Most brain-training products begin by demanding patience from the exact brain that is already distracted, stressed or exhausted. Jamie Wood wants the opposite: ask as little as possible, find the highest-leverage behavior and deliver something useful fast.

In part two, Jamie walks me through AUTONOMIC as if it were a personal trainer for the brain. We unpack the questions, recommendations, feedback loops and social proof behind a product trying to make better sleep, focus, mood, energy and stress regulation feel less like homework.

Transcript

Jamie: To explain to everyone what we're doing and what we were doing, was being able to analyze an individual by asking specific questions that are tied to neuroscience-based proven behaviors that either enhance brain function or detract brain function.

Pablos: Ask me some question that neuroscience has indicated is tied to brain function.

Jamie: Do you look at your devices within 30 minutes of waking up? Yes. Do you drink caffeine past 3:00 PM? Do you provide corrective feedback within 48 hours of an instance?

Pablos: I don't know what that means. Instance of what?

Jamie: Do you drink water before caffeine in the morning?

Pablos: I don't know. Probably not.

Snooze function?

Jamie: Snooze function on your phone. Does an individual want to lower stress, improve energy, focus, mood, or sleep? From there: different surveys that we've created, anywhere from eight to 14 questions max. We like rapid time to value. We shouldn't have to ask you 45 things. You watch that annoying progress bar on your quiz: "Screw it. I'm out."

Pablos: I don't do surveys.

Jamie: We begin to ask the least amount of questions to gain the biggest understanding of who you are. You can fine-tune later. Asking questions like that, you can profile me, figure out the things this guy could do to focus, to reduce stress. That's step one.

Now that you've profiled me as the particular kind of fringe wacko that I am, what's the example? It doesn't have to be accurate. Let's say you want to improve your focus, which from our conversations, you don't have a problem with. I'm kind of amazing at everything. If you can find a cost to it, I'll take it.

Pablos: I don't know what problem I would complain about.

Jamie: Improve focus. From there, your AI coach will make a habit recommendation which comes along with a science of why. We want you to ask why. As an example, get off your device 30 minutes before bed. Let's say you want to improve sleep. We're off focus now.

Pablos: "Why do I have to do that?" And you're gonna come back with blue light and, or whatever.

Jamie: You take the user through a journey. What's the one sentence direct science? What's the 50-second video? Where is it found in any sort of social proof? In Science Daily or whatever publication? Reddit. Subreddit. People who can't sleep saying, "It's amazing if you stop using your phone 30 minutes." Where is the peer-reviewed paper this is tied to? Is that paper properly funded, cited? Who are the researchers?

You can go down an information rabbit hole if you want. Some people do. It's great to have it staged.

Pablos: Tweet. TikTok.

Jamie: YouTube. Paragraph. Paper. From there, based on getting a piece of information you didn't have before, you make a decision to accept that habit into your training or reject it. Or you could skip it and push it for a bit. If you reject it, you have to provide your coach a reason why, which provides even more information around how we can personalize and iterate for you.

Pablos: The 30 minutes before bed: stop using your phone. How many are there? Dozens? Hundreds?

Jamie: Right now, in our platform, we're talking about subjective analysis—one modality on the platform. We look at 92 neuroscience-backed behaviors across 26 categories administered through 17 proprietary surveys for the subjective user.

Pablos: There's 92 possible problems I could have.

Jamie: Yes, what you're going to resonate to—and one of the most important things in learning for any individual—is getting the right information at the right time that's applicable to them.

Pablos: Of 92 possible problems, you're gonna figure out which ones you think are significant for me. Of those, I'm gonna figure out which ones I agree with you on or are palatable enough for me to work on.

Jamie: Some people wanna change.

Pablos: Not quitting caffeine.

Jamie: Exactly. Throw that one out. You're supposed to be the expert. Figure it out. Let's say you're working on two habits—two behaviors that are helping you improve your sleep. Where I nerd out, in terms of being excited around what can happen, is: you start getting into the patterns, insights, and trends based on why you engage in a habit sometimes but don't.

Where's that deeper data insight that we can feed back to you? "It's really interesting, Pablos. We see that every Friday you don't do X." Then it influences these other metrics. This is subjective, but you can move into biometric data capture, passive data capture, and it becomes even richer.

