Ex-OpenAI CTO Builds AI That Works With You, Not Instead of You

What if the future of artificial intelligence isn't about replacing humans, but empowering them instead?

Mira Murati, former Chief Technology Officer at OpenAI, is betting big on that vision. Her new company, Thinking Machines Lab, just unveiled AI models designed to collaborate with people rather than automate them away.

"At some point we will have super-intelligent machines," Murati told WIRED. "But we think that the best way to actually have many possible futures—good futures—is to keep humans in the loop for as long as possible."

This week, Thinking Machines previewed what they call "interaction models." Unlike typical voice assistants that simply transcribe your words, these new models truly understand how humans communicate in real time.

They pick up on pauses when you're thinking. They adapt when you interrupt yourself or change the subject mid-sentence. They respond to shifts in your tone the way a thoughtful colleague would.

The company demonstrated these capabilities in several videos, though the models aren't publicly available yet. It's a stark contrast to how most big tech companies are building AI today.

OpenAI, Google, and Anthropic are racing to create systems that can write entire software applications from a single text prompt. Humans barely need to lift a finger.

Murati sees a different path forward. Her team wants people to customize their own frontier AI models with their own data, then work alongside those models to achieve their personal goals.

The Ripple Effect

Thinking Machines isn't alone in this vision. Startups like Humans& are also prioritizing human collaboration over automation. Even prominent economists have started calling for AI development that focuses on empowering workers rather than replacing them.

Alexander Kirillov, a founding team member at Thinking Machines, explains the bigger picture. "The model constantly perceives what you're doing and is constantly there to be able to reply and give you information," he says. It's AI that understands your intent and predicts what you might need next.

The company already launched one product called Tinker last October. It lets researchers and engineers fine-tune open source AI models using their own custom data.

Murati raised billions of dollars to build this human-centered future. She left OpenAI in 2024 specifically to pursue this vision with several other prominent engineers.

"This is showing the first bet on human collaboration," Murati explains. "Where this is going is really amplifying people's own preferences and values."

At a time when headlines focus on job losses and AI concentration in the hands of a few tech giants, Thinking Machines offers something refreshingly different: technology that makes humans more capable, not obsolete.