Sam Altman’s Merge Labs Taps Caltech Scientist for Non-Invasive Brain Tech

A Different Path to Brain-Computer Interfaces

Sam Altman is quietly assembling the pieces for his next major venture, and it’s shaping up to be a direct challenge to Neuralink’s vision of how humans should interface with machines. According to sources familiar with the matter, Altman has recruited Caltech professor Mikhail Shapiro—an award-winning biomolecular engineer—to join Merge Labs, a brain-computer interface startup expected to launch publicly in the coming weeks.

Shapiro’s hiring reveals crucial technical details about where Altman wants to take this technology. Unlike Neuralink’s approach of implanting electrodes directly into brain tissue, Shapiro’s research at Caltech has focused on using sound waves and magnetic fields to interact with neurons. His lab has pioneered non-invasive techniques for neural imaging and control, specifically exploring how ultrasound can read brain activity without requiring open-skull surgery.

The Ultrasound Advantage

What makes Shapiro’s approach particularly intriguing is how it combines ultrasound technology with genetic engineering. During recent talks, the scientist has explained that rather than inserting physical electrodes into the brain, it’s “easier to introduce genes into cells” that modify them to respond to ultrasound. He’s made this his mission: developing ways to interface with neurons that would be far less invasive than current methods.

This technical direction aligns perfectly with Altman’s publicly stated preferences. At a press dinner in August that was attended by The Verge’s Alex Heath, Altman made his skepticism of Neuralink’s approach abundantly clear. “I would definitely not sew something to my brain” that destroys neurons like Neuralink’s interface does, he stated. Instead, he described wanting technology that could let him “think something and have ChatGPT respond to it,” potentially in a “read-only” capacity that feels more reasonable to him.

Funding and Positioning

Merge Labs appears to be positioning itself as a serious contender in the emerging brain-computer interface space. Sources indicate the startup expects to raise hundreds of millions of dollars from OpenAI and other investors, as The Financial Times initially reported. While Shapiro’s official title remains unclear, he’s reportedly been presented as a key leader in discussions with potential backers and will be part of the founding team alongside Altman and co-founder Alex Blania.

The connection between Altman and Blania is worth noting—they previously co-founded Tools for Humanity, the company behind the controversial Worldcoin eyeball-scanning orb. At Merge, however, Altman is expected to serve as chairman rather than taking a day-to-day operational role, similar to his arrangement at their other venture.

Broader Implications

This development comes at a fascinating moment in the evolution of human-machine interfaces. While companies like Neuralink push forward with increasingly sophisticated implantable devices, Altman appears to be betting that consumers will prefer approaches that don’t require brain surgery. The ultrasound method being developed by Shapiro could potentially offer a middle ground—more direct than external EEG headsets but far less invasive than implanted electrodes.

Altman has been thinking about these convergence points between humans and technology for years. Back in 2017, he wrote about “The Merge” on his personal blog, noting how Silicon Valley was obsessed with predicting when humans and machines would merge or when AI might surpass human capabilities. Most estimates, he observed at the time, fell between 2025 and 2075.

Now, with Merge Labs, Altman isn’t just speculating about that future—he’s actively building what could become one of the key technologies that makes it possible. The recruitment of a respected scientist like Shapiro suggests this is more than just another tech startup; it’s a serious attempt to redefine how humans might eventually interact with advanced AI systems directly through thought.

Neither Shapiro nor representatives for Altman and Blania responded to requests for comment, leaving many questions unanswered about the specific timeline and technical capabilities of Merge’s approach. But the pieces now coming together suggest we’re about to see a significant new player enter the brain-computer interface arena—one that could challenge fundamental assumptions about how these systems should work.