The Benefits Of Measuring Brain Activity with Ryan Field, CTO at Kernel

(Intro Narrator at 00:00:03) Hello, my friends. Today we're talking to Ryan, CTO at Kernel, and we discuss how Kernel Flow is reading signals in the brain, what different substances such as alcohol do to our minds, and how objectively measuring brain activity can inform our subjective human experience. All of this right here, right now on the Modern CTO Podcast.

(Joel Beasley at 00:00:34) This is the Modern CTO Podcast.

(Ryan at 00:00:42) Hey, Joel.

(Joel Beasley at 00:00:43) How are you doing today?

(Ryan at 00:00:44) Good, thanks. How are you?

(Joel Beasley at 00:00:46) Well, I'm doing pretty good. There is a mouse in the studio today. You may see me jump up at some point. I've got a bucket and I'm going to catch this mouse.

(Ryan at 00:00:56) So we may have a guest appearance. Is that what you're saying?

(Joel Beasley at 00:00:59) Yeah, yeah. I told my wife, I said he's not going to come out until the cameras start rolling. He's a popular mouse. Well, alright. I'm going to give you my quick background, okay?

(Ryan at 00:01:10) Okay.

(Joel Beasley at 00:01:10) Just so you have some context about where my experience is at and all of that, and then we can learn about you, learn about Kernel, what you guys are doing with Flow. So the abbreviated story is a software developer for 17 years. So I went from individual contributor to building a team to teams of teams, and I did that for, you know, repeat, rinse and repeat for a while. And then I started writing and sharing what I learned. You know, here's the mistakes I made. Next generation, don't make these mistakes. And that turned into a blog, turned into a book, turned into the podcast. So I kind of went from software engineer to podcaster to podcast production company owner. That's sort of my trajectory. That's how I got to this call today, man.

(Ryan at 00:01:53) Cool. That's quite an evolution, I would say. And I'm guessing when you started 17 years ago software developing, this is not the endpoint you imagined.

(Joel Beasley at 00:02:03) No, I'm not a billionaire yet, so still working on it.

(Ryan at 00:02:09) Cool. Well, I can tell you a little bit about myself. So I did a PhD in electrical engineering at Columbia, and in that I was developing custom camera chips that look at not the color of light, but the temporal properties of it. So how it changes over time. And these timescales are really short, on the order of nanoseconds. And the reason this is interesting is because a lot of biological molecules have properties that can be measured optically through these very short timescale measurements. And so it's a technique called fluorescence lifetime imaging, and it's all based on time-of-flight measurements. So I had a PhD basically designing custom time-of-flight chips. And in that I kind of learned everything from the device layer, so how the PN junctions and diodes that detect the light all the way up to the software layer work. So kind of everything in between, designing the circuit, designing the boards, writing some low-level firmware, writing software. So it gave me this really broad experience to start out with and kind of a good understanding of whole systems. From there I had a short time at Intel working in their research labs on some biosensing platforms, and then went on to a startup called Quanergy where I was developing a LiDAR sensor. So doing, kind of going back to my roots and what I did my PhD on, and doing time-of-flight measurement. We send these pulses of laser light out and see how long it takes to come back, so we know the speed of light.

(Joel Beasley at 00:03:39) That's why you're measuring them. I was like, why would anybody care about how light changes on a nanosecond level?

