The Brain as Computer: Max Hodak on Engineering the Mind

Lena: Imagine you’re standing in a quiet room, and for the first time in years—maybe decades—you can actually see the light catching the edge of a window. It’s not just a fuzzy shape; you can see the grid of a Sudoku puzzle and actually solve it. That’s not science fiction anymore. It’s happening right now because of a fundamental shift in how we think about the human brain. Today, we’re looking at the work of Max Hodak, the founder of Science and formerly of Neuralink, through a fascinating interview about the future of neural interfaces.

Miles: It’s a pretty provocative starting point, isn’t it? Max Hodak argues that if you want to understand where medicine is going, you have to accept one very specific, very literal premise: the brain is a computer. And he doesn't mean that as a metaphor or a poetic way of describing our thoughts. He means it as a physical reality. You arrange matter in a certain way, you press "go," and it solves computational problems. When you look at it that way, the skull isn't just a bone—it’s a vat. We are, quite literally, brains in vats, connected to the outside world by a few thin cables we call the cranial and spinal nerves.

Lena: That "brain in a vat" analogy really recontextualizes everything. Usually, when we talk about Brain-Computer Interfaces, or BCIs, we think about someone using their mind to move a cursor on a screen or control a robotic arm. It’s "brain-to-computer." But Hodak is flipping that. He’s saying the interface is about treating the brain itself as the central computational system that needs better input and output.

Miles: Exactly. It’s a shift from "how do I talk to my laptop" to "how do I repair or upgrade the processor that is my mind?" And he’s got high confidence in this—he considers it a plain, clear fact of the universe. If the brain is a computer, then blindness or hearing loss isn't just a biological tragedy—it’s a broken peripheral. It’s a cable that’s been cut or a sensor that’s failed. If you can bridge that gap with engineering, you’re not just treating a patient; you’re debugging a system.

Lena: It’s an intellectually rigorous way to look at medicine, but it’s also one that makes people a bit uncomfortable. Hodak mentions that telling the internet the brain is a computer is a great way to make people angry. But if you can move past the philosophical discomfort, the engineering results he’s seeing are hard to argue with. So let’s dive into the actual technology that’s making this real—starting with a tiny chip that’s restoring sight in a way we’ve never seen before.

Miles: So let's talk about the Prima retinal prosthesis. This is the flagship product for Max Hodak’s company, Science, and it just received regulatory approval in Europe as of July 2026—meaning it’s moving from a research project to something that’s commercially available. To understand why this is a big deal, you have to look at what it’s actually doing. It’s a tiny chip, implanted right under the retina at the back of the eye.

Lena: And it’s specifically for people who have lost their "photoreceptors"—the rods and cones that normally catch light—due to things like macular degeneration. But how does the image actually get to the chip? It’s not like they’re just opening their eyes and seeing.

Miles: Right, it’s a system. The patient wears special glasses equipped with a laser projector. That projector beams the visual information onto the chip in the back of the eye. The chip then stimulates the remaining healthy parts of the retina directly, bypassing those dead rods and cones to send a signal up the optic nerve. Hodak calls it a "cochlear implant for the eye".

Lena: That’s a powerful comparison. We’ve seen how transformative cochlear implants are for hearing. But the results Hodak describes from the clinical trials sound even more precise. He mentions patients actually reading books or solving Sudoku puzzles. That’s "form vision"—being able to see actual shapes and patterns, not just vague flashes of light.

Miles: That distinction is crucial. Previous attempts at this, like the Second Sight system from about a decade ago, mostly gave patients flashes of light that they had to mentally piece together. They got paid about $150,000 per patient and offered very limited utility. Prima is a massive leap forward because it allows for that coherent, "mind’s eye" image. We can categorize the existence of this form vision as a [HIGH] confidence result—it’s been demonstrated in the clinic.

Lena: But even with that success, Hodak is very clear about the current limitations. It’s not "Terminator vision" yet. Right now, it’s basically black and white, and the field of view is tiny—he compares it to looking through a straw.

