This Week in Startups - Brain-computer interfaces and the future of neural engineering with Dr. Benjamin Rapoport | E1682

Episode Date: February 21, 2023

Precision Neuroscience’s Co-Founder and Chief Science Officer joins Molly to discuss how the brain communicates with the body (7:08) and the breakthroughs that allowed companies like Precision to exist. (12:52) Then, they discuss Precision’s founding principles, the process of getting FDA approval for its neural interface, and much more. (28:55) (0:00) Molly kicks off the show (1:27) Dr. Rapoport’s origin story (7:08) How the brain communicates with the body (11:22) Squarespace - Use offer code TWIST to save 10% off your first purchase of a website or domain at https://Squarespace.com/TWIST (12:52) Breakthroughs in neuroscience (22:37) Contra - a commission-free marketplace for freelancers and independent creators. Get $500 off your first hire at https://contra.com/twist  (24:06) The tension between public and private institutions (28:55) The founding principles of Precision Neuroscience (33:21) The Layer 7 interface and preparing for FDA approval (40:39) LMNT - Get a free sample pack with any purchase at https://DrinkLMNT.com/TWIST (42:07) Minimizing risk (45:25) Interfacing with different areas of the brain (47:06) Medical infrastructure and the business model of med-tech CHECK OUT Precision Neuroscience: https://precisionneuro.io FOLLOW Jason: https://linktr.ee/calacanis FOLLOW Molly: https://twitter.com/mollywood

All right, everybody, today we have a special and fascinating interview. Get ready to talk about our brains. Joining me is the co-founder and chief science officer of precision neuroscience, Dr. Benjamin Rappaport. We have a fascinating conversation about how our brains communicate, the scientific breakthroughs that led to the founding of Neurrelink, where he started doing this kind of work. And then what led him to start a whole new company doing a different kind of brain interface? Precision Non-Invasive Layer 7 interface. It's basically we live in the future as an interview about science that could have a profound

impact on our lives. It's going to be a fascinating show. Stick with us. This week in Startups is brought to you by Squarespace. Turn your idea into a new website. Go to Squarespace.com slash twist for a free trial. When you're ready to launch, use offer code Twist again to save 10% on. off your first purchase of a website or domain.

Contra is a commission-free marketplace for freelancers and independent creators. Get $500 off your first hire at contra.com slash twist. And Element is a tasty electrolyte drink mix with everything you need and nothing you don't. That means lots of salt and no sugar. Get a free sample pack with any purchase at drinklmn.com slash twist. Dr. Benjamin Rappaport is with precision neuroscience, and I'm going to let him explain what they're working on, but by way of background, precision has raised $53 million to date to come up with minimally invasive neurosurgical implants, right? I'm going to let you take it from there. Welcome to the show. Thanks, Molly. Thanks for having me. It's a pleasure to be here.

So please tell me what you're working on with these brain implants. Well, we're working on brain computer interfaces is the general term for the technology that we're building. And yes, they are a form of brain implant. And they're designed to connect the brain directly to computer systems as ways of helping to treat some forms of neurologic disorder that are currently basically untreatable. And those include things like certain forms of paralysis, stroke, traumatic brain injury. forms of disorder in which the brain can think but the body can't act. And brain computer interfaces are designed to enable a direct communication between the brain and a computer bypassing the part of the body that isn't able to act in order to reconnect the brain to the digital world.

Right. How common are these disorders? Well, there are definitely millions of patients, millions of people in the United States alone living with some form of paralysis from spinal cord injury or other other disorders. Okay. They're pretty, almost everybody knows somebody. Right. I, this is, I do not mean this in any way to sounds insensitive, but I just went through

a version of this with my dog, where he was losing the use of his back legs. And it was simply, you know, the vet was like, his brain is not telling his legs to work. And it was super terrible. I want to kind of go through the history here because you're at precision now. You were at Neurrelink before that. But I want to go all the way back to. your background and what, like, what is the origin story here? What got you interested in this? And more importantly, what was the moment when you realized that this could be possible?

Well, that's a great way to start. And, you know, no one ever really begins something completely de novo, right? And I come from a family of doctors and engineers. And in a way, I guess I've been working on on this my entire life. My dad is a neurologist who specializes in electrophysiology, which is the the electrical aspects of the way the brain and nervous system work. So, and my grandfather was an electrical engineer, a radio operator in the second World War. Actually, my father trained to be an electrical engineer and was exposed to the very earliest forms of artificial intelligence. And in a way, that was how he made the transition to becoming a doctor.

