MIT dropout Claire Wang is simulating nervous systems in worms to unlock better brain-computer interfaces

Claire Wang — simulating worms to build better brains

Claire Wang dropped out of MIT as a junior in electrical engineering and computer science to pursue something most labs won't touch: biologically accurate simulations of nervous systems. No company yet. No funding round. Just a bet that the field of brain-computer interfaces is solving the wrong problem.

The BCI critique

Current BCI progress is real but limited. Devices can read rough motor signals, translate them into x-y coordinates, and help paralysed patients move a cursor or a limb. Wang argues that's post-hoc pattern matching: throw enough data at a brain model and it learns a shortcut. The ceiling is low because nobody actually understands the underlying signals.

Simulate the brain accurately enough, and you can decode exactly which regions to activate to produce natural, multi-degree-of-freedom movement rather than a constrained output. That's the gap Wang wants to close.

“A lot of my interests has been in the field of neurotech, which includes a focus in whole brain emulation with a focus on doing this with worms first, but also how useful this could be for brain computer interfaces and understanding consciousness. C. Elegans is 300 neurons — we can't even simulate the C. Elegans, so we have to start there.”

Why C. elegans

C. elegans has 302 neurons, is translucent, and amenable to fluorescence gene therapy that makes imaging far easier than in any mammal. Wang's argument is blunt: we can't even simulate this correctly yet, so that's where to start.

The organism is useful less for what it resembles and more for what it can answer: does voltage data matter, or is calcium imaging enough? Is light-sheet microscopy sufficient, or do you need electron microscopy? Those methodological questions, resolved in a simple system, inform the hardware and data requirements for zebrafish, mice, and eventually humans.

The constraint binding the field right now is imaging, not compute. There is currently no way to image a living mouse brain in full, which is why the stack has to be built from the bottom up.

Why now, independently

Wang sees the current moment as unusually permissive for solo or small-team biotech research: AI tooling is accelerating what one person can do, alternative science funding is growing, and the stigma around "research risk" in startups has largely dissolved. A decade ago, she notes, research risk was a red flag for investors; today it's table stakes across the sector.

She's still early — no co-founders confirmed, scientific bet not fully locked, likely relocating to San Francisco. The Thiel Fellowship gives her runway to keep that open longer than a typical startup formation would allow.