(Ryan at 00:03:45) So in LiDAR that's kind of on a microsecond level, so it's a different timescale, but same principle. So you're pulsing laser light and looking at microsecond timescales. So I worked there for a few years, developed some detector technology, built up a team, kind of grew that detection team up quite a bit. And then I was approached by the team at Kernel and Bryan Johnson, who had this idea of measuring the brain using time-of-flight on the nanosecond timescale again. So going back to the shorter timescales that I'd done my PhD on. And so the concept is you put in a pulse of laser light and you look at how that light changes over time for this, you know, short, you know, few nanoseconds. And that tells you something about the properties of the tissue inside the head. And you can, you know, by detecting changes in those properties, tell what parts of the brain are active in a certain time. And this is kind of a well-understood phenomenon. It's been around for many decades, but it had previously existed as kind of racks of equipment in a research laboratory and never as a, you know, integrated device. And so it sounded like a challenging problem, and, you know, I had the right skills. I had the experience building the teams and kind of decided to join and start helping build up that technology and figure out what exactly to build and how to build it and what the right technology stack was and, you know, everything along the way to get to what is Kernel Flow today. So it took us about four years, a little over four years to get here, but we've got something working and we're kind of putting the finishing touches on the production release for later this year.

(Joel Beasley at 00:05:26) What can I do with it?

(Ryan at 00:05:27) Okay, it's a great question. So I'll tell you what it does fundamentally, and then we can talk about what that could tell you kind of from an applications perspective. So fundamentals, are you familiar with a pulse oximeter? So you may have a smartwatch that has an SpO2 sensor in it, or if you've gone to the doctor or hospital, they clip one on your finger. That's a pulse oximeter. It's measuring the oxygen in your blood. And so fundamentally what we do with Kernel Flow is we generate a map of blood oxygenation across the entire brain. So we're seeing small differences in oxygen in your blood throughout the brain. And what that tells you is what parts of the brain are active. So as the brain does work, or parts of the brain do work, those parts need some oxygen in order to do the metabolism to fire the neurons to produce the function. So Kernel Flow measures that change in oxygenation, and, you know, this is what's called a hemodynamic measurement. So we're looking at changes in hemoglobin oxygenation. So, you know, the hemoglobin is the thing that carries oxygen around in your bloodstream, and we're measuring how much of it has oxygen versus how much of it doesn't. And we use a technique called near-infrared spectroscopy. So we use two wavelengths of light, which is the spectroscopy component, and it allows us to differentiate the oxygenated versus deoxygenated. So high-level, we're measuring blood oxygenation as a map of your whole brain. And, you know, what you can do with that is you can start to ask questions like, what is my brain doing when I'm just sitting, kind of daydreaming, staring off into nothing? And this is what's called, you know, resting state. This is what the brain looks like when nothing is going on.

(Joel Beasley at 00:07:17) Resting brain face.

(Ryan at 00:07:18) Yeah, resting brain face. And so it's just one very simple example. Another is, I think you probably saw some of the work we've done playing video games. And so what is happening in the brain when a person is playing video games? And we have this hypothesis that, you know, we would see actually almost the entire brain light up. You're using so much of your brain playing a video game. You're using your motor cortex to control a mouse or keyboard. You're using your prefrontal cortex to plan out activities and how you're going to move or what your next point in the game is. Your visual cortex is lit up as it processes all of this visual information coming in. The parts of your brain handling auditory or processing sounds, trying to, you know, a lot of games use spatial audio-type tricks to give you signs and cues for what's happening in the world you're in. So so much of your brain is being activated by video games, and we actually saw this in some of the measurements we took is that, you know, when you play video games, tons of stuff is happening. It's not actually bad for your brain. You're actually working out the entire brain when you play these games. And those are just a couple of examples. I mean, we could probably talk all day, could you measure the brain while doing X? And the answer is probably yes. And one of the things that Kernel Flow enables is asking questions in a new way because it's now a portable device and you can kind of use it in more naturalistic environments. And we could talk a little bit if you're interested in kind of what the legacy devices, what the whole history of functional brain measurement was built on and how those don't really quite provide the same type of experience.

(Joel Beasley at 00:08:55) So you took something, made it better, portable, and you're doing an infrastructure play where you're building these electronic infrastructures that people can then build applications and things on top of?