So I grew up with electrophysiology and clinical neuroscience as part of the everyday. and by the time I was about 20, finishing college, the most interesting thing in the world to me was what was just becoming possible or just seeming to become possible at that time in the late 90s, early 2000s, which was the notion that even though for a long time, it was possible to interface electrically from a scientific and clinical perspective with the nervous system. And in fact, all through the 20th century,

research neuroscientists and doctors had been using the electrical properties of the brain and nerves to diagnose and treat disease and to study the nervous system. The electrical nature of the brain and nervous system is kind of what makes it special in the human body. But it was not possible to do that in a kind of high bandwidth way until the very end of the 20th century. And what I mean by that is that you could maybe record from a small number of nerves or a small number of nerve cells at a time using specialty hardware until the very end of the 20th century, it became possible. All of a sudden, through some breakfuls, that maybe we'll talk about later, to record

from many, many neurons at a time. And that change in the bandwidth of our ability to interface with the brain and nervous system made the current generation of brain computer interfaces possible. And I saw that happening. And to me, it seemed incredible. And I basically spent the rest of my life, had spent the rest of my life, had spent the rest of my life working in that in that space and early on very early on scientists neuroscientists understood that it might be possible to restore function to paralyzed patients

amputees spinal cord injury patients even in some cases blind patients and so the promise of the technology has been around for for quite some time maybe 20 years now but it wasn't until about the late 20 teens that there was kind of a general consensus that it was ready to emerge from academia into the real tech world to really translate what had been proven in academic settings into clinical reality. But that was something that I had wanted to do for basically a long, long time, I and many others, and that's what we're set to do at precision. That's amazing.

So you were in the science fiction part of it, where you imagined a future future. that could be possible and then that has become true within your lifetime, which is amazing. I guess you could say that. Yeah. I mean, in a way, that's how a lot of good science gets done, right? The science fiction of the prior generation inspires the next generation to try to turn science fiction into fact. Start with, if you wouldn't mind actually, give us a primer on the electrical nature of the brain. For people who may not be familiar, you know, you mentioned, this is the thing that makes the brain in the nervous system special. If you wouldn't mind, just give us that kind of like 100 level why that is the case and why there became this idea

that you could potentially tap into that in some way. Sure. Well, neurons, which are the cell in the brain that are responsible for conscious thought and communication and many of the functions of the brain and body that we think of as making us human, neurons, communications, communications, communicate with one another using electrical impulses. And those electrical impulses, neurons are tiny. They're about, if you put them side to side, the body of a neuron, you might be able to put 20 in the space of a millimeter. So they're very small.

And the electrical signals that they produce are also tiny. but if you think about it, if you make an analogy to kind of sound, I think sometimes that's a easy way to think about it. Instead of thinking about electricity, think about sound, and think about the ways of interfacing with neurons and the brain kind of like listening to the brain

instead of electrical, just because I think that's an easier way to associate with what's going on. So if we talk about in terms of listening, electrically interfacing with the brain is kind of like, in some ways listening to or speaking to the brain. And intervasing with a neuron or with groups of neurons is kind of like building a tiny little microphone that we bring just up close to these tiny little neurons or that we place within a group of neurons.

And we try to listen to their chatter. We listen to the way they speak to one another. And there are distinct patterns of electrical activity that we can hear and make sense of. And we call that decoding. So the electrical chatter of single neurons and groups of neurons is a language. Every brain speaks a little bit differently. The problem of learning how to interpret what the electrical signals are in the brain, what they mean is a little bit different from individual to individual.

And so that is, that's kind of where the artificial intelligence angle comes from. And that was part of the technological change that occurred in the field of brain computer interfaces. As we moved from the 90s into the 2000s and the 20s, the ability to the material science that allowed us to record from many neurons or large groups of neurons at once coincided with increases in computational power and sophistication that allowed us to decode what those conversations among neurons meant. And so that in a way, that confluence of technological paradigm shifts is what has given, is what has given rise to a very powerful new technology, which is the brain computer interface. That's fascinating. So you needed this combination of the ability to manufacture an implant that is smaller than a human hair, as I understand it, combined with the ability to have adaptive compute, basically, on the other end.

to say, okay, I'm taking these signals, I can decode them, and I can put, it's not a brute force problem. You can't apply the same, you know, zap to every issue. That's right. That's right. So neural decoding is essential to bring computer interfaces. And so it's not like you can implant a tiny little microphone and give the system a dictionary and say, translate.

Everyone speaks with a little bit, everyone's brain speaks with a slightly different language or a slightly different accent, and the system needs to be programmed to adapt to that. And so that's sort of a fundamentally different paradigm of medical implant that we're talking about designing and has ever really been designed before. Hey, everybody, we're back with another show us your space contest in partnership with our friends at Squarespace. We did this last year.

