Noam Shazeer, a researcher and engineer regarded as one of the contributors to the Transformer architecture, has left Google’s Gemini project to join OpenAI. Shazeer most becomes part of the team behind ChatGPT after several years working at Google.
Career background and significance
Shazeer is less well known to the general public than figures like Sam Altman, but he is a prominent authority within the AI research community. He joined Google in the early 2000s and spent years working on search and language model development. His most notable contribution is linked to the 2017 paper "Attention Is All You Need," which introduced the Transformer architecture that later underpinned ChatGPT, Gemini, Claude and most modern large language models.
Within the professional community Shazeer is credited as a key creator of the Transformer; many of today’s chatbots and AI systems likely would not exist in their current form without the solutions he helped develop.
Previous departures and founding Character.AI
This is not Shazeer’s first departure from Google. At Google he worked with Daniel De Freitas on the Meena chatbot, which was ahead of its time in several respects, but Google chose not to release it widely. As a result, Shazeer and colleagues left in 2021 to found Character.AI. The startup rapidly became one of the fastest-growing AI companies.
Following Character.AI’s success, Google reportedly struck a deal last year worth roughly $2.7 billion (about 950 billion forints) to obtain usage rights for the technology and brought Shazeer back. He then took a leading role in developing Gemini, Google’s answer to ChatGPT.
Why this move matters
Shazeer’s move is noteworthy because the competition between Gemini and ChatGPT has long ceased to be purely technical. Google’s strategy focuses on embedding AI into its existing ecosystem—search, Android, Workspace apps and cloud services—while OpenAI has built a global platform from a single product that has expanded into developer and business ecosystems.
In recent months OpenAI has rolled out new capabilities at a rapid pace, including multimodal features, agent systems and automated workflows. As model differences narrow, hiring top researchers and engineers has become a primary source of strategic advantage; major breakthroughs still often come from the work of a few outstanding individuals.
Implications for the future
Shazeer joining OpenAI may signal that the company is preparing its next development cycle. OpenAI has been speaking increasingly about more advanced autonomous systems, more complex agents and steps toward general artificial intelligence. Recruiting a specialist who helped create the Transformer is more than a routine hiring: it represents a strategic investment in building next-generation systems.
Sam Altman, CEO of OpenAI, responded to the move by noting that Shazeer is among the few experts he wanted to work with from the early days. The ongoing evolution of AI will continue to depend heavily on how leading companies attract and deploy top research talent.
