Ex-Neuralink Founder: AI Enhanced Bodies Are Nearly Here w/ Max Hodak | EP #171
29:54 Share

Giving sight to the blind is a very biblical statement. Literally. Literally. Max Hodak. Max Hodak. From rewriting biology to brain-computer interfaces. Pushing the boundaries of what it means to be human. It's crazy how, like, biblical miracles eventually become tech companies.

So I love entrepreneurs who've got this incredible moonshot, but also have a real business at the same time. I think we are incredibly close to super exciting breakthroughs, but it has to be supported by something that can fund it sustainably. Let's establish how fast our brain is actually inputting, outputting information. So spoken language is about 40 bits per second.

Our attention processes the world at about 10 bits per second. Penetrating cortical electrodes like Neuralink, like BrainGate, like others have been able to get five to seven bits per second. - What is a possible future here? - All I know is that these devices are technically capable to build and I have no idea what it will feel like, but we're gonna find out. Now that's the moonshot, ladies and gentlemen. - Super pumped about what you have been building.

You know, we're going to talk about two things. One, a product you have today, which is a revenue engine. So I love entrepreneurs who've got this incredible moonshot, right? Something that's big and bold, but also have a real business at the same time on the path there. That's extraordinarily unique and critically important when you're actually building a business and you've done, you've done both. So before we jump into the BCI of it all,

Would you talk about Prima? Yeah. First of all, thank you for having me. Pleasure. Super cool to be here. And I don't want to undersell the near-term product, which is still a huge, I think, huge deal. Yeah. So we...

- Have the world's first retinal prosthesis that really works. So there's a couple of slides here. There's three layers of cells in the eye that transmit vision from light coming in to a signal going into the brain for patients that have lost the rods and cones in the back of their eyes. So their retina is intact, their brain can see, but the eyes are no longer light sensitive. - And these are typically what type of diseases? - This is specifically macular degeneration, especially age-related macular degeneration. - Anybody know anybody with macular degeneration here in the room? A lot, right? It's pretty prevalent.

And retinitis pigmentosa. My mom's dad had retinitis pigmentosa, so I grew up around blindness. So we have a chip that can be implanted in the back of the eye. So each one of these little honeycomb structures is essentially a solar cell.

So the patient puts on glasses that has a laser projector that strikes the implant in the back of the eye to excite the remaining by the cells in the retina to get the visual signal into the optic nerve at the first possible opportunity beyond the dead photoreceptors. It's a super simple one hour outpatient procedure. The surgeon makes a little blub under the retina, places the chip,

And then the patient goes home and recovers in a couple of days, can put on these glasses. And we finished a clinical trial last summer, 38 patients in Europe in that trial. There's about 50 people around the world that have had it so far. It's the first time in the history of the world, as far as we know, that these blind patients have been able to read again. And so we're super, super excited about this. Thanks.

You know, it's interesting because, you know, giving sight to the blind is a very biblical statement. Literally. Literally. Yeah.

And so I definitely, I mean, I'm very excited about the BCI technology that we have coming. I think one of the things that I've learned is the end state is often obvious, like Ray Kurzweil was like in the 90s, he was saying, oh, we would get here. The end state can be inferred. The question is, how do you get there? And the kinds of investment that it's going to take to make these technologies work. But I definitely don't want to undersell the retinal prosthesis. Let's talk about Primo for a second. Yeah.

If you were going to describe the state of the technology today,

Inhumans working, ready to sell? So I have to be very careful with what I say on that piece. It's not approved yet. But yeah, we're planning to submit for marketing approval in Europe in the next month. We're discussing with the FDA on what exactly additionally they need to see. But we're hoping to have this on market in definitely the EU and hopefully the US early next year. Amazing. Amazing. Awesome.

And that's a big product. And how big is that market, that potential total market? So there's very – it's many billions of dollars a year. There's very strong reimbursement precedents at a couple hundred thousand dollars, probably around $200,000 a patient. In the U.S., the payer that matters here is Medicare because all of these patients are over 65. In Europe, there's reimbursement precedents around $150,000 a patient.

There's tens of thousands of patients that this is directly relevant for. The whole population is probably about 250,000. If you can get to even reach two or 3,000 patients a year, this is a half a billion dollars, three quarters of a billion dollars. I love it when an entrepreneur describes their on-base single as a billion dollar opportunity. Yeah. But I mean, when you look at AI...

what are the companies that get to invest the most, most sustainably? It's the profitable tech companies. Yes. Having to go back to the well every 18 months is a huge strategic disadvantage. I think there's been a ton of capital and enthusiasm that's flown into BCI, but what...