It was a huge hit. Here's how it works. We're going to give one twist listener $1,000 in Squarespace, Space credits, but we're doing it vertical-specific this time. If you run any kind of an e-commerce related business, it could be a DDC brand, a consumer marketplace, a consumer subscription service, online course, you get the idea. Head to Show Us YourSpace.com. That's it. And that's going to redirect you to one of my tweets from Matt Jason. Reply to the tweet with a short video, an image, a link, a gift, whatever, that shows off your e-commerce site on Squarespace. Then the team is going to pick a winner,

and we're going to give them $1,000 in a Squarespace gift card. That's right. If you want to be an entrepreneur, you want to start a side project, the hustle, Squarespace is how you do it. On Squarespace, you can build or sell anything. We love it here at launch. We use it for a remote demo day, countless other projects, and the features are amazing. They've got templates, analytics, inventory management APIs, everything. And it's optimized for mobile. It's going to look great on an iPhone and Android phone. Everything just looks perfect. And you can even sell courses directly. inside of Squarespace and keep the 15% that other platforms are taken. Listen to your money. Keep it. Here's your call to action. It's so simple. Head to Squarespace.com slash twist to start your free trial.

And when you're ready to launch, use the offer code twist to save 10% off your first purchase of a website or domain. So now let's talk about thank you for that. That was an incredible primer. Also, you're a total poet. The analogies are perfect. So then talk to me about the process of forming, of, of forming, companies around this. So you started by founding or co-founding or partly founding Neurlink around this technology. Tell me about sort of the history there, how you became involved with NeuroL. I was one of the, one of the eight co-founding team members at Neurling back in 2016, 2017. Okay. And then explain the talk to me about the innovations that were being worked on there. Yeah. I mean, what I'll what I'll say is that that

I kind of mentioned historically the way the way the field has developed. And I can maybe take things back in history a little bit further just to give some context. So, you know, the origin origin in a way, which just to give some deep perspective is, you know, the 20th century was when biology really discovered the electrical nature of communication. in the nervous system, and that's what we were talking about earlier. And the way that scientists and doctors interrogated the brain and nervous system was with electrodes. And an electrode is just a device. Sometimes it is a wire, sometimes it is another kind of conductive material that allows you to either touch or come in close proximity with the part of the nervous system. Could be the brain, could be a peripheral nerve, could be the spinal cord.

that's generating electrical signals. And up until the late 20th century, there was no real standardized manufacturing process for manufacturing those electrodes. And neither was there a really standardized way of processing the signals. And so it's not like audio engineering, there was a whole industry

that standardizes the manufacturer of microphones and equipment for amplifying and processing sound and filtering it and recording it. There are standards and known equipment that you can buy. That didn't really exist very much until the late 20th century. And what did exist was relatively large scale devices. And let's say on the scale of fractions of a millimeter, okay, so that's large in neuroscience. Lots of human hairs.

A ponytail. So when I say large, it's still relative. And then in the late 80s, early 90s, what I think was, really a turning point for neuroscience and for what became the field of brain computer interfaces was the development of a device called the what's now called the Utah electrode array. And that was a microarray of tiny little electrodes spaced at a fraction of a millimeter from a part, approximately 100 of them. And they were made using the same manufacturing process, the same microfabrication process that is used to manufacture microchips.

So you could make this electrode array of 96 or 100 electrodes. You could make many of them. They would all be exactly the same. And you could give them to researchers to use. And everyone would be recording using a standardized microfabricated device. And looking back on it, I think that was the moment when Moore's Law arrived in neuroscience. Wow. Okay.

And remind us, I'm sorry, what year this was? The late 80s, early 90s. Okay. Was when Richard Norman developed the Utah Electric Array. Got it. So then you had the chip. And, well, that was the beginning. You know, it doesn't happen all of a sudden like that.

But you went from artisanal manufacture of electrodes to a standardized, microfabricated manufacturing technique, and one that allowed high-performance microelectronics to interface with the electrodes themselves. And, you know, we are all familiar with the, you know, the kind of concept of Moore's law, the notion that some scaling property could be applied to the technology. And the scalability of microelectronics is what is one aspect of what has powered the, you know, the revolution in computing that has, that began in the last century and continues today. That scaling paradigm is essential and the ability to connect the electronics to the end

of vector was essential. So that only arrived in neuroscience and basically let's say 1990. And that paradigm was pushed into the early 2000s. Remember that in high performance computing as we think about it today, the kind of things that enable modern artificial intelligence applications that didn't exist until the mid-20 teens. So things that we kind of take for granted, high performance computing applications that we take for granted, even things like image processing, you know, did not exist at that time.