(Ryan at 00:09:08) Yeah, I would definitely call it kind of a platform play wherein you built this, you know, the worst best analogy I have is it's like the iPhone, right? It's the hardware technology that's needed to enable so many new applications. And can I dream of all those applications today? No, right? Who knew when the iPhone launched that the most used app would be TikTok in 2022? Right? No one would have predicted that. And I would say the same thing about a device like Kernel Flow, which is really kind of a platform for building applications on the brain, on information about the brain. And I don't think any of us today can predict ten years from now what the significance of that will be or what the most used app would be in that ecosystem.

(Joel Beasley at 00:09:56) How much time do you spend on TikTok?

(Ryan at 00:09:58) Zero.

(Joel Beasley at 00:09:59) I have not been able to get into it. Yeah, I've tried multiple times. I just...

(Ryan at 00:10:04) Maybe it's an engineer thing. It's like I think I'm a little bit of a control freak. I don't know. It's like I want to control my own destiny. I don't want content served to me randomly or based on some algorithm. I wanted to use it. I'm in control here.

(Joel Beasley at 00:10:20) Yeah. Does hair gel interfere with it reading your brain? How does hairstyle, hair products, does that interfere?

(Ryan at 00:10:28) Yeah, so we use an optical technique. So as I mentioned, we're firing light into the head, and you can imagine anything that blocks light is going to block signal. And so hair is one of those things that's a little bit of a challenge for us. Hair gel, I think the only problem it would cause is that it makes the hair kind of clump together more. So you kind of have a harder time fighting through it. But what we've done is actually kind of neat. We designed these modules with spring-loaded tips on them, so they're kind of like mini brushes that you kind of put on your hair and kind of wiggle around as you get it into place so it can get down and touch the scalp. So if you had gel in your hair and a nice hairstyle when you started the day, you take a measurement with Kernel Flow, that's no longer going to be true. But, you know, then you have Kernel Flow hair, which is, I think, a mark of pride in itself.

(Joel Beasley at 00:11:19) Is there anything at your labs... Where are you located, by the way?

(Ryan at 00:11:23) We're in Los Angeles.

(Joel Beasley at 00:11:24) So if I fly out to Los Angeles, right, and put on one of your helmets... What do you call them? I'm sorry, you probably don't call them helmets. What do you call them?

(Ryan at 00:11:32) So I try and call them headsets, but helmet is a very common thing. Headset sounds a little nicer. Helmet's like I'm putting this on to protect my head, whereas headset is...

(Joel Beasley at 00:11:42) You're too deep in the space. You're too deep in the space. Because from an outside, I'm like, oh, that's a helmet. If it covered my eyes, I'd probably say headset. Does it cover your eyes?

(Ryan at 00:11:52) It does not, no. It's just there.

(Joel Beasley at 00:11:54) Like a metal hat type deal. I completely forgot my... Fly to LA.

(Ryan at 00:12:01) Headset on.

(Joel Beasley at 00:12:02) Put the headset on. What would I be able to do with it right now, today? Could I play a game and think about things and how they move characters? What could I do with it today?

(Ryan at 00:12:12) Yeah, so one of the, I think, most interesting demos that we've been working on these days is a focus training demo. So it's kind of a neurofeedback application where you try to focus on a topic or focus on a thing or let your mind wander. And so you kind of train your brain to be able to kind of modulate between these two different states. And so it's something that, I don't know if you've practiced any mindfulness or meditation-type things, but it's something that kind of fits into that space where you can get a feedback, a score based on your brain activity of how well you were focusing or how well you were mind wandering. And so you can use that score as quantified information to say I'm doing better, I'm doing worse. Let me try different techniques. Let me change what I'm doing so you could improve. And so it's a really interesting thing because it gives you immediate feedback. It is something that we measure in real time, so you don't have to wait for it to go to the cloud and get processed where a lot of our earlier kind of demonstrations and things we're doing... We still have all of that capability and use it quite heavily when we look for kind of larger scale, big data-type results. But this is kind of an individual personalized feedback system that we've been developing. And there are a lot of applications for it outside of just mindfulness, right? You can imagine all the things you could do if you could improve focus or improve, you know, some dimension of your thinking, whatever it is, whatever label you apply to it, whether it's focus, mind wandering, attention, stress.