Pablos: A great example: I get up pretty late. I'll have lunch relatively early. It's my only meal, lunch and dinner. I don't have breakfast. Say I get up at 9:00 or 10:00. I think I'm gonna have caffeine at 2:00, but I usually do it by 1:00. I have a Red Bull once a day. It's the only caffeine I have, and I'm good to go till midnight, 1:00, 2:00 AM.

I'll have dinner at 6:00 or 7:00. I'm on a schedule with those things. With the food, if I miss the schedule, I'll get hangry. Nobody likes me anymore. I'll feel abandoned and alone.

Jamie: Judged.

Pablos: I'm on schedule for the food 'cause I need it. The caffeine—I need it at that time. I'm a junkie. I'm not sensitive, and I'm not self-aware about performance on anything. I don't know: eating more or less or sooner or later, having caffeine more or less or sooner or later, or any of those things. I can take any supplement Thorne makes, and I won't feel a thing. I don't know if I'm just wasting this stuff.

I don't know if it's doing something good for me or not. Having some passive monitoring—TikTok probably knows if any of this stuff is affecting me. TikTok, I assume, knows everything. TikTok knows if I should be having more or less caffeine, or sooner or later. If you could tap into some of the monitoring that's available: I have an Apple Watch on. You probably can't get access to the data you would want from that. I have an iPhone on me 24/7. Probably you can't get meaningful data off of that.

I'm not wearing an EEG yet, but I assume one day I will be.

Jamie: Very quick.

Pablos: It's coming very quick, is it?

The point is: with active monitoring, you could start to correlate whether these behavioral things I'm doing are affecting, in a positive or negative way, stress response or focus.

Jamie: So if we keep our conversation in the realm of brain performance and measurement of that, and we think of a scale, and at one end of the scale it's low fidelity, high adoption, and that's like your surveys as an example. Lower fidelity 'cause it's subjective but high adoption 'cause it's really easy to get that to everybody. And then at the other end of the scale, it's Neuralink implants. High fidelity, low adoption. Then we think about what else is on that scale. So from surveys, you can then move to heart rate variability.

A lot of people are wearing Fitbit, Whoop, what have you. From there you could move into voice. Voice is showing some of the best promise for early detection of Alzheimer's. We're all talking on our phones. There's gonna be apps coming out that can help predict, monitor, measure. Then we move into mobile EEG: wearables you can have out of the lab, like a Muse headband, Cognionics—kinda like an octopus-looking thing that sits on your head—and everything in between. Even gaming, VR—all of those sensors are coming in for mobile EEG. Being able to measure the brain in that way, we can go into fNIRS, fMRI, so on and so forth. That's the scale: low fidelity, high adoption to high fidelity, low adoption. Autonomic is agnostic to the measurement system. Our objective was to first prove it out with the surveys, then begin to integrate with all of what's coming in the industry. That platform takes it all in. That seems pragmatic.

How far along are you with this now? We've got our first product in the market: brain performance coach. We explain the value prop like it's a personal trainer for your brain. That's the subjective piece that we've just talked about. This is a logical extension of what you started out trying to sell to finance executives, pivoting into selling to tech people because they were more open-minded. Totally more open-minded, for better and worse.

Those people became customers of the popsicle stick version. Then you built the first software version. That's where you're at now, deploying that to more and more people. One of our incredible partners, Dr. Olav Krigolson—he runs the Krigolson Lab in UVic—is also one of the world leading mobile EEG neuroscientists. He has one of the most cited papers in the world proving that Muse, that headband, is good enough to capture full readings from the brain. We have an IP assignment from him.

We're the only company that can commercialize his algorithm, and we've done a proof of concept of the next phase of our product suite: being able to have a brain scan and customized training based on different brain health biomarkers. Can you get those from a single scan in a lab, or do you need to monitor these people over time? You would always wanna monitor over time. We want to make this accessible. You could buy a headset and run it on a Muse headband.

Pablos: Are people wearing these things in daily life and giving you data, or is that too big of a jump?

Jamie: They're not doing that right now. You want somebody to do a brain scan 10 days to two weeks apart—two brain scans to establish a baseline. Some things you can do that influence brain function right away. Other things take time. You can give people Autonomic, they can use it, and they can measure their brain again and see how it's changed and improved.

The product now: if I'm a user, I sign up, I do some surveys. The thing tells me to stop using my phone.

