What They Did
Character AI was a popular chatbot platform that let users create and interact with conversational AI characters. Founded by Noam Shazeer and Daniel De Freitas, both former Google researchers who helped create the transformer architecture that underpins modern LLMs, the platform attracted millions of users who engaged with AI personas for entertainment, roleplay, and companionship.
How LLMs Killed Them
Character AI became another casualty of Big Tech's AI talent consolidation. Google paid $2.7 billion for a non-exclusive license to the company's technology and to hire back its cofounders. The deal effectively gutted the company's leadership and technical vision. Character AI shifted away from training its own LLMs to a cheaper, more sustainable business model, but lost its founding team and competitive edge.
Timeline
- August 2024: Google paid $2.7 billion for a non-exclusive technology license and to rehire cofounders Noam Shazeer and Daniel De Freitas.
- 2024: Company shifted strategy away from training proprietary LLMs to reduce costs.
- 2024: DOJ opened an investigation into the deal's structure and potential antitrust implications.
By the Numbers
- $2.7B paid by Google for license and to rehire founders
- 2 cofounders returned to Google (Shazeer and De Freitas)
- Both cofounders were former Google employees who helped create the transformer architecture
- DOJ investigation opened