What this base really needs is a company making $100 million or more a year. And until that happens, there will always be a risk of winter. But I think we are incredibly close to super exciting breakthroughs, but it has to be supported by something that can fund it sustainably. I think most of you know that the news media is delivering negative news to us all the time because we pay 10 times more attention to negative news than positive news.

For me, the only news worthwhile that's true and impacting humanity is the news of science and technology. And that's what I pay attention to. And every week I put out two blogs, one on AI and exponential tech and one on longevity. If this is of interest to you,

And it's available totally for free. Please join me. Subscribe at diamandus.com slash subscribe. That's diamandus.com slash subscribe. All right, let's go back to the episode. And you've acquired the manufacturing and built up the manufacturing capacity for this, yes? So we've done two acquisitions, including one as a captive MEMS fab in North Carolina. I mean, the vertical integration is essential. I absolutely received the gospel of vertical integration from my forefathers.

former co-founder and prior boss. I mean, we routinely void the warranty on million dollar fab tools to place atoms like exactly where we want them. And that being able to do that and also go from design change to surgery in a couple of weeks is absolutely enabling at being able to innovate.

Fantastic. So hold that in your mind as a company that is solving something of extraordinary difficulty where the tech is up in operational regulatory approvals coming very shortly, revenues falling very shortly thereafter. And I think that's an extraordinary accomplishment on its own. Now let's move to sort of the grand slam home run potential. I remember when

when i was talking to ray kurzweil about his predictions and again if you google his predictions he's got like an 86 accuracy if you look on uh on wikipedia and one of his predictions was high bandwidth bci by the early 2030s like 2033 and i was like ray this one i don't see it happening in that time frame it's just you're wrong about this one and then i met max and i was like okay ray you're right again

So, Max, just for a moment, I won't linger on it, but you were the co-founder and president of Neuralink. How long were you there for? About four and a half years. Four and a half years. Okay. I wouldn't say this, but, I mean, he wouldn't say this, but, you know, it's, actually, I won't just, I won't even say it. All right. You broke away and founded Science. Yes.

You had a unique idea. Yeah. Which I think is extraordinary. Describe the problem with all the current neural implants. So you've got external BCI, which is looking at EEGs. You've got something under the skull, above the dura. Then you have...

sort of wires placed in the upper parts of the neocortex and you have deep brain stimulation. All those are different types.

But let's talk about products like Neuralink and others. What's the challenge they have? Yeah, so there's many, many different ways to try to record and drive activity of neurons throughout the brain. Like neuroscience as a field has been trying to do this for the last 100, almost 150 years.

And the first thing I want to say is that BCI is a field, not a product. There are many different products that will use many different modalities for different things. But there do seem to be very serious fundamental physics limitations to the types of...

like the resolution and the accuracy that you can get with purely non-invasive devices. So then once you start thinking about putting something below the skull, the main approaches that are used today are putting just wires into the brain. The idea there is very simple. Neurons communicate, they have these electrical fields that they generate. And so if you put an electrode into the brain, you can detect this. There's other groups that are using the

they're genetically modifying neurons in the brain to make them light sensitive or make them emit light. And then there's other groups that are interested in using ultrasound. The problems with ultrasound and, and opt the optical methods, optogenetics is these really require genetically modifying neurons throughout the brain. And so doing this in an adult human is really pretty tricky. That,

That seems like a non-starter for many cases. You're irreversibly modifying these neurons in the brain of adult humans using these viral vectors, and they don't get perfectly distributed. And even then, there's still really severe limits to the depth that you can image or the resolution that you can get.

Now, the problem with placing wires into the brain, which allow you to get single neurons, is that we are used to these cartoons of neurons floating in space where you can place electrodes safely between them. But the reality is that there's no space in the brain. The brain is this wet, warm, squished thing. And no matter how small or how flexible your device is, it might look like it's a tiny fraction of a human hair floating off a finger.

every time you place one of these into the brain, you destroy thousands of cells. So that's your typical, that blue line is your typical thickness of a... Yeah, of an electrode. Some electrode. And now if you have a serious spinal cord injury...

destroying 50,000 cells in cortex to get five bits per second to recording from 500 neurons might be totally indicated, but it does mean that you can't scale up this approach to millions of cells. And that is really what you want in order to get these next generation applications. So I think about like, what is an idealized neural interface? I've been thinking about this question really since I've been in fifth grade. And can I, can I set a piece of, uh, of context for folks? I love that since I've been in fifth grade, you heard that, right? Uh, uh,

How, in terms of bits per second, in terms of baud rate, how would you describe the human brain interface on communications, on speech and so on? Let's establish how fast our brain is actually inputting, outputting information. So there's two ways to answer this question. So the figure of merit for any brain-computer interface is a bandwidth in bits per second.