(Joel Beasley at 00:13:47) I was curious, can you tell if your brain is tired or fatigued? Because, you know, at the end of the day when you're just burnt, can you put the helmet on and be like, am I burnt, or do I have some more in me?

(Ryan at 00:13:57) Yeah, we can't yet. We don't have an app for fatigue yet. Kind of our first app that we're building is this focus type of one. But, yeah, I think it is possible, right? And the reason I think it's possible is we've done some other things. We did a study, it's maybe about halfway through right now data collection, where we're giving people alcohol and looking at a placebo and then two different doses of alcohol up to the legal limit, 0.08, and seeing how their brain changes in response to the alcohol. And what you find is that there are some compensatory mechanisms. So you see the brain activating parts of it to overcome alcohol to an extent. So when you kind of get this lower dose, people on average show this compensatory mechanism. But then at a certain point, close to the 0.08, you actually see that that compensation fades. So your brain is no longer able to make up for the impairment caused by the alcohol. And I bring this up in the context of fatigue because it's probably something similar. It's like I'm feeling a little bit tired, I'll just push through. You can do it a little bit longer. You'll probably engage a little more of your brain, overcome some of that fatigue you're feeling. But at a certain point you're probably just like, alright, I am completely burnt. Time to call it a day. So we haven't measured it, but I could see it being something within the possibilities of what we could measure. And it kind of parallels, like I was saying, this kind of impairment-type study that we had done or in the process of doing.

(Joel Beasley at 00:15:31) Nice, nice. And then sorry, I see the mouse.

(Ryan at 00:15:35) How do you? By the...

(Joel Beasley at 00:15:37) Way, if you were in studio with me, we'd be dealing with this together. It would be quite fun. I actually had an in-person podcast yesterday, and I'm so glad the mouse wasn't here.

(Ryan at 00:15:46) Have you ever put this on mice? Have you ever used mice in your testing at all?

(Joel Beasley at 00:15:53) No. Every test we've done has been on humans. So the technology itself, fundamentally, is safe for human use. And actually, one of the interesting things about Kernel is that it makes it so more people can get their brain measured without a kind of medical condition. So I never had any kind of brain measurement done before I joined Kernel. And I've done dozens or maybe hundreds at this point of brain measurements.

(Joel Beasley at 00:16:20) Like size? Like size of your physical brain?

(Ryan at 00:16:23) No. Just like, you know, what's going on inside my brain? Like, what's my brain doing? So, okay. So we've done all our testing on healthy volunteers. Important. So, but, you know, people choose to participate and help evaluate this. It's safe. We need safe levels of light, so we never had to do anything in animal models or mice, pigs, anything like that. And I don't know. You mentioned seeing our headset or our big metal helmet, as you described it, and that isn't really mouse-size. So we'd have a little bit of an engineering challenge, I think, just to make a system small enough that could go on a mouse.

(Joel Beasley at 00:17:02) Oh, man. Have you found anything new or interesting about humans in all of this research?

(Ryan at 00:17:10) Yeah. That's a good question. So we've had the technology online for a little over a year, and a lot of that time has been doing things that have already been done, so what we call validation tests. So showing that our system can reproduce the same results that others have gotten previously. The one kind of exploratory thing that we started doing was this study with ketamine. So we had a partner who sponsored a study on looking at could a person wear Kernel Flow while experiencing a psychedelic experience? And we chose to use ketamine for that because it is probably the most easily accessible psychedelic. We went through the FDA and had an ethics review board kind of oversee the study and did everything the proper ways. But it's still a lot easier than going to one of the more intense psychedelics. So we looked at in healthy individuals, what does ketamine do to the brain? And we had a small study. It was only 15 people, but we saw some interesting results from that. And we're going to share, I think, next week at a conference, some of the preliminary analysis. And then probably later this fall, we'll release a paper on some of the findings that we had from that ketamine study. So it looks promising. I don't want to reveal any results right now. But we are on the verge of maybe seeing something new for the first time with Kernel Flow. And part of that is enabled—I mentioned before, if you look at the legacy devices that were used to do these type of measurements, they all exist in research settings, hospitals, big rooms that are dedicated to these types of equipment. And we just took our device to a clinician's office in Marina Del Rey who treats patients. And they are able to look out over the ocean and be in this very comfortable, more natural environment, like calming, a great place to have that type of treatment done. So we took our device there and did all of our measurements in that setting. And we're able to record data in a relatively short amount of time and get some information about what's happening inside the brain when a person is experiencing a psychedelic trip due to ketamine.

