OpenAI Offers Billions to Poach Google AI Researchers as Talent War Heats Up
-
OpenAI is attempting to recruit top AI researchers from Google through multi-million dollar equity incentive plans. This hiring spree comes as OpenAI nears completion of its employee stock sale, with the company's valuation expected to double to over $80 billion.
Image Source Note: The image is generated by AI and licensed by the service provider Midjourney
OpenAI's recruitment efforts are primarily focused on senior AI researchers responsible for Google's Gemini AI model. According to The Information, OpenAI recruiters are offering these researchers annual compensation packages mainly ranging between $5 million to $10 million, primarily in the form of stock. This move aims to strengthen OpenAI's capabilities in the artificial intelligence field and further drive the company's technological innovation.
It is noteworthy that, due to changing market conditions, OpenAI has not only been generous with senior employees but has also increased salaries for junior staff. This demonstrates the fierce competition for talent, with the company willing to spend heavily to retain and attract top professionals.
In this talent war, Google recently managed to poach some high-profile researchers from OpenAI by offering higher salaries, further heating up the competition between the two.
Sam Altman, one of OpenAI's founders, revealed internally that Google will maintain its advantage in computational resources until Microsoft provides more AI chips next year. This highlights that OpenAI and Google are not only fiercely competing for talent but also battling in the hardware domain.
As major AI companies seek alternatives to Nvidia's dominance in the AI chip market, Microsoft is collaborating with AMD to develop the next-generation AI chip MI300X, which is expected to be released in the fourth quarter.
Since at least 2015, Google has been developing and using its own AI chips called Tensor Processing Units (TPUs). TPUs are specifically designed for high-volume, low-precision computing, offering more input/output operations per joule compared to Graphics Processing Units (GPUs).
Google has utilized TPUs in various applications, including AlphaGo's match against Lee Sedol, Google Street View text processing, Google Photos, and training its current AI models. Since 2018, TPUs have been made available to third parties through Google's Cloud TPU service.