In a massive blow to Google’s AI ambitions, Noam Shazeer, the behemoth’s vice president of engineering and a co-lead of its Gemini AI models announced that he’s joining ChatGPT maker OpenAI.
“I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there,” Shazeer – who’s widely recognised as one of the most prominent and influential figures in AI – said in a post on X.
One of Time Magazine’s 100 most influential people in the AI world, Shazeer, was brought back to Google for $2.7 billion in 2024 through an acqui-hire arrangement after he quit the tech giant out of frustration in 2021 to start his own company Character.AI. Shazeer had quit in 2021 because the search giant had refused to release a chatbot he had developed.
“It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you,” Shazeer said on his second-exit from Google.
Sam Altman, the co-founder and CEO of OpenAI could barely contain his excitement on the development. Replying to Shazeer’s post, Altman said, “Noam is one of the people I have most wanted to work with since the very beginning of OpenAI. Only took 10 years. I think it will be worth the wait!”
Shazeer is a Google veteran, having joined the company in its early years in 2000 as one of the company’s first 100 hires. In many ways, he’s one of the key people responsible for catazlysing the AI boom. A seminal 2017 research paper he co-authored is largely seen as the kickstarter to present-day large language models (LLMs) that is the basis of the likes of ChatGPT and Gemini.
Shazeer’s appointment comes as a shot in the arm for OpenAI in the run up to its imminent IPO (initial public offering), as the key appointment will bolster market confidence. Last month, OpenAI confidentially filed for IPO. Rival Anthropic too has fconfidetntially filed for IPO.
Shazeer studied mathematics and computer science at Duke University in Durham, NC, on an academic scholarship. Following Duke, he briefly enrolled in a graduate program at UC Berkeley but left before finishing to join Google.
The Perfect Scorer
Before becoming one of the most prominent figures in artificial intelligence, Shazeer had already shown glimpes of his genius as a teen.
Shazeer represented the USA team at the 35th International Mathematical Olympiad in Hong Kong as a teenager, and won a Gold Medal with a perfect score.
In his very first semester at Duke, he ranked 6th in the US on the prestigious William Lowell Putnam Mathematical Competition. He subsequently anchored the Duke math team to 1st and 2nd place national finishes in 1996 and 1997.
In 2020, Shazeer and his colleague Daniel De Freitas built an advanced conversational chatbot at Google named Meena (later evolved into LaMDA). Shazeer actively lobbied with Google executives to release it to the public, predicting it would completely change search. However, when Google leadership resisted out of safety and reputational concerns, he famously walked out to build his own platform in 2021.
While the rest of the world grapples with job loss fears, AI talent is in an enviable position. Last year, in a bid to boost its AI efforts, Mark Zuckerberg led Meta — that owns Facebook, WhatsApp, and Instagram — acqui-hired Scale AI founder Alexandr Wang to lead its Superintelligence Lab for $14.3 billion.