Ironically, it also required sort of a new chip architecture, right? The shift from CPUs to GPUs. That's true, correct. So GPUs did not exist at that time. So the early interfacing of software with the new generation of microelectrodes, was all CPU-based computing, which worked just fine for tens of electrodes, and it was kind of strained at the hundreds of electrodes level. But now, you know, in the 20-teens, computing was able to catch up.

And concurrent with all of that, you know, there was a generation, I include myself in this, of, you know, engineers in training, masters and doctoral level engineers who kind of cut our teeth on circuit architecture, algorithm design for how to interface with the signals coming off of these sorts of electrodes. And by the early 20 teens, I think there was a general consensus that most of the major science problems, or many of the major science problems, even some of the engineering problems, had basically been solved. In other words, how to build back-end electronics and how to encode software,

that would make sense of many simultaneous electrical signals coming out of the brain in a way that we could understand what the brain was trying to tell the arm or the leg or the mouth to speak. And at that point, there was a sense that to take the next step to really transition academic science into clinical reality that would benefit people, it was time to move into a commercial and industrial setting

out of the lab. And I want to give you one other piece of context, which is that, you know, think about the historical backdrop of all this work as it was happening in the United States. You know, from the early 2000s, 20 teens, we were seeing a lot of, you know, wounded young people coming back from Iraq and Afghanistan. And the National Science Foundation and DARPA and other funding agencies had a strong mandate to try to do whatever was technologically possible to take care of these motivated young people who had been serving our country. And so that led to a tremendous amount of attention being paid to advanced prosthetics development.

Not all of that was neuroprostetics. The work took many forms. But certainly that motivation catalyzed a lot of tremendously important work. and so the government funding agencies invested heavily in the development of this of this technology. But at some point, it became clear that that form of investment and the timescales that were required to secure the government grants and do the work in an academic setting, the funding wasn't enough and the time scales were too long. and in order to really translate the advances that have been developed in academia into clinical reality, it was time to move into a commercial setting. And so in 2016, a couple of major efforts shifted attention from the academic to the commercial setting.

Neurrelink was one of them. Facebook had an initiative around brain computer interfaces. A company called Kernel also was started around. that time, each of these entities put a huge, what at the time seemed like a huge amount of capital behind moving people and resources from academia into a more commercial research and development setting to build the brain computer interfaces that would actually go into the clinic. And that has given rise to a small ecosystem of advanced startup companies and a tremendous amount of talent being brought to bear in the field. And I would say that that's the

best thing that happened out of Neurlink is that engineering talent has been really focused on what I think is a tremendously important problem for our generation. And the more talent we bring into the field, the better. So it was sort of a beacon. And it attracted all of this best and brightest. And you could sort of go off and found related companies. Hiring freelancers and doing that on project-based work is a brilliant way for you to grow your startup sustainably, right? You can't just hire everybody in every little vertical. And listen, there is a cunt of top talent right now out there looking for work. Do it all the layoffs in tech, you know that.

So you need to check out Contra, C-O-N-T-R-A. Contra is a commission-free marketplace for freelance and independent creators. So all that money that's going back and forth between you and your freelancers, it's not taken by some marketplace. No. There's no percentage-based upcharge when you do hire somebody. And they do all the vetting. They find the best people on the other side of the marketplace.

Hey, if you're one of these laid-off tech workers and you've got tons of skill, sign up for Contra. It's an amazing platform for you. And remember, like I said above, creators on Contra keep 100% of what they make. There's no fees. They specialize in design, engineering, social media, video, writing, and of course AI. This is a really easy way for you to get great talent and to do it quickly if you need project-based work. You need to check out Contra. It's that easy. And you know what? The best thing about freelancers is you

only spend what you need to spend. You might have a really important social media project, but it's only for six months of the year. Or you need some videos, but you only need 10 of them, not a hundred of them. They're going to do it fast. They're going to do it right. So here's your call to action. I can't believe it. $500 off your first hire at contra.com slash twist. That's right. Five crisp hundies waiting for you at Contra. C-O-N-T-R-A.com slash twist. I wonder, talk to me a little bit more about that. I think we're all familiar with the kind of academic valley of death that can occur with R&D, where it doesn't become commercialized or there is just not quite enough investment, right?

It's like you can get here, but you can never make the final leave. And yet, there is also that question about what happens if you attract all of the best and brightest researchers and scientists to private industry, that if that doesn't work, then you've left universities unable to continue this research too. Like how do you tackle that tension behind doing this as private companies that have, you know, investment return expectations as a result, you know, because of their VCs, versus having this happen kind of in an academic way that might be more open? Yeah, it's a great question.