(Joel Beasley at 00:19:30) And so you're watching the blood flow during the trip?

(Ryan at 00:19:34) Yeah. Exactly. And it's not just blood flow. What we're looking at specifically is how different parts of the brain activate in different sequence or with each other. So some parts will be synchronized. They'll be activating in tandem, and others will be asynchronous. And so you can kind of look at how these different parts of the brain are connected and working together, and that tells you something about what's going on in that person's state of consciousness or whatever it is that is being represented by those, what are called networks, brain networks.

(Joel Beasley at 00:20:07) So then that's your software sitting on top of the physical device. Just the physical device only measures the oxygen in the brain and the location of the oxygen in the brain, and then all the other interpretations, sequence of what areas are firing, all that stuff is in the software layer.

(Ryan at 00:20:25) That's right. Yeah. So we have a team of neuroscientists and data scientists that kind of take both the understanding of what the brain does and how the function is understood from a scientific perspective and then kind of apply data science techniques onto that understanding and the data we've collected to try and draw conclusions. So it's a really—you know, our team is so talented in, I don't know, like 13 different dimensions from, as I mentioned before, kind of the device that we start with all the way up to analysis layers, where we're doing like machine learning and looking at ways to infer information from the signal we measure.

(Joel Beasley at 00:21:07) What about, like, have you ever noticed that people could do something physical that changed the results somehow? Like, if my breathing pattern changed over the course of the measurement, like any sort of physical change in my body would also affect what's going on in the brain?

(Ryan at 00:21:29) Yeah. So these are what are could be considered as confounds. So you would say, like, how do you know it's brain activity and not just, you know, something due to change in respiration? And there are a couple of ways. So one, we have one heart and one set of lungs. So changes in respiration or heart rate, like pulsing, those are global changes. We tend to see them everywhere. So it's kind of a common mode signal, if you were to think of it that way. The other way is by looking at the frequency content of these different signals. So, you know that respiration—I think, you know, adult humans are something like 15 breaths per minute or somewhere in that neighborhood. So you know the frequency of respiration, you know the frequency of heart rate, right, like, you know, resting heart rate somewhere in the 60, 70 beats per minute. And so you can use that information to kind of look at the frequency content of the signals you measure and kind of isolate what is coming from different parts of it. And then you can intelligently kind of pick that out and say, like, this is not due to the brain. This is. But the biggest thing is that really being able to measure these maps give us an idea of these global patterns so we can see what is common mode due to our cardiovascular system and what is differential due to changes in brain activity in those regions.

(Joel Beasley at 00:22:52) Can we take a look at some of this stuff? Can I pull up, like, a website or something?

(Ryan at 00:22:57) Sure. Kernel.com.

(Joel Beasley at 00:22:59) All right. Let's make this bigger. Oh, it looks much nicer.

(Ryan at 00:23:02) Yeah. The version we're about to release is even—I think looks even nicer than that. But these three still images here are pretty interesting, and I can talk to those quickly to kind of give you a sense of what the data looks like.

(Joel Beasley at 00:23:15) Yeah. It looks like heat map for people that can't see it, which they should go watch on YouTube. It looks like a heat map of the brain along with some charting of whatever that you'll tell me. Ketamine dose?

