Mira Murati debuts Thinking Machines Lab, her AI startup
Add Axios as your preferred source to
see more of our stories on Google.
Mira Murati. Photo: Thinking Machines Lab
Former OpenAI exec Mira Murati on Tuesday unveiled her new startup's name, Thinking Machines Lab, and its goal: developing AI systems focused on the interaction between humans and AI.
Why it matters: Other startups founded by former OpenAI executives — from more mature AI firms like Anthropic to other just-out-of-the-gate startups like Ilya Sutskever's Safe Superintelligence — have more single-mindedly dedicated themselves to creating AI that's more powerful than humans.
- "Instead of focusing solely on making fully autonomous AI systems, we are excited to build multimodal systems that work with people collaboratively," the company said in a blog post announcing its formation.
- "While current systems excel at programming and mathematics, we're building AI that can adapt to the full spectrum of human expertise and enable a broader spectrum of applications."
Zoom in: Thinking Machines Lab says it has about 30 employees, including a number of Murati's former OpenAI colleagues.
- Murati is CEO, Barret Zoph is CTO and John Schulman is chief scientist. Zoph left OpenAI in September; Schulman departed OpenAI in August for Anthropic, and said earlier this month he was leaving Anthropic for a new opportunity.
Murati isn't disclosing a timeline or specifics on Thinking Machines' first product, nor any details on funding, though the company is confident in its abilities to raise the money it needs.
- The company said it wants to be open in its work, though that doesn't necessarily mean its models will be open source.
Between the lines: One of the key gaps that Murati wants to address is the gulf between the capabilities of AI systems and humans' ability to understand and use them.
- "The scientific community's understanding of frontier AI systems lags behind rapidly advancing capabilities," the blog post said. "Knowledge of how these systems are trained is concentrated within the top research labs, limiting both the public discourse on AI and people's abilities to use AI effectively."
Friction point: One of the key challenges is at the interaction layer.
- Chatbots are good at answering questions, but often struggle to refine or adjust their output.
- "Despite their potential, these systems remain difficult for people to customize to their specific needs and values," the blog post notes.