And there is that tension always. There always will be that tension. and it's hard to solve the general case, but I can speak to the specific case of how that sort of dynamic has played out in our field. And maybe to give a little bit of perspective, I like to think about what's happening now in neural interfaces as somewhat similar to the genomics revolution of the early 2000s. And if you sort of think about it, you know, actually in the,

the year 2000, it was not common for a computer scientist to be working in biology. True. Right. That was kind of a new thing. We don't think about it so much nowadays, but it was not really a standard thing to have computer scientists working in biology, but the human genome project gave them jobs in biology. There was no such thing really as computational genomics, you know, in the 80s. Biology wasn't really a platform. It wasn't enough data.

There was not kind of an. infrastructure. And then, you know, what happened was that the human genome project in the final stages became kind of a competition between, you know, industry and an academic consortium. And it doesn't really matter so much who won. It just matters that that competition made it clear that it was time for high throughput gene sequencing to take its place in industry. as well as in academia. And so it gave people who had been working in a purely academic sense an opportunity, some of them, to move and to take care of the engineering and scaling challenges required

to actually bring what was essentially an academic endeavor until then to patients, doctors, and healthcare systems. And that itself has been challenging in all kinds of ways. But that transition has been made by, you know, quite a few companies that have gone on to be tremendously successful to generate jobs and have economic impact that has been far greater than the total amount of federal investment that's gone into the field. And most importantly, you know, it has had a tremendous impact on medicine. You know, to the point where now you can, you know, you've had, you've had some of these people on your podcast and, you know, in the past,

But, you know, you can, as a consumer, you can have your genome sequence. You can have all kinds of insight into your family and past and future health, as well as, you know, patients undergoing advanced medical therapy can have not just their own genome sequence, but, you know, the genome of a tumor, for example. Right. So I mentioned that as background because I think that something very similar is happening in neural interfaces today. That to me, I think the year in 2016 was for.

neural interfaces, kind of like the year 2000 for genomics, in that several major entities were formed that drew tremendous talent out of academia into industry with the mandate to try to take academic science and bring it to patients. And, you know, like you say, that transition is fraught and it's challenging in all kinds of ways. It's not usually, usually, it's not usually well done in an academic setting. It kind of needs, it kind of needs professional engineering and oftentimes a profit motive to really, to really develop a robust engineered system that meets all the quality control standards and regulatory standards that allow it to be patient-facing. Right. So then on our sort of journey here from 2016, say, to today, what, if you don't mind my

asking caused you to leave Neurrelinkin co-found a new competitor, if you will. I don't know if it's a competitor directly or not, but what's happening at precision and what made you want to go do that? Yeah, I mean, I think that as was the case in, as was the case in genomics and high throughput gene sequencing, we've seen that there was tremendous opportunity there and a number of very successful high-impact companies emerged, all attacking aspects of that scientific endeavor from different ways. And so maybe some of them are competitors, but nevertheless, they've been able to have tremendous impact side by side.

And I sort of see something similar happening in neural interfaces today. There are, there's no one-size-fits-all neural interface. There is, I think, a consensus that the ability to interface with many, many neurons or to interface with the brain and nervous system in a very high bandwidth manner is critical. And that's the general trend. That's what modern brain computer interfaces are. They are high bandwidth connections between the brain and the digital world. But there are different ways of doing that.

So I guess that's the question. Is precision neuroscience solving a different problem than neuralink was tackling? It's solving it in a different way. And that way is going to have some different applications. So one of the things that I have, let's put it this way, some of the founding principles of precision are a little bit different from what others are doing in the field today.

And we feel that in order for neural interfaces to really have a major impact clinically in patients, we have to be able to reach many patients and to do it very safely in a way that poses minimal risk to patients for maximal benefit, and in a way that is extremely scalable. So the performance of a brain-computer interface is very much dependent on the bandwidth of the interface,

which makes sense, right? I mean, we all live through this transition from dial-up to high-speed Internet and all of the changes and advances in technology that we've seen go along with that have been completely transformative. There are things that we can do with high speed that we could never even have dreamed of with early generation modems. And the same is true with neural interfaces. The scale, the bandwidth is tremendously important.

And so we have developed a platform around those principles that safety, which to us also comes with minimal invasiveness. So that means basically not damaging the brain with the interface and yet being able to deploy an interface that scales to very high bandwidth. Those are kind of the guiding principles, the guiding design principles at precision. And so that resulted in us designing an electrode interface that rather than being many tiny little penetrating electrodes

that penetrate the surface of the brain, that is the nature of the Utah electrode that we mentioned before, that is the nature of the neural ink electrode and some others. Rather, we're using tiny little electrodes. Think of it like saran wrap. So a kind of saran wrap that coats the surface of the brain with many, many tiny little electrodes, each one of which is about the size of a neuron.