(Ryan at 00:23:28) Yeah. So this was a pilot we did before we did our 15 ketamine participants. And what we did this pilot for was to see if we could measure the difference between before and after administration of ketamine. And on the leftmost figure there, what you see is this brain map that I was telling you about. So this is a functional connectivity map. So we've identified or kind of carved the brain up into a hundred different regions. And then we look at how different regions are correlated with one another. And the strength of that correlation is represented here in this map. And then once ketamine is administered, we look at how those, the strength of those correlations change. And so if you look at the after ketamine takes effect, what you can see is that there's a decrease in activity in a lot of these regions. And that's thought to be kind of the dissolving of some of the networks that the brain has to kind of maintain its conscious, awake state. So as it enters this kind of psychedelic, more free-form brain state.

(Joel Beasley at 00:24:36) It got rid of the red ones, so it turned down their neural connectivity?

(Ryan at 00:24:41) So, Joel, the best analogy I love to use for this is if you kind of imagine all the major cities of the world. You know, they each have an airport, and you've seen like these maps of flight paths between like major airports. And you can see that like New York's a major hub, London's a hub, Tokyo, wherever the main hubs are, Hong Kong. And you can see the strength of connection between those hubs. And what you're looking at in this map is really the strength of connection between regions in the brain. So it's like how many flights per day between those paths, kind of. But it's really just how those brain regions are connected, how strongly linked they are. And then if you think of major world events, like a volcano erupting in Iceland, and you look at how that changes the flight patterns, you could see—you could very clearly see just by looking at flight patterns alone, without knowing anything else about the world—but the flights between New York and London are disrupted for some reason. Right? So there's something going on in the world between New York and London, and there's no connectivity between those two nodes now. And that's a similar concept here. So that would be a blue line in reducing the amount of connection between—in the bottom. So on the bottom panel, the blue would be a reduction in connection. It could be a disruption between New York and London. Does that help kind of relay, relate it to something a little more tangible? So what we're looking at, a strength of connection between different regions.

(Joel Beasley at 00:26:06) So were there, are there any new connections between the two or no?

(Ryan at 00:26:12) So this isn't reforming any connections. At least this pilot analysis shows no reduction here. And, again, I don't want to overstate any claims here because this is pilot data that's presented. And as I mentioned, we'll release a full publication later this fall that will go through kind of full peer review process.

(Joel Beasley at 00:26:32) I mean, the premise that drugs can slow down brain activity is not something that's hard to sell. Just being real here. So no new connections. Those are two identical maps, and then so, basically, it just cooled down the connections. So, basically, is slowing down the brain the right word or no?

(Ryan at 00:26:54) No. It's not really slowing down the brain. It's just changing the way the brain is synchronized, connected, however you want to think about it. But it's the strength of two regions working together is this connection.

(Joel Beasley at 00:27:07) Oh, I'm—you're helping me now. I'm starting to get it. Okay. All right. So, basically, like, you could reroute data more evenly.

(Ryan at 00:27:15) Right. Yeah. Yeah. So you can think about it as like these red ones are like really deep ruts. And then the drug comes along, and it's smoothing over those ruts a bit. So they're a little bit shallower across the whole brain rather than having a few deep ones.

(Joel Beasley at 00:27:31) Got it. Okay. That's kind of cool. So I wonder if you had somebody wear it while they sleep if you would get like similar results from the processes of like when we sleep pruning our synapses and such.

(Ryan at 00:27:45) Yeah. I mean, that's a really good question. I would say our headset today is not really designed to be worn during sleep. I think we'd have to make some adaptations in order to support something like that. But it would be a really interesting study. And what I hear you getting at, though, is just like, it would be really interesting to see what happens to the brain during blank. And you could fill in that blank with so many different activities. And I would try to argue that Kernel Flow can't solve all those activities today, but it's greatly increased the set of possibilities that you could fill in the blank with just by its new form factor and ease of use.

