Nudge.

Here is a thing that sounds like science fiction, told the way an equity analyst might tell it. There is a large, extremely valuable, extremely poorly served market - the inside of your head - and for most of human history the only ways to intervene in it have been crude. You can swallow a pill that floods your entire body to nudge one region of your brain. You can, if things are dire, let a surgeon drill through your skull and place electrodes. Or you can do nothing, which is what almost everyone does, because the other two options are, respectively, imprecise and terrifying.

Nudge, a San Francisco company founded in 2024, thinks there is a third door. Its pitch is that you can reach deep into the brain - the amygdala, say, or a circuit implicated in chronic pain - with beams of focused ultrasound, aimed from outside the skull with millimeter precision, and gently turn the activity of that circuit up or down. No incision. No implant. You put on a headset, and sound waves converge on a spot the size of a grain of rice, somewhere behind your eyes.

The company that is trying to build this is led by Fred Ehrsam, who you may remember co-founded Coinbase and spent years as a partner at the crypto venture firm Paradigm. This is a notable career move. Ehrsam spent a decade making it easier to move money around; his second act is a hardware company trying to move brain states around. The through-line, if you want one, is that both are enormous, heavily regulated, infrastructure-shaped problems where the interesting part is the plumbing nobody sees.

The physics is the whole company

The reason this hasn't already happened is that skulls are annoying. Bone scatters and distorts ultrasound, which is why aiming a focused beam through it and hitting a precise deep target is genuinely hard - a stack of problems in physics modeling, signal processing, and real-time feedback that has to be solved all at once. Nudge's argument is essentially that the modeling and compute have finally gotten good enough to correct for the skull's distortions and steer the beam where you want it, and to image what's happening at the same time. Its first human-ready system, Nudge Zero, is the embodiment of that claim - and, importantly, it is real. It exists, it is worn by people, and it has been in near-daily use in research studies to refine the platform.

That last detail matters more than it might seem. A lot of neurotech is beautiful renderings and confident timelines. Nudge Zero is a working device generating data today. The order of operations here - build the thing, put it on heads, learn - is the difference between a company and a pitch deck.

Clinical first, headphones later

Nudge is playing what its investors politely call a "clinical-to-consumer arc." Translated: start where the stakes justify hospital-grade, MRI-guided rigor - acute medical conditions like chronic pain, depression, anxiety, OCD, PTSD, addiction - and only later shrink the technology into something that looks and feels like a pair of headphones, for healthy people who just want to sleep better or focus harder. This is the sensible order. It is much easier to earn trust treating a hard medical condition and then move down-market to wellness than to start by promising strangers you'll adjust their mood and hope regulators shrug.

It is also where the story gets genuinely strange, in a good way. The far-future version of Nudge is a device that lets you change how your brain is behaving the way you'd change a thermostat - a little more focus here, a little less anxiety there, break a habit on the way out the door. You should be appropriately skeptical of that, and Nudge, to its credit, mostly talks about the clinical near-term. But the reason $100 million showed up is that if the platform works at all, the ceiling is "how humans feel," which is not a small market.

In through the skull, or around it

The obvious comparison is Neuralink, and Nudge invites it, partly by having hired several of Neuralink's engineers - including its president and CTO, Jeremy Barenholtz, who ran product and technology there. The philosophical split is clean. Neuralink goes in: surgery, implanted electrodes, the highest possible bandwidth, and all the risk that comes with opening a skull. Nudge stays out: lower resolution than an implant, sure, but no surgery, no scar, and a device you can simply take off. One approach treats invasiveness as the price of fidelity. The other treats it as a bug to engineer around.

Both can be right for different jobs, and the honest answer is that nobody yet knows exactly how much a non-invasive ultrasound system can do. That's what the research studies and the Early Feasibility Study underway in San Francisco are for. But the strategic logic of "reach a billion healthy people who will never consent to brain surgery" is hard to argue with, and it's why the non-invasive camp - Nudge, Openwater, Neurosona and others - keeps attracting capital.

The money, and what it's buying

The $100 million came from Thrive Capital and Greenoaks, two firms that write large checks into companies they expect to be category-defining, and it closed in July 2025 with the law firm Goodwin handling the paperwork. What's worth noticing is what the money is not buying: there is no product on shelves, no revenue line, no consumer waitlist to point to. Investors here are underwriting a platform and a team - the belief that if you can safely and precisely reach the brain from outside it, the list of things you can eventually do is very long. Reporting has put the company's valuation somewhere in the neighborhood of a billion dollars, though Nudge hasn't confirmed a figure, so treat that as directional rather than exact.

That kind of pre-revenue conviction only works with the right people in the building, which is where the hiring becomes the strategy. Pulling Neuralink's former product-and-technology lead and several of its engineers into a company built on the opposite technical premise is a pointed bet: the best people to pressure-test the invasive approach are the ones who lived inside it and concluded there was another way. The team is small - roughly 32 to 40 people spanning hardware, machine learning, embedded systems, neuroscience and clinical operations - which for a company this ambitious reads as focus rather than caution.

Why any of this should interest a normal person

Strip away the neuroscience and Nudge is making a bet about defaults. The default state of the human brain is that you mostly can't intervene in it deliberately - you can't decide to be less anxious the way you decide to be warmer. Medicine has chipped at that with drugs and, at the extreme, surgery, but the everyday middle has stayed off-limits. Nudge is wagering that focused ultrasound moves that middle into reach: first for people whose conditions make the current options worth the trade-offs, and eventually, maybe, for everyone. That's a big "eventually," and it deserves scrutiny rather than applause. But it is a specific, testable claim with a working device attached, which is more than most companies selling the future can say.

So that's Nudge: a small team - somewhere around 32 to 40 people - a working prototype, a $100 million Series A led by Thrive Capital and Greenoaks that closed in July 2025, and a plan that starts in a clinic and ends, if everything works, on your nightstand. The technology may or may not reach the ambitious version. But it is a rare thing to see a founder walk away from an enormous, finished company to spend the next decade on a physics problem this hard. Whatever else Nudge turns out to be, it is not a hedge.