And yet it doesn't penetrate the brain. So it can listen to the brain at very high resolution, can even stimulate the brain. so it can listen to and speak to the brain cells. And yet it can be removed with no damage to the brain. And it can be replaced or upgraded. That's the nature of the system. And it can be deployed in a way that doesn't require a very complex open brain surgery.

And that is not the case. That is distinct from other solutions. I'm not trying to get you too. I'm not trying to be provocative here. That is distinct from other solutions. Got it. Okay. How safe is this?

Like how far along on the road to true commercialization and widespread adoption are you? So we are, we're getting ready for FDA submission this year. Our, we have, we work, we have done a lot of work in large animals. And all the early work that we've done, all the work that we've done to date suggests an extremely good safety profile. So our goal is really to never damage the brain through the interface. And just to make it clear, that truly is a different paradigm from the brain computer interfaces of the 90s, 20 teens, all of which were most of which, many of which, let's put it this way, many of which were developed around tiny little electrodes that penetrate the brain. So think of the little microphones that we discussed before,

the little microphones that listen to groups of neurons, the way those are, think of those as little wires or little needles. And in order to, in order for them to do their listening, they need to be placed inside the brain itself. Like if this microphone in front of me was actually extending a little tendril into my vocal cord directly

so that it could hear me, I would prefer that it not do that. Exactly. You prefer that it just listened to your voice. Yes. And so, you know, with a microphone, you can have something that's completely external to your body. Right. So that's, that's, and, you know, for a long time, people have been trying to ask the question,

can a high band with neural interface be completely non-invasive? Right. Right. And there has been a lot of work that's gone into what kinds of electrical signals can we record from outside the brain completely, from outside the body, from the scalp, or, you know, something like that. And certainly there are detectable electrical signals that one can detect from outside the head completely, but they are not, that kind of system does not permit high bandwidth information exchange between the electrode and the brain.

Just the physics of the situation doesn't permit us to listen at high resolution, spatially or temporally. So you need to be really close to the brain to get the best quality information. And so what the precision system is doing is to get as close to the brain as you can without damaging. So that sounds like, it seems like what you're saying is this is a big deal. That's a really big breakthrough. Yeah. I think so. That is a completely new paradigm.

Congratulations. Thank you. Thank you. You're like, I need you to understand. This is major. Like this is a. I mean, nothing, let's just, you know, nothing comes in isolation, right? I mean, there is, we're conscious that there is, you know, decades of neuroscience and engineering

that have allowed us to, you know, to get to this point. And likewise, there is an ecosystem around us that we are interfacing with. And that is an ecosystem of, you know, high-performance manufacturer, advanced manufacturer in the United States, the whole medical device industry as it, as it exists today. The regulators, the FDA, and insurers, hospital systems, neurologists, and neurosurgeons and patients and their families. I mean, there is an entire ecosystem required to move, you know, a promising technology from the laboratory into patients' lives. And that is kind of what you were asking before, is how do you decide when to make the leap?

How do you make it work? And we're very conscious this is not tech development that happens in a bubble, especially medical technology. The environment is highly regulated and the ability to, to the ability to, to, the ability to get early feedback from our users in medical technology is very much limited relative to almost any other high-tech industry. And we're sensitive to that.

And I will say that is one thing also that is, that is unique about the precision system. And you asked before about what our next steps are and how close we are to actually be. able to deploy the system and because our electrodes don't damage the brain, they're what we call a surface microelectrodes as opposed to penetrating microelectrodes, they can be removed or they're designed to be removable without damaging the brain. And so we're designing the first system to be a temporary system, meaning it will be usable for a temporary interface early on for diagnostic use in conditions like epilepsy. And then it will be able to be removed.

moved. And that has certain advantages. One of them is that the first generation device, we hope, will do a lot of good clinically. It will certainly deliver to, we hope, will deliver to clinical practice, the ability to interface with the brain at spatial and temporal resolutions that have never been possible before. We'll be able to see, we've seen it, you know, we've seen it in development, but hopefully we'll see it very soon in patients. the real-time activity of the brain at spatial resolutions that have really never been seen before. And to us, that's tremendously exciting. We think it will have a significant impact on clinical care.