Pablos: Then it gives me alerts to do things that are good for me. At some point I get to a lab, I get a couple of fMRI experiences under my belt. I get a Muse headband, and this starts to inform the app on how to make better recommendations to me. Is that all? I'm getting that so far?

Jamie: If you had integration with your Apple Watch, and you're reporting your stress at a six out of 10, but we're seeing strain, maybe your stress is higher. Maybe you should be doing this particular behavior to help activate the parasympathetic nervous system more. The most important thing around wearables and the ability to capture data is that it's all well and good, but unless you're answering the now what and how, you're leaving people hanging. We wanted to focus on answering the now what and how. Say you keep reporting your stress is high. We're making this recommendation around this particular behavior you should engage in that we know would help lower your stress. Let's say there's 14 or 16 of them. This one resonates with you. "Stop being a parent. Quit your job. Pay off all your debts." I get it. Great recommendations.

It's awareness development. You cannot shift behavior without awareness. Is this app able to recognize things that maybe I wasn't aware of and let me know? Yes. That's what we're looking to scale. That's what you wanna get to. You don't necessarily need a sensor. The interesting thing is that we've seen a ton of success and proven efficacy just on the subjective alone.

Pablos: You send me an alert saying, "One more survey question. Are you stressed?" "Yes." "A lot." "What's stressing you out?" "We have a recommendation—"

Jamie: "For you."

Less survey questions.

Pablos: Can you give me a concrete example of what a user might experience?

Jamie: A user would experience receiving a highly personalized training that iterates and customizes to them based on their interactions with habit recommendations, and being able to communicate with their coach.

Pablos: Let's walk through an example. Is this all iPhone type thing—an iPhone app or a web app?

Jamie: Currently web app-based.

Pablos: I have a web app on my phone. It asks me some questions, and it's profiling me over the course of days or weeks, and scheduling a Zoom call with my coach. Am I getting this right so far?

Jamie: There's no Zoom call. It's all done asynchronous. Very quick. A product like this had to integrate with an individual quickly and not be distracting. You have to feel like you're getting value. You have to feel like it's not taking a lot of your time.

Pablos: That's very important. Is the thing sending me alerts?

Jamie: It does send you alerts. It does send you reminders: check in to rate those five key metrics. Answer yes or no if you've performed these habits. It's also providing you an education piece. You have a personalized library of the science of why and papers that you're working on. If you think about an N of one experience, just because a certain paper has said, "Don't drink caffeine past 3:00 PM," does that sample size actually reflect you as a human? That's what becomes fascinating in terms of the data we can collect.

Pablos: I think it's super important. I'm not trying to quit caffeine.

I don't have a problem with falling asleep, so I can use my phone, go to bed. I've been training on screens since I was nine. I'm good at that. I don't have any negative effects that I notice. I can look at my phone in the morning. My day'll be fine even if I read about whatever war broke out on Instagram, TikTok, or Twitter. I'm not trying to quit those things. You can figure out that I don't care and move on.

Jamie: Exactly. That's exactly it, and also it's back to that curiosity piece. Learning with you about learning about yourself as it relates to brain performance. That's a really interesting interaction that occurs between you and Autonomic. And right now we still have human in the loop on the back end. Very specific around who we went after to build our data sets to then start to train the models, and who was making these recommendations. And what is the information that is being shared, and what are the papers that it's tied to. We haven't let an AI loose yet because we actually want to then prove efficacy on the AI, 'cause we've proved efficacy on the human in the loop experience, but the next phase is proving it on the AI. That's probably not gonna be too hard to do well at this point. Right. Agree there's a thing I've observed about humans, which is that, they seem to be a lot more convinced about things to try

Pablos: By the friends on their softball team than they are by

Jamie: Scientific journal articles.

Pablos: It seems like one of the things you might be really benefiting from is, or could benefit from, is to make a sort of social interaction version of this where it's like, oh, you're like these people who've

Jamie: Been on the app for a year longer than you. Here they are to tell you how it's working for them, why you should do it.

Pablos: You can ask them questions. That kind of thing might really supercharge this because humans are just evolved to believe humans even if they're uninformed dumb shits, right? That's just like any random human with an opinion is way more influential than a scientific journal article.

More Resources

Recorded on January 16, 2025