Miles: Exactly. It’s a proof of concept that they’re on the right track, but it’s an engineering floor, not a ceiling. They’re already looking at how to add grayscale and potentially red and green color vision. The goal is to move from hundreds of thousands of potential patients to millions as the technology improves. And that gets us into the "why" of it all. Why tackle the eye first if the ultimate goal is the whole brain?

Lena: One thing that really stands out in Max Hodak’s approach is his skepticism toward traditional drug discovery. He points out that humanity just isn't that good at understanding the molecular details needed to make a drug for something like Parkinson’s or vision loss. You can spend a decade and billions of dollars on a small molecule, run the trial, and the answer might just be "no".

Miles: It’s the "random walk" of biology, as he calls it. And he contrasts that with the predictability of engineering. If you put electrodes in the motor cortex of a paralyzed person, they’ll probably be playing a video game or using a computer within an hour. The "effect sizes" in BCI are massive compared to what you usually see in medicine.

Lena: It’s about choosing a path where you can actually iterate. With the retinal implant, they know exactly how to make it better—depth of gray scale, we think we could see a path to get at least red and green. It’s a compounding engineering process rather than a binary "does this chemical work" gamble. This is why Hodak views Science more as a tech company than a traditional biotech firm.

Miles: And that tech mindset led them to some interesting choices. They didn't just stick to one method. They looked at everything—gene therapy, ultrasound, electrical stimulators. They even developed an in-house gene therapy that’s likely going to human trials next year. But their most ambitious research project—one that feels like the ultimate bridge between tech and biology—is what they call "biohybrid neural interfaces".

Lena: That sounds like something out of a cyberpunk novel. What does "biohybrid" actually mean in this context?

Miles: Instead of just sticking metal wires or silicon chips into the brain, you engraft living neurons. These neurons grow in and form new biological connections with your existing brain. It’s high-risk, long-term research, but it’s part of a 10 to 15 year pipeline they think could revolutionize medicine. We’d categorize the near-term success of biohybrids as [LOW] confidence right now since it’s still in the research phase, but the engineering roadmap for the electrical implants is [HIGH] confidence.

Lena: It’s fascinating because it suggests that we don't have to choose between "natural" and "artificial." We can combine them. But to do that effectively, we have to understand the language the brain speaks. And that brings us to one of the most provocative ideas in the whole interview—the connection between the human brain and Artificial Intelligence.

Miles: This is where things get really "trippy," as Hodak puts it. There’s a theory called the "platonic representation hypothesis," and it’s a major reason why there’s so much excitement in the BCI field right now. The idea is that when you train a massive AI model on a huge amount of data, the mathematical objects it creates—the way it represents concepts—look a lot like what we see in human neuroscience.

Lena: So, if I’m understanding this right, if you give a computer enough data about the world, and you give a human brain years of experience in the world, they eventually "agree" on how to map out reality? Like they’re both grasping at the same underlying truth?

Miles: Precisely. Hodak says they use this practically at Science. They can find "alignments" between neural recordings from animal brains and the internal representations of AI models. It’s as if there’s a "true underlying data manifold" of the universe, and any sufficiently powerful intelligence—biological or silicon—will eventually converge on it.

Lena: That is a massive claim. It suggests that AI isn't just a "gimmick" or a statistical trick, but something that’s actually tapping into the same fundamental logic as our own minds. Hodak even jokes that the best way to do neuroscience these days is to work at a place like OpenAI or Anthropic, because it’s way easier to study the "neurons" in a model than in a living brain.

Miles: It’s controversial, of course. Some people really don't want this to be true. And we should be careful here—the existence of these alignments is being observed and used constructively. It’s being observed and used constructively, but we don't fully understand the "why" or the global structure of it yet.

Lena: But if it is true, it changes the goal of a BCI. It’s not just about building a "brain keyboard" to type faster—which Hodak actually thinks is a bit of a dead end. Instead, it’s about "redrawing the border around your brain". It’s about merging these two systems that already speak a similar language. And that raises some really heavy questions about what it even means to be "you."

Miles: When Max Hodak talks about the future, he doesn't just talk about fixing broken eyes. He talks about "substrate independence". This is the idea that the "human experience" could eventually be sustained on something other than a biological brain. But he’s very skeptical of the way most people think about this—like just "uploading" a copy of yourself to a computer.

