Mira Murati is a prominent AI engineer who worked on several of OpenAI’s products, including ChatGPT, which started the AI revolution in late 2022. She served as the company’s CTO until her departure last September, and you might remember her as the interim CEO of OpenAI during the brief coup in late November 2023 that saw Sam Altman ousted from the firm and then rehired as CEO.
You might remember Murati from OpenAI’s live events, like the introduction of GPT-4o in May, or from certain interviews.
Murati infamously said that some of the creative jobs the AI will replace should not have existed. Her answers concerning OpenAI’s training of the text-to-video Sora model were more annoying. The CTO then said she didn’t know whether Sora was trained with specific video sources, such as YouTube.
Or you might not remember Murati at all, given that we’re still in the early days of AI, a technology not many people will follow closely.
Whatever the case, you should know Murati is leading a mysterious AI firm of her own, as it happened with other OpenAI engineers who departed the company. And the ChatGPT creator has certainly bled talent since the internal turmoil that saw Altman briefly deposed. However, it’s unclear what Murati is working on, and I’m curious to see what sort of AI product we expect from her startup.
Of course, Ilya Sutskever is the most prominent departure from OpenAI. He left last summer and quickly formed Safe Superintelligence Inc. (SSI), a company that will focus on developing artificial superintelligence (ASI). The former OpenAI exec said that ASI would be the only product from the company. Make that safe superintelligence, as the company’s name suggests. We’ll never see a commercial product from Sutskever & Co. until that level is attained.
Mira Murati doesn’t have the same prestige as Sutskever. While anyone would be ready to wait for the latter to deliver on that safe ASI promise and fund the startup generously, Murati can’t expect the same generosity from investors.
It’s unclear what Murati is working on, but Wired has learned that her mysterious AI startup has started hiring engineers and researchers from OpenAI and other AI firms. Murati won’t pivot to a non-AI field after working on ChatGPT, Sora, and other OpenAI products.
Wired’s report focuses on Jonathan Lachman, the previous head of special projects at OpenAI, calling him a “major get” for Murati’s startup. Lachman left OpenAI recently to join the new company. Murati reportedly poached 10 researchers and engineers from OpenAI, Character AI, and Google DeepMind.
The report further explains that while Murati is looking to explore AGI (artificial general intelligence), the key evolution of current AI that will lead us to ASI, the startup doesn’t have a name or a firm product direction.
As for funding the new startup, Murati was trying to raise $100 million for her company in mid-October. Wired notes that the figure has not been finalized. Comparatively, Sutskever raised $1 billion for his ASI firm in the three months since he left OpenAI.
The report drops at an interesting time, considering the chatter on social media about OpenAI’s potential AGI and ASI breakthroughs. As I said earlier, it’s not just OpenAI working on such big innovations; other parties might be more enigmatic. Sutskever is the best example; his firm won’t launch commercial products.
If Murati is working on any AGI-to-ASI products, it’ll be interesting to see whether it targets customers with commercial AI software or adopts an approach similar to Sutskever’s.
The post Mira Murati, a former high-ranking ChatGPT exec, has a mysterious AI project of her own appeared first on BGR.
