Former OpenAI CTO Mira Murati is preparing to launch her own artificial intelligence startup, joining several other former OpenAI executives who have recently struck out on their own in the AI industry.
According to sources speaking to Reuters, Murati is already in discussions with venture capitalists to secure funding for her new venture. The planned company will reportedly focus on developing AI products based on proprietary models. It's not yet clear if Murati herself will take on the CEO role.
While talks are still in early stages, Murati's reputation and the significant capital required to train custom AI models mean the new company could potentially raise over $100 million, sources say. Exact funding targets have not been disclosed.
At OpenAI, Murati played a key role in projects like ChatGPT and DALL-E, as well as the company's partnership with Microsoft. She may receive support from Barret Zoph, OpenAI's former head of post-training for AI models, who left the company on the same day as Murati in late September.
If Murati is indeed pursuing proprietary models, she likely has a unique approach in mind. The AI model market is already a red ocean, with many commercial and open-source models offering similar performance levels. These models often become outdated within weeks, despite costing millions of dollars to develop.
What did Ilya not see?
Murati was one of the OpenAI members who shouted the now infamous slogan "OpenAI is nothing without its people" during Sam Altman's brief ouster in November 2023. And she's not the only former OpenAI leader to strike out on her own.
Several other prominent staff members have recently left to pursue new projects. Most notably, Ilya Sutskever, co-founder and former chief scientist of OpenAI, has launched "Safe Superintelligence Inc.", which focuses on developing safe ASI. The startup recently raised $1 billion in funding, which is remarkable for what is essentially a research project with no short- or medium-term revenue in sight.
Andrej Karpathy, a leading AI researcher at OpenAI and former head of AI at Tesla, founded a company to create new AI-based educational opportunities.
Jan Leike, former co-leader of OpenAI's Superalignment team, joined competitor Anthropic to work on AI safety. John Schulman, another OpenAI co-founder and a key developer of ChatGPT, also joined Anthropic.
The reasons for these departures vary, ranging from concerns about OpenAI's priorities to internal conflicts. They also suggest that former OpenAI employees continue to see significant potential in AI technology but are looking for new ways to pursue their visions.
This wave of departures and new startups raises questions about the true extent of OpenAI's lead in the field. If the company were on the verge of releasing transformative artificial general intelligence (ASI or AGI), it seems unlikely that well-informed former employees would consider launching competing ventures or that investors would fund such efforts.
So maybe it's not what Ilya saw, but what he didn't see.