But the fact that the first-generation device is removable allows us access to a regulatory pathway called the 510K pathway, which is what the FDA calls the pathway for devices that are sufficiently similar to something that's been done before. And so the path is somewhat expedited relative to what we call the PMA pathway, which is used for class three, high risk permanent implant devices. And we're very excited about that because we feel that even as we deliver first generation benefit to clinical medicine, we also will be able to understand how the device works in the hands of clinicians and in the lives of patients. And that will be, we hope, the first, or one of the first, if not the first, you know, approved high-band with electrophysiology systems in clinical medicine today. And that,

we hope, will set the stage for the generation of perinin implants that come in the years to follow. But that will generate a lot of learnings in our, in our view. And we hope will be good for the feel. Yeah. I mean, not to oversimplify. but you'll be effectively, your first product will be both a diagnostic tool. Like it will be a test strip for the brain. In a sense, yes. Yeah. It will, it will deliver, we hope that it will deliver clinical benefit while also breaking,

kind of breaking open the space of high bandwidth electrophysiology. Yeah. It will, it will be one of, if not the first truly, extremely high bandwidth, neural interfaces available to patients. So that's what we're, that's what we're working towards in the next. 12 to 18 months. All right, everybody, I have to tell you about the most delicious electrolyte drink I've ever had. Element is a tasty electrolyte drink mix with everything you need and nothing you don't.

That means lots of salt with no sugar. It contains a science-backed electrolyte ratio. 1,000 milligrams of sodium, 200 milligrams of potassium, 60 milligrams of magnesium. An element has none of the junk, no sugar, no coloring, no artificial ingredients, no gluten, no fillers, no BS. It's formulated to help anyone with their electrolyte needs. And it's perfectly suited if you're keto, low carb, or paleo. I've been using it. I was up in the mountains in Tahoe.

I was drinking it twice a day. I felt great. I slept better. And you know what? It helps prevent and eliminate headaches, muscle cramps, fatigue, which boy, at my age, I'm feeling, and other symptoms of electrolyte deficiency. Element is used by everyone from professional athletes to Navy SEALs to everyday people. I love that chocolate salt flavor.

You know, it gets me in that chocolate space that I'm. I love so much, so sweet, so savory. Element is offering Twist listeners a free sample pack with any purchase. That's eight single serving packs free with any element order. Go try all eight flavors right now. DrinklmnT.com slash twist. Element offers no question asked refunds, so you have nothing to lose. That's drinkelement.com slash twist.

D-R-I-N-K-L-M-N-T dot com slash twist for a free sample pack with any purchase. I mean, what could we learn from that? Like, let's have the sci-fi part now. Because there is still so little we know about the brain, all things considered. And it feels like this is a huge opportunity to really get in there. Yeah. Well, I love that. I love that way of thinking about it.

At the same time, you know, when we think about this question that you asked before, when is it time to move from academic science into, a commercial enterprise. And so in a sense, we try to do as little new science as possible, right? We're really trying to just professionalize the science and engineering that has already been done. Which, to be clear, is already awesome. I'm not trying to jump ahead.

No, no, don't get you wrong. This is already amazing. It's a huge lift. It's tremendously exciting and I think very, very high impact. we everyone here at precision believes that. I just, you know, since we're talking about, and in a sense, you know, this, this podcast is about strategy and ways of thinking about startups

and the world of new technology. And medical device development is high risk enough. And we want to take as much of the risk out of that development process as possible, or at least to quantify it to the extent that we can and to take on the risk in as a small bite-sized chunks as we can. And so you're right. There is a lot that is unknown about the brain,

and we try to deal with the parts that we know the most about. And so when you ask, you know, what will we be able to learn? What will be able to learn from this device once it's cleared for use in human patients? And the answer to that is, I hope that we learn a ton. I hope that we, you know, are going to be delivering both a diagnostic tool and a scientific tool. and a scientific tool to the community that will teach us all kinds of things. And certainly, you know, the last generation of electrodes and electronics that were delivered to that were made available for clinical use, we have learned unbelievable things,

you know, really, really, I would say, transformative things. You know, if you think about, you know, 50 years ago what the first generation of electrodes was able to teach us. It taught us about where language was located in the brain in great detail. and all kinds of detailed things. And I think that we'll learn a ton when we can, basically what we're providing is a tool that will provide sub-millimeter resolution electrical information from the brain.

So it will provide a window that will allow us to see the active brain in real time kind of at a microscopic scale. So think about, you know, looking at the brain under a microscope, but actually being able to see how the brain is computing, not just being able to see what the brain looks like. At the same time, you know, we, when we think about the clinical applications that we're designing around, we try to design around those areas of brain physiology that we feel we know a lot about. So, you know, one of the, one of the patient populations that we're designing for are patients with, various forms of paralysis, and that includes paralysis of the limbs, as well as paralysis of the

articulatory muscles of speech. So in a sense, aphasia and certain forms of inability to speak are also forms of paralysis to the extent that it's the articulatory apparatus, the mouth, the tongue, the pharyngeal muscles and so on that cannot move. And all of those muscle systems have spatial representations in the brain. And we basically know where they are, but being able to interface with them at the scale that we're talking about, we hope will enable unlock functionality for patients with those kinds of disorders that has not been possible with lower resolution electrode systems.

