Former OpenAI CTO Mira Murati launches Thinking Machines Lab

Former OpenAI CTO Mira Murati has launched a new AI startup, Thinking Machines Lab Inc., which will focus on developing multimodal models capable of processing both text and multimedia files. The company, which includes former OpenAI research executive Barret Zoph as CTO and John Schulman as chief research officer, aims to prioritize infrastructure quality to enhance research productivity. Unlike OpenAI’s models, Thinking Machines Lab’s models will not be specifically tailored for programming and math tasks but will instead adapt to a broader spectrum of human expertise. The startup plans to offer customizable AI models and intends to open-source some components of its AI stack to collaborate with the wider research community.

Thinking Machines Lab, which currently employs 29 people, including former staffers from OpenAI, Google, Meta, and other major AI players, plans to focus on AI safety by publicly releasing code and technical assets to prevent harmful AI output. The company will use established methods like red-teaming to test algorithm safety. While the startup did not address previous reports of raising over $100 million in funding, it emphasized its commitment to advancing AI understanding through collaboration and frequent publication of technical content.

Key takeaways

Former OpenAI CTO Mira Murati has launched a new AI startup, Thinking Machines Lab Inc., focusing on developing multimodal models.
The startup's initial team includes former OpenAI research executive Barret Zoph as CTO and John Schulman as chief research officer.
Thinking Machines Lab aims to prioritize infrastructure quality and plans to make its AI models customizable for a broad range of applications.
The company intends to open-source some components of its AI stack and focus on AI safety by releasing technical assets and using methods like red-teaming.

Former OpenAI CTO Mira Murati launches Thinking Machines Lab - SiliconANGLE

Key takeaways

Discussion (0)