Mira Murati, former tech chief of OpenAI, unveils new AI startup

Mira Murati, once the secret weapon of OpenAI as its chief technology officer (CTO), has launched a new artificial intelligence (AI) start-up aimed at broadening access to the technology.

On Tuesday, the 36‐year‐old unveiled Thinking Machines Lab (TML), a product and research organisation that seeks to make “AI systems more widely understood, customisable and generally capable.”

A blog post on the start-up’s website explained that “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.”

Based in San Francisco, TML has also recruited several senior former OpenAI employees, including co-founder John Schulman, Jonathan Lachman (the former head of special projects), and Barret Zoph, who previously served as vice-president.

Lilian Weng, OpenAI’s former vice president of safety, has also joined the startup, as per Bloomberg.

In addition to her role as interim chief executive during the failed coup against founder Sam Altman, Murati has built a team that includes researchers and engineers from competitors such as Google, Meta, Mistral, and Character AI.

This team will focus on developing models specifically geared toward scientific research and programming.

“Scientific progress is a collective effort,” Thinking Machines Lab stated. “We believe that we’ll most effectively advance humanity’s understanding of AI by collaborating with the wider community of researchers and builders.”

The organisation also plans to share its work openly by publishing technical blog posts, research papers, and code, because it believes that “sharing our work will not only benefit the public but also improve our own research culture.”

Mira spent more than six years at OpenAI. During her watch, Murati led efforts to establish ChatGPT as a standalone product and contributed to key advancements in the company’s large language models (LLMs).

In November 2023, following claims that Sam Altman was not “sufficiently candid” with the board, OpenAI’s directors named Murati interim chief executive — though Altman was reinstated days later amid protests from employees and investors.

Lachman confirmed his new role as founding Head of Operations at Thinking Machines Lab in a LinkedIn post, stating: “I am excited and beyond grateful to be joining the founding team of Thinking Machines Lab, a new AI research and product company led by Mira Murati, alongside a remarkable group of colleagues from OpenAI, Anthropic, Character.AI, Google DeepMind, Mistral AI, and Meta.”

Meanwhile, Ilya Sutskever, an OpenAI cofounder and chief scientist involved in the earlier leadership dispute, has left the company to launch a new venture called Safe Superintelligence. His start-up raised $1bn in September to develop safe AI systems that aim to achieve human-level or superior intelligence.

TML explained how its AI intelligence models and products will support more “human-AI collaboration”:

“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,” the company said. 

Out of the nearly 30 current staff members listed in the blog post, more than a dozen were previously at OpenAI, according to  those employees’ public LinkedIn profiles. 

Bloomberg reported that Murati has been in talks with venture capital firms about a funding round, according to people familiar with the matter. 

In recent months, she was said to be seeking about $1 billion, one of the people said, speaking on condition of anonymity to discuss private conversations.

The company declined to discuss funding plans.

While TML does not have a product or model out yet, it claims to have a different philosophy than some other AI companies. 

The startup is having researchers and product leaders “co-design” in tandem to “make AI systems more widely understood, customizable and generally capable,” according to the blog post.

“Instead of focusing solely on making fully autonomous AI systems, we are excited to build multimodal systems that work with people collaboratively,” the startup said,  referring to AI models that can work across mediums such as text, audio or video. 

The company is building models designed to excel in domains like science and programming, with an aim to unlock new breakthrough discoveries in those areas.

TML continues to hire talent in areas such as machine learning and research management, per job listings online.

With inputs from Bloomberg.