And so there's this is another way to look at the different approaches. There's groups that are placing stents with electrodes into blood vessels. There's something very elegant about getting into the brain through the body's natural road system. But because of where that limits you and how far you are from the cells, those only get half a bit per second.

Penetrating cortical electrodes like Neuralink, like BrainGate, like others have been able to get five to seven bits per second. So spoken language is about 40 bits per second. Okay. So just to hear that, right? 40 bits per second is when you and I are speaking.

Right. So Neuralink is probably getting how much you think? I think that what's been published is about seven bits per second. Seven bits per second. Yeah. And that's an interesting result because if you take many different human languages, some are spoken more quickly and convey less information per token. Some are spoken more slowly and convey more information per token. But if you plot these, they all come out to about 40 bits per second. There's also a lot of neuroscience evidence that our attention processes the world at about 10 bits per second.

So the amount of information that you can perceive and remember is that there's kind of this evolved cognitive bottleneck about 10 bits per second. And so when I think about high bandwidth BCIs, I don't think in terms of communicating faster. I think that you're not going to make it so you can just convey thoughts more quickly, but it might be possible. Like getting information into the brain is very straightforward, is very easy. You can see, you can hear, you can feel. These are...

much, much more than 40 bits per second, the bandwidth of a vision. But you can't get these out of the brain. For everything that you can perceive, you can imagine, but you can't get imagery out of the brain or audio out of the brain. And that might be possible. Or we think about adding new cortical areas in the sense that even if you're still communicating at 40 bits per second in terms of number of tokens,

is can you make those much smarter? Can you have skills or knowledge or memory that you can get into the brain? - Like a Chinese character versus a letter of an alphabet. - Something, yeah. - I just want you to get those numbers because maybe you think you communicate in megabits or gigabits like your computer does. We're at 40 bits. - 40 bits.

So when I think about like, what is the idealized brain computer interface, the one that would really solve a lot of these problems. The thing I think of, if you've seen the avatar movies is this thing. Yeah.

It's a big externalized cranial nerve. So all of the information that flows in or out of the brain goes through a relatively small number of wires. There are 12 cranial nerves. The optic nerve is nerve two. The vestibulocochlear nerve that carries hearing and balance is nerve eight. Then you've got 31 spinal nerves that connect out to the muscles. You're bringing back memories from medical school. Yeah.

And so when we think about our retinal prosthesis, what we really see is a nerve two interface. But the question is, could you grow a 13th cranial nerve that has kind of interhemispheric class bandwidth, like the bandwidth that connects the two hemispheres, this fiber bundle called the corpus callosum? Could you have a branch of that that comes out and gives you a USB cable at the end? And so-

This was an idea that I had back in college, but really was beyond the field, like all of our collective ability to build at the time. But the idea that we had is what if we, instead of placing something into the brain, we load an electronic device with heavily engineered neurons, stem cell derived neurons, embed them in a hydrogel so that the cells don't go anywhere. And then engraft the wet side of this into the brain.

So there's no sterile injury to the brain. We don't place any electrical or mechanical parts. The only thing that penetrates into the brain are the biological processes of these graft cells. But at the far end, you get chemical synapses.

and so we can activate these cells optically to fire them selectively they grow both accents and dendrites so we can get input and output and we can record from them elect electrically so the optical stem and electrical record allows us to avoid crosstalk so we can drive all of them at once and this is a cool device because you can easily make

a hundred thousand electrode device because you're much closer to the cells so you can have much tighter electrode pitch you can load that with a million neurons and when that grows in you'll get a billion synapses throughout huge areas of cortex really critically important here so if you look at brain gate neural link how many total electrodes are they placing so brain gate places 100 neural link so far has placed a thousand okay we're talking about a hundred thousand or millions

And the other thing that's beautiful is that these neural growths, these axons and dendrites, when they grow into the brain, because they're native to the brain, they're not disrupting the tissue. They're pushing it aside. Yeah. So if you were to do this for real, you would see an image that looks like this.

This is a mouse brain. You can see at the top, there's a bolus of cells where the device was removed for sectioning. So this is what you did. Yeah, yeah. This is us. It's functional in a mouse. And so the graft cells that we've added are labeled in green. The host neurons of the mouse are in blue. And the thing, if we look carefully, you can see all these little green dots kind of really throughout it.