Lena: Right, he uses that thought experiment: if you were dying of cancer and someone made a perfect software replica of you, would that make you feel better? Most people would say no, because there’s no "continuity". You’re still you, dying in the hospital, while some computer program goes off and does your job.

Miles: Exactly. It’s not about making a copy; it’s about the "phenomenal continuity" of your experience. We accept "drift" in our identity over time—you’re not the same person you were at five years old—but as long as the experience is continuous, you feel like the same "self". This is why he’s so focused on things like general anesthesia—it’s a break in that continuity that we’ve all just decided to be okay with.

Lena: So the real quest isn't "uploading," it's "expansion." It’s about slowly replacing or augmenting parts of the brain while maintaining that continuous stream of consciousness. Hodak thinks we need a much more fundamental explanation for how the brain partitions experience—why you have "your" vision and "your" hearing and never mine.

Miles: It’s a foundational question. And he’s looking at "connectomics"—the mapping of all the connections in the brain—as a key missing piece. We’re close to a mouse connectome, but a human one is still a ways off. But the goal is clear: if you can achieve substrate independence, you can take the human experience almost anywhere.

Lena: It’s a very different focus than what you see from other BCI companies. While others are trying to let you summon an Uber with your thoughts—which Hodak thinks would be incredibly annoying if it accidentally summoned two because you were just thinking about an Uber—he’s looking at how to make humans less fragile. He wants to solve the "jeopardy" of the human condition.

Miles: And that jeopardy usually comes from our support organs—the heart, the lungs, the pancreas. Hodak calls them "support characters" whose only job is to keep the brain activity interesting. He says he’d be "fairly disappointed" to be "murdered by his pancreas". If we can replace those parts, we solve the leading causes of death, like cardiovascular disease or cancer that metastasized to the brain. But there’s a bottleneck to all this expansion, and it’s not what you might think.

Lena: One of the most surprising parts of Hodak’s philosophy is his take on "brain keyboards" and high-bandwidth communication. You’d think a BCI expert would be all about "mind-reading" at lightning speed, but he points to a very specific, deeply evolved "cognitive bottleneck" of about 10 bits per second.

Miles: It’s a fascinating observation. If you take someone with a perfect memory, put them in a helicopter over Manhattan, and then ask them to draw everything they saw, they can only output information at about 10 bits per second over a few hours. Language, writing, talking—it all seems to roll up to this same slow speed.

Lena: So the idea that we have these perfectly formed, high-bandwidth thoughts just waiting to be "unlocked" by a BCI is probably a myth. Hodak argues that "talking or writing is thinking". You might feel like a thought is fully formed, but it usually isn't until you actually try to express it.

Miles: Exactly. So a "brain keyboard" might be nice for walking down the street and talking to your AI, but it doesn't fundamentally change who you are or how you think. This is a [HIGH] confidence claim based on multiple independent lines of evidence.

Lena: This is why he’s focused on input—vision, hearing, balance—rather than just output. If you can give the brain a "kilobit per second" of motor control and high-fidelity sensory input, you’re "halfway to the Matrix". You’re not just trying to type faster; you’re trying to expand the actual richness of the experience itself.

Miles: And that expansion is what leads to the ultimate goal: interstellar travel. Hodak believes that if we’re ever going to explore the stars, we have to adapt ourselves to that environment. We can’t just "export Earth" with us everywhere we go. Our bodies were designed for this planet, but if we want to survive the vacuum of space, we need to be able to upgrade and replace our parts. Substrate independence isn't just a medical goal—it’s an evolutionary one.

Lena: It’s easy to get lost in the talk of space travel and "the Matrix," but Hodak is also a CEO. He’s very clear that for any of this to happen, Science has to be a real, profitable business. He actually pushes back on the modern definition of a "moonshot" as something that’s just a way to "vaporize investor money".

Miles: Right, he points out that the actual moonshot worked. We left footprints on the moon. To him, a moonshot should be a successful project, not an excuse for failure. And the way you ensure success is by having a product people actually need right now—like the Prima retinal implant.

