Noam Shazeer, CEO of Character.AI, Makes a Return to Google

In a noteworthy development, Character.AI’s co-founder and CEO, Noam Shazeer, is making a return to Google, his former employer. Shazeer left Google in October 2021 to establish the startup backed by a16z. During his earlier tenure at Google, he led the team responsible for creating LaMDA (Language Model for Dialogue Applications), a foundational language model instrumental in the development of conversational AI technologies.

Alongside Shazeer, Character.AI’s co-founder Daniel De Freitas and several other team members are set to join Google. Dominic Perella, Character.AI’s General Counsel, will serve as the interim CEO of the startup, signaling that the core team will remain with Character.AI.

Furthermore, Google has entered into a non-exclusive agreement to leverage Character.AI’s technology.

Shazeer expressed his enthusiasm about rejoining Google and integrating into the Google DeepMind team. “The accomplishments at Character.AI over the past three years fill me with immense pride. I am optimistic that the non-exclusive licensing agreement with Google, coupled with the extraordinary team at Character.AI, will ensure the startup’s ongoing success,” Shazeer commented to TechCrunch.

Google has confirmed Shazeer’s appointment to the Deepmind research team but has yet to detail the specific positions for him or De Freitas.

Google’s statement highlighted the excitement surrounding Noam’s return, acknowledging him as a leading researcher in machine learning. “This partnership will allocate additional resources for Character.AI to expand and concentrate on delivering personalized AI solutions to users globally,” a representative from Google described.

Character.AI has successfully garnered over $150 million in funding, with significant contributions from a16z.

Character AI outlined its founding mission in a blog post, emphasizing the desire for personalized superintelligence necessitating a comprehensive approach, including pre-training and post-training models to create the unique experiences central to Character.AI, as well as developing a product platform with worldwide reach.

With the evolving landscape over the past two years, characterized by the availability of numerous pre-trained models, Character.AI sees merit in integrating third-party LLMs with its proprietary ones. This strategy allows the team to allocate more resources toward post-training and innovating new product experiences for an expanding user base.

Possible scrutiny from regulatory entities, such as the Federal Trade Commission (FTC), the Department of Justice (DoJ) in the U.S., and the EU, over such reverse acqui-hires may emerge. For instance, last month, the U.K’s Competition and Markets Authority (CMA) released a notice examining Microsoft’s strategy of hiring pivotal personnel from Inflection AI to determine if it represents an attempt to circumvent regulatory scrutiny. This follows the FTC’s decision in June to initiate an antitrust probe into Microsoft’s $650 million transaction.

For further information, this reporter can be contacted via email at im@ivanmehta.com and on Signal at ivan.42.