPT-next project.

This article is from Wall Street Observer. Welcome to download the APP to view more.

Original: https://www.toutiao.com/article/7521536332884083252/

Disclaimer: This article represents the views of the author. Please express your opinion below using the [Like/Dislike] buttons.