And so what we've seen is that when we engraft these devices, they grow in and they wire up very promiscuously. They kind of form connections everywhere. And then after about four to six weeks, they start undergoing an activity dependent pruning. And so the really interesting possibility there is that how they wire up is not necessarily genetically defined. It can be informationally defined based on the types of activity that you're getting in the device.

And then in addition to kind of growing down and wiring up throughout cortex, the first layer on the surface of the brain, cortical layer one is a white matter track. It's these long projection. It's like this highway between different areas of the brain. And we often see in the devices that we'll get a fiber bundle that will follow that for millimeters. The mouse brain is very small, but we see these things project all the way through to subcortical structures.

- And where these neurons and dendrites grow and they wire and then they connect and where they don't, they die off. - Yeah, they'll retract. I mean, so the cells mostly don't die, but they'll retract the axon growth cones and the dendritic arbors, yeah. - Everybody, I hope you're enjoying this episode. You know, earlier this year, I was joined on stage at the 2025 Abundance Summit by a rockstar group of entrepreneurs, CEOs, investors,

focused on the vision and future for AGI, humanoid robotics, longevity, blockchain, basically the next trillion dollar opportunities. If you weren't at the Abundance Summit, it's not too late. You can watch the entire Abundance Summit online by going to exponentialmastery.com. That's exponentialmastery.com. So this is, we have these, this is the kind of looking at one of the chips. There's these cells loaded in these trenches. And

Here, this is a Z stack. So each frame is starting at the surface and it's looking deeper and deeper into the brain. You can see the circles of the cell bodies on the surface. All of the green that we see are the graft neurons. But you can see the shadows of the blood vessels in these later layers.

And so this is, this is super cool because when you place an electrode into the brain, you always get bleeding. And if you hit a descending blood vessel, you could stroke out a whole mini column here. These grow in around the blood vessels, the capillaries remodel around it. And so this is like a really, it's a perfectly biocompatible way to get, get chemical synapses. And we see these things even where it looks like it's, it's fallen off. You, uh,

you see the processes of these cells growing in. - You know, the theme of this event, this year's summit is convergence. What technologies have to converge here to make this possible?

A lot of this was enabled by recent advances in cell engineering. So one of the things that we have to do is hide the graft cells from the immune system. We want to be we do a lot of editing to these cells. If we were to do this on a per patient basis because the immune system would have to recognize them, this would take over a year and be prohibitively cost like would be prohibitively expensive.

There's been a lot of advancement recently making what we call hypo immunogenic stem cells, the whole CRISPR toolbox and a lot of other there's technologies now like small molecule trigger kill switches. So we can make it that if you take a vitamin, the graft cells will die so you can keep an eye on them. Once you've hidden them from the immune system, you kind of want that built in.

When these go into humans, these will be by far the most heavily edited cell therapy to reach people. And I assume material sciences. And then materials. Yeah. Materials like silicon carbide, copper.

Big improvements on the materials. There's we, we talk in the BCI field, we talk about the smartphone dividend. We rely heavily on the same tech stack that smartphones and wearables build on, but Apple and Samsung and others have poured over a hundred billion dollars onto that. Our field is too small to afford that today, but we get to build on that. And that has really been enabling and has really advanced a lot in the last few years. All right. Talk to me about where, when this enters primates and potentially humans.

And when can I get mine? Yeah. So we currently have some primates getting trained up on behavior. So you're training them in advance? Yeah, we're training them in advance. We also need to figure out things like how well can they reason, which actually hasn't been that well studied in many cases. We're hoping to do the first primate engraftments in a few months later. So going into primates in a couple of months, which is amazing.

right i mean once you're in operation once you're operational in primates you're effectively other than regulatory prohibits prohibitions functional and yeah we'll be able to prove the the neuroscience that is like the the big questions for humans and primates and the first humans to get this will almost certainly be for stroke yeah um and if you'd asked me this three weeks ago i would have said i thought that it would be four or five years before the first human will get it i actually think this is not going to be much faster i think there's we uh

there might be a collaboration that allows us to go to humans a lot faster than I'd realized. And again, that'll be almost certainly for stroke. And the primates are actually a pretty good model of human stroke patients because a human that's lost a cortical area can be modeled by a monkey that never had it in the first place. And if you can restore that capability, then there's an argument that you'll be able to do it in humans. And the other thing I'll say is from the beyond stroke rehabilitation or adding these capabilities to humans,

When we think about scaling this up, I see this as a way to redraw the borders around the brain. Your head has two hemispheres. These are connected by a fiber bundle called the corpus callosum that gives you the experience of one agent in the head. But really, you've got two subbrains that are mostly independent. And people have, like a long time ago, they don't do this so much anymore. They used to cut the corpus callosum, this connection in epilepsy patients.