So, for example, understanding the detailed structure of the articulatory muscles of speech or the fingers and where those lie on what we call the motor cortex of the brain, only electrodes of the scale that we're talking about can really interface with the neural structures that give rise to function in those subtle aspects of the body. We kind of know where they live, but we need to, we need sensors and computation that are on the right scale to interface with them. So that, to me, those are some of the more exciting applications that we're looking to develop in the years ahead. That's amazing.

And then one last question, I promise I will let you go, because we also talk about investment on this show. Talk to me about cost and business model. The manufacturing process you've described, you know, I mean, I know from just semiconductor manufacturing, it's a clean room situation, unbelievably expensive boundaries. How, what does it cost to produce one of these layer seven devices and what will? it cost on the other side of things? That is a great set of questions, and it's not a seven-minute discussion. I'm not going to answer exactly. So please invite me back, you know, for...

I would love to. But no, it's a great question. It's absolutely critical, and I'll try to address it in a couple of minutes. But that question gets at, you know, some deep aspects of how medical device development works in the United States and in the world. and in what medical device what the medical device industry is going to look like in the years ahead. And to date, really, there is almost no example of implantable medical technology

that really depends on microfabricated sensors and actuators. Almost all medical devices today are artisanally finished, meaning human hands are making and finished. the devices. So unlike the semiconductor industry, which has all of the highly expensive infrastructure that you mentioned, MedTech has not relied on that to date. I think that is going to change in the coming years. Certainly, the neural interfaces industry is facing that, and we are driving some aspects of that change, meaning the sensors that we're developing do require and microfabrication.

They don't require the kinds of single-digit nanometer resolution that advanced semiconductor manufacturer requires today. But nevertheless, they require similar processes. And so we are seeing a need for advanced manufacture in medical technology. And to me, that's actually very exciting. But also it does, of course, come with a number of considerations, including investment dollars for how to scale up that manufacturer. But the question of how much it costs to manufacture the device and how that cost is borne by insurers and so on is a good one.

I can say that these devices, I think, will be more expensive than the current generation of implantable neural devices like deep brain stimulators, then pacemakers, then cochlear implants, and so on. but not, you know, probably not 10x. And if you think about the kind of medical economics of what we're trying to do, it's easy to understand how it makes sense. What we're really trying to do with these devices is to enable, say, a young quadriplegic patient who may be 30 years old and have, you know, 50 years of life ahead of them and, you know, 35. of them maybe in the workplace, we want those patients to be able to have a level of independence

and dignity and financial self-sufficiency and the ability to go back to work if they want to. And if you think about the change that that is possible, we firmly believe that that is possible. We're not really that far away actually from enabling that transformation. But if you think about the meta economics there, it's not a hard case to make. you're taking somebody who right now forget about that we need to do it as a society. I mean, that's just a given, okay? But from an economic standpoint,

you're taking people who whose medical care is largely born by disability insurance and state-run programs and things like that. And in the workplace, they pay for commercial insurance. So the impact of that economically on the medical system is a completely sensible model. So even if the devices are several times more expensive than current generation devices, the technology will, the impact on the system as a whole will pay for itself. And I will also say that most likely, I mean almost certainly,

We're looking at a slightly, certainly hope, we're looking at a different model in the years ahead. There has been a shift in the medical device, medical technology industry, towards more software as a service, software as a medical device models. And a lot of the functionality, we spent a lot of time talking about the material science and the electronics and so on. And we talked a little bit about the computation involved. But a lot of the functionality over the years will come well after the implant as software upgrades are pushed to the device. And the ability to push software upgrades and enhanced functionality will continue for years and years after the initial implant. And one of the things that we say at precision is that every patient should, the data that is generated by every patient should provide some benefit to the data to the data. data to every other patient's care that comes after.

You know, every patient should be helping every patient that comes after them. And machine learning does make that possible. You know, data science today makes that possible. And so, but that will also be part of the economic model. You know, the software upgrades will not be free. And, but also, you know, people will only be paying for, for functionality that benefits them. Yeah.

So I hope that answers your question in a nutshell. Definitely. It's fascinating. and now I want to have part two. Dr. Benjamin Rappaport is co-founder and chief science officer at Precision Neuroscience. Your present really is our future. It is fascinating world you live in. Thanks for having me. Real pleasure being here. Thanks for the time.