Thinking Machines Lab Is OpenAI Ex-CTO Murati’s New AI Startup

thinking-machines-lab-is-openai-ex-cto-murati’s-new-ai-startup
Thinking Machines Lab Is OpenAI Ex-CTO Murati’s New AI Startup

If you have been closely following OpenAI, you probably know who Mira Murati is. She used to be the CTO before she decided to leave the AI tech company last year in September. After about four months of vanishing from public view, Mira Murati has finally come forward to share that she is actively working on her AI startup, Thinking Machines Lab.

OpenAI’s former CTO, Mira Murati, announces her new AI startup, Thinking Machines Lab

The former OpenAI CTO started Thinking Machines Lab with thirty researchers and engineers from competitors like OpenAI, Meta, and Mistral. It’s worth noting that one-third of Thinking Machines Lab’s team comprises former OpenAI leaders. One popular name that works with Murati’s new AI startup is John Schulman, who was co-founder of OpenAI.

Barrett Zoph, former leader of the post-training research team at OpenAI, has joined Mira Murati’s startup as CTO. While there’s no mention of releasing any products, Murati’s startup website mentions its goal is to “make AI systems more widely understood, customizable, and generally capable” for bridging gaps.

See also  LG Unveils Ultra-Light "Hybrid AI" Gram Laptops

Thinking Machines Lab also promises public transparency by frequently publishing technical blog posts, papers, and codes. It will further focus on building “multimodal systems that work with people collaboratively” instead of making fully autonomous AI systems.

AI safety and infrastructure quality will be a major focus area of the startup

AI safety will be another focus area of Thinking Machines Lab. The startup will prevent the misuse of models it will release in the future as it “maximizes users’ freedom.” Infrastructure quality will also be a top priority as the startup aims to “build things correctly” without taking any shortcuts.

A part of Murati’s tweet from earlier in the day reads, “Our goal is simple: advance AI by making it broadly useful and understandable through solid foundations, open science, and practical applications.