Um, to prevent a seizure from spreading across the hemispheres. And so, you know, that if you cut that, you really get something that looks like two agents in one head. And there's a natural example of going the other way. Um, there's a pair of twins in Canada, the Hogan twins that have one head with four hemispheres and they can share meaningful elements of their consciousness. And there's elements of task transfer between them. And so I think a way to conceptualize this is imagine if this was a tech product.

And that might be coming a lot sooner. So like we know this must be possible because nature has done it. And hoping to have this, I mean, it'll be in humans hopefully pretty soon. Which is pretty extraordinary. Let's give it up for that, huh? Max, I want to dive a little bit further about what will this mean? So this becomes enabled.

other than me being able to think in Google or sort of watch a 4K video with my eyes closed. What does this actually mean in terms of increasing intelligence connecting to AI? What is a possible future here? Yeah, I mean, for a lot of my life, I always felt like I could see the future. And I've got this event horizon somewhere between 2030 and 2035 now that is just impossible to see past. The

AGI and ASI are definitely happening. And I think that this is, I mean, everybody knows about it now, but it is still, there's basically no way to overrate the impact of that. And this is the...

Is the merge the only way through? I don't know. That was a conversation we had last year. Do we need to couple? By the way, I should have said, please get your questions ready. I'm going to be bringing Max still on stage and bringing Mo out. So we're going to be going to your questions next. So please dive in here. But if we do merge, you've talked about the idea of pre-training in some ways of these biohybrids. Can you speak to that a little bit?

Yeah, I mean, I think neural interconnects, like brain-to-brain connections, are a really interesting technology, both for merging with AI. Now, I think transformers are a pretty good explanation for cortex. But to get real agency in a way that is interesting or dangerous, you need something else to add on to that. And so people have these loops that are prompted. But that is still coming from the human.

And so it might be that the agency remains with the humans, but these technologies are so powerful, so adaptive that people who participate in this have a huge advantage. And this is a thing that societies need to need to think about. I also see it as a longevity technology.

If you can get, I think, like, how do you let someone into your head is a tricky question. I think that the first use case for this would probably be things like long married couples where one has a terminal disease. For the last year, you can get like a brain to brain. So rather than merging with AI, it's merging with your spouse or a close family member. Talk about a level of intimacy. Yeah. And can you turn...

And there, throughout your life, small groups of neurons are constantly dying. There's a smaller number that are being generated, but this is turning over. And so can you turn the death experience into basically a stroke that you recover from?

All communication is about creating correlations between brains, long relationships already store memories in each other's brains. Is there a threshold where you can get phenomenal binding across the interface where you really get one agent out? And then when you lose some group of neurons, that's not, you still get continuity of consciousness and continuity of experience through that transition. That

i see that as an alternative path than the like biological longevity companies but feels a lot more like an engineering problem to me um i think it'll be possible on the on the time scale of the next decade and in that view it's you can merge with other people you could merge with ai or you could have these super organisms that are composites of big groups yeah i call them a meta intelligence when we're able to connect millions of people's our thoughts our feelings and a level of intimacy and connection i mean you are a collection of 40 trillion cells

that you don't think of yourself as 40 trillion cells. You think of yourself as you. Imagine if millions or billions are connected through the cloud together and you become conscious on yet another level. Yeah, I mean, the really interesting question here where we're still missing some physics, but I'm increasingly confident we're going to get this in the next five years is,

What is the point where you go from having two conscious experiences into a single experience? Or do you keep multiple attentional windows? I mean, this is tough to talk about without sounding like a lunatic. All I know is that these devices are technically capable to build, and I have no idea what it will feel like, but we're going to find out.

All right. On that note, let's give it up for Max Hodak. Everybody, thanks for listening to Moonshots. You know, this is the content I love sharing with the world. Every week I put out two blogs, a lot of it from the content here, but these are my personal journals, the things that I'm learning, the conversations I'm having about AI, about longevity, about the important technology transforming all of our worlds.

If you're interested, again, please join me and subscribe at dmadness.com slash subscribe. That's dmadness.com slash subscribe. See you next week on Moonshots.

This episode is brought to you by LifeLock. Not everyone is careful with your personal information, which might explain why there's a victim of identity theft every five seconds in the U.S. Fortunately, there's LifeLock. LifeLock monitors hundreds of millions of data points a second for threats to your identity. If your identity is stolen, a U.S.-based restoration specialist will fix it, guaranteed, or your money back. Save up to 40% your first year by visiting LifeLock.com slash podcast. Terms apply.