(Joel Beasley at 00:28:23) Yeah. You just need to put this out here and let the monkeys monkey with it.

(Ryan at 00:28:26) Yeah.

(Joel Beasley at 00:28:27) That's pretty cool, man. You must really enjoy your work.

(Ryan at 00:28:30) It's great. It's a lot of fun. And, you know, I should also qualify all the information I just said. Like, I'm an electrical engineer. My neuroscience training has been on the job. So, you know, I could have got a fact here or there wrong. But for the most part, general idea is correct. It's a lot of fun learning new things and really trying to understand what are the most accessible questions we could ask and how can we ask them in a way that will get good and meaningful results.

(Joel Beasley at 00:28:59) Why psychedelic? Why ketamine? Why was that your team's first thought of, like, hey. Is it because you guys just want to get your hands on some ketamine, like, legally? Or did, like, why was that the first test to run?

(Ryan at 00:29:12) Yeah. So we actually had a customer come to us. So that was done in partnership with a company called Seydin, and they were interested in understanding more about what happens in the brain during the administration of certain psychedelics. Ketamine is not their primary psychedelic. You may guess from their name that they're interested more in things like psilocybin. Yeah. They're developing novel molecules to treat things like depression and anorexia, and they want to understand more how the molecules functionally affect the brain. So this was their first question is just, can you measure psychedelic effect? Can you measure someone having a psychedelic experience with this device on? Is it safe? Are there any risks trying to understand what the landscape looks like?

(Joel Beasley at 00:29:54) I've been seeing a lot more literature and content around treatment using these sort of psychedelics for PTSD. So it's been, I think, what, the past, like, two, three years or so, it's like really kicked up. So the idea that somebody came to you and, you know, did a business with you in order to do some research on this and I don't know. Can they use psilocybin legally now through the FDA? Or did—because, I mean, ketamine's used as like a vet drug a lot. Right? So was it because you couldn't get your hands legally on psilocybin? Or...

(Ryan at 00:30:29) There are paths. They're just more difficult. So you're correct. Ketamine is an approved anesthetic. So if you, you know, have anesthesia for surgery, part of that cocktail may include ketamine even as a human. And it has also been approved in a form, S-ketamine, for the treatment of depression. And so it's like a nasal spray that's sold. So these are things that are out on the market already and FDA approved. Psilocybin, it's more difficult. It's a controlled substance, so there's a lot more approval and work that goes into doing it, at least in the U.S. So as a first kind of starter, psychedelic ketamine in the right dose. So you don't want to go to full anesthetic doses. Otherwise, you know, person just goes blank. So it's a sub-anesthetic dose of ketamine that produces a psychedelic effect.

(Joel Beasley at 00:31:19) Nice. What do we want to get out there to the world? We got, we have like six more minutes or so here. I want to make sure if there's any call to action for you guys or any publicity type stuff that you want to get out there that we do that.

(Ryan at 00:31:29) Yeah. I mean, the big call to action process is just if you are working on something that affects the brain, and you want to measure what that effect is, and you want to measure it in a quantified way, then come talk to us. We're very interested in partnerships. We have a lot of things going on with other drugs that are legal in most states in the U.S. and drugs that are used for treating certain conditions. So all those things are not yet public, but they're in discussions and things that we're planning in the background.

(Ryan at 00:32:01) I mentioned alcohol as kind of an easy, accessible starter drug. Again, you know, have some alcohol. We can go buy it off the shelf and administer it safely, again under oversight of an ethics board. So we're always doing our experiments on humans in a safe and responsible way.

(Ryan at 00:32:23) So to date, I would say a lot of things that affect the brain have been built on subjective measures, right? Can you tell me how you feel after I administer this drug to you, or after you take this treatment, or you play this video game, or whatever? Can you tell me how fatigued your brain feels today? And at Kernel, really what we want to do is take the subjective out of that and bring in an objective measure, which is directly from the brain.