Lena: And the market for that is huge. One in two people have some early-stage age-related macular degeneration by the time they’re 80. If you can restore even a "minimal" amount of vision, it’s worth a massive amount, both in terms of quality of life and commercial value. He mentions that even a gene therapy that only slows degeneration can reimburse at almost half a million dollars per eye.

Miles: So the strategy is to build a "real, profitable business" in vision first. That gives the company the "ability to do this forever" and fund the much more ambitious, longer-term projects like biohybrid interfaces and substrate independence. It’s a "lateral move" in medicine—instead of trying to solve every single disease, you just make the brain less dependent on the parts that break.

Lena: It’s a pragmatic way to approach a wildly futuristic goal. He even raised his Series A mostly from tech investors because the conventional biotech industry has a very different culture—one he finds less suited to this kind of iterative engineering. He’s betting that the "tech" way of doing things is what will finally crack the code of the human brain.

Miles: It’s a bet on the brain as a computer. If that premise is correct, then all the tools we’ve developed for computers—iteration, debugging, scaling, AI alignment—suddenly become the most powerful tools in medicine. So where does that leave you, if you’re looking at this from the outside?

Lena: If you’re trying to wrap your head around what this means for the future, there are a few key takeaways from Max Hodak’s worldview. First, stop thinking about BCIs as just a way to "talk to computers." Start thinking of them as a way to "repair and expand the brain". The focus is on the brain as the central object, not the laptop or the phone.

Miles: Second, pay attention to the "engineering floor." The current limitations of things like the Prima implant—the straw-like field of view, the lack of color—aren't permanent barriers. They are just the starting point for an engineering process that we already know how to navigate. When you see a "low-res" medical breakthrough, don't dismiss it; look at the path to 2.0 and 3.0.

Lena: Third, be skeptical of the "mind-reading" hype. If the 10-bit-per-second cognitive bottleneck is real—and Hodak is very confident it is—then the real revolution isn't going to be typing with your thoughts. It’s going to be in sensory expansion. It’s about what you can take in, not just what you can put out.

Miles: And finally, keep an eye on the "platonic representation hypothesis". If AI models and human brains are indeed converging on the same way of representing the world, the bridge between us and artificial intelligence is already being built. It’s not just a software connection; it’s a structural one.

Lena: It’s a future that’s both incredibly exciting and deeply provocative. It challenges our most basic assumptions about health, aging, and even what it means to be an individual. But as Hodak points out, we’re already "brains in vats". We’re just looking for a better vat.

Miles: And a better way to connect that brain to the universe. Whether it’s through a chip in the retina or a biohybrid bridge to a computer, the goal is the same: to make the human experience less fragile and more expansive than it’s ever been before.

Lena: As we bring this to a close, it’s worth reflecting on that central idea of "jeopardy". We all live under the constant threat that a single organ—a heart, a lung, a pancreas—could fail and end the entire "interesting" project of our lives. Max Hodak’s work is a bet that we can engineering our way out of that fragility.

Miles: It’s a shift from "curing" the body to "transcending" it. If the brain is the only organ that is, in principle, irreplaceable, then everything else is just a support system that can be upgraded or substituted. It’s a lateral move that could make the very concept of "healthspan" look completely different twenty years from now.

Lena: But it also leaves us with a big question to think about. If you could maintain a continuous stream of consciousness while slowly replacing the biological parts of your brain with something more durable, at what point do you stop being "you"? Or is the "you" just the pattern, regardless of the substrate?

Miles: Hodak would say the pattern is the point, as long as it never stops flowing. It’s a lot to process, and it certainly makes you look at your own "skull-vat" a little differently today.

Lena: It really does. Thank you for diving into this with me. It’s been a fascinating look at the edge of what’s possible.

Miles: Absolutely. It’s not every day you get to talk about Sudoku and interstellar travel in the same breath.

Lena: Definitely not. And to you, listening—thanks for joining us. I hope this gives you a lot to think about the next time you consider the "computer" between your ears. Until next time, take care of that processor.