(Ryan at 00:32:51) Right? So a lot of things, you know, to get these subjective measures, you're using proxies. It's like a survey or questionnaires or looking at things like, I don't know, your heart rate, how your heart rate changes, and saying something about how maybe your focus or attention is. And what we want to do is get away from proxies and get straight to the source, right?

(Ryan at 00:33:12) So much of what makes us who we are and makes us unique as individuals is in our brains and how our brains respond to things. And, you know, we'd really like to get there and quantify that.

(Joel Beasley at 00:33:24) What are your thoughts on objective results from your—let's say there's objective results, whether it's from your device or not, just objective results of brain activity in different people—causing the look identical on the data, but they're having different subjective experiences? Do you think that that's possible?

(Ryan at 00:33:46) I think it is. At least what I would expect is that if you look at people having a similar experience—if you look at, let's assume for a bit that reports of that experience are normally distributed—so I would expect that the objective measures would have a tighter standard deviation than the subjective ones. Like, you'll see higher variance in what people report, because what is a level five pain scale to me maybe a level two pain scale to you, for instance.

(Ryan at 00:34:16) Right? Like, we all have different reference points for how we convey what our subjective experience is. So I would expect, kind of, you look at things in group or population level, you start to see these broader distributions from subjective measures than you would from objective. And if we could show that, then we could start to build trust in the objective measure that, like, yes, we were able to form consistent distributions. It's not just random.

(Ryan at 00:34:40) And with this, we're able to more deterministically develop things. And, you know, even if your subjective experience is slightly different, right, like we're applying different gain or offset to our relative conveying of these scales on one to ten, we at least now have an objective measure to push it in the direction we want it to go. So maybe we want to reduce some kind of suffering score or increase some sort of pleasure or focus or attention score. Then you can start to have objective measures to do these things regardless of what subjective experience may be.

(Joel Beasley at 00:35:12) Yeah. Well, as an interviewer and an engineer, my job is to find the edge cases. But, yeah, in general, you can use this, collect data, and find directions that you might need to go. So maybe someone comes in and they're like, I don't know what's wrong with me. I just don't feel like myself.

(Joel Beasley at 00:35:26) They could put it on. They're like, oh, okay. Your brain flow is that of one that matches this type of depression. And generally, these two or three treatments are where you would want to start. It would just help. Is that kind of an area?

(Ryan at 00:35:38) I mean, I think that's one of our long-term visions for a device like this, especially in the realm of depression. Is, you know, can we differentiate types of depression and people who are good candidates for treatment in certain directions? I think that would be—if we could achieve that, I would personally be very satisfied. And I think the team as a whole would be as well.

(Joel Beasley at 00:36:01) Does anybody ever refer to you as somebody who, like, reads minds?

(Ryan at 00:36:07) We get all kinds of interesting inbound things through our [email protected] email address. Yes. And some of those are, like, you know, mind reading and wanting to know, like, what this chip that was implanted in their brain was doing through whatever, you know, NSA or someone. Like, we get all kinds of wild stories that come in. But at times, yes, we're referred to as people who read minds.

(Joel Beasley at 00:36:34) That's a podcast we need to start, like, your support inbox. That should be the podcast. And we just talk about, like, the—because I'll tell you what, people send the craziest things in. It's really great to meet you, man. Did we get everything done that we needed to get done? Kernel.com. Easy name for the tech community to remember.

(Ryan at 00:36:53) Yeah.

(Joel Beasley at 00:36:54) Good job getting that domain, by the way.

(Ryan at 00:36:56) It was work. We were kernel.co for a while, so it took us some time to get that com.

(Joel Beasley at 00:37:02) Cost you? Oh. It's private.

(Ryan at 00:37:05) I don't actually know, but I don't know that I want to know either, so...

(Joel Beasley at 00:37:09) Correct. Correct. Well, this is fantastic. We made a podcast. How do you feel?

(Ryan at 00:37:14) Oh, I feel great. This is fun.

(Joel Beasley at 00:37:16) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email: [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.