Button TextButton Text
Article

Inside the Science of Nonsurgical Brain-Computer Interfaces with Subsense Co-founder and Head of R&D Dr. Scott Meek

Singularity Discussions Recap

Singularity
•
November 16, 2026

Brain-computer interfaces (BCIs) could transform how we understand and treat neurological conditions, and how we interact with computers. But today's technologies come with major trade-offs. Noninvasive systems like electroencephalography (EEG) can monitor brain activity without surgery, but provide limited insight since they sit far from the brain’s neurons. Implanted devices can access much more precise neural signals, but they require invasive procedures that most patients would rather avoid.

Subsense, a startup developing a new kind of BCI, wants to bridge this gap. Dr. Scott Meek, co-founder and head of research and development, details the company’s technology and how it could change BCI in this episode of Singularity Discussions. Subsense uses nanoparticles that are delivered through the nose and guided into the brain using magnetic fields. The particles can read or stimulate neural activity, with a wearable headset serving as the interface.

The goal is to access neural tissue without surgically implanting electrodes, a development that could make brain-computer interfaces far more accessible. “A lot of this technology is about taking neurotech and moving it to something that is scalable beyond very few people,” says Meek. Subsense is targeting a treatment for Parkinson’s disease to start, but sees potential to expand to many other medical and non-medical uses.

Two Types of Nanoparticles, Working Together

Nanoparticles are microscopic pieces of matter just 1 to 100 nanometers wide. “For reference on the size scale, they're much smaller than most human cells,” says Meek. “They're on the order of the size of a virus particle.” The particles are synthetic and predominantly made of metals. They’re divided into two different types, each with its own function.

The first is magnetoelectric particles, which are designed to stimulate neurons. When they’re exposed to a magnetic field, the particles compress, and that mechanical motion creates a local electric field that can activate nearby neurons.

The second type, called plasmonic nanoparticles, are designed to read neural activity. These particles scatter light at different wavelengths depending on their surroundings. “If there's a neuron nearby, it will change the wavelength of the scattered light based on if the neuron is firing or not,” Meek explains. By detecting those changes using near-infrared light—which can pass through biological tissue more easily than visible light wavelengths—the system can monitor neural activity.

Meek describes the particles as a kind of wireless electrode. The initial wearable device, which looks like a baseball cap, will have optical components and magnetic-field generators. A separate device will be used in a clinical setting to deliver and guide the particles to targeted areas of the brain.

A Non-surgical Treatment Method 

Subsense hopes to develop a nonsurgical alternative to deep brain stimulation (DBS) for Parkinson's disease. DBS uses surgically implanted electrodes to stimulate specific brain regions and help manage symptoms. Meek says that while an estimated 10 to 15 percent of Parkinson's patients may be eligible for DBS, only about 1 percent receive it. 

“By the time people are eligible for this treatment, they’ve already moved through drug treatment and are typically medically frail, and they just don't want another surgery,” Meek says. “It's too much.” A less invasive approach could make brain stimulation accessible to more patients and potentially allow treatment earlier in the course of the disease.

The company has successfully tested key parts of its system in mice, including neural reading, neural stimulation, and nanoparticle navigation. Its next step is to test the technology in larger animals, including pigs and sheep, while continuing to develop the hardware and conduct safety studies. It’s submitting its biosafety testing protocols to the FDA “so that patients can have assurance that this is all being done at the highest level of safety,” Meek says.

A New Window Into the Brain

Subsense scientists are using AI simulations to model the physical and chemical properties of nanoparticles, screening thousands of possible designs. AI also helps model magnetic fields, inform hardware development, and analyze unusual neural signals.

Beyond the initial goal of treating neurological conditions, Meek sees a much broader opportunity in collecting neural data over time. A continuous stream of information about brain activity could help researchers learn more about diseases such as Alzheimer's, Parkinson's, and Huntington's, whose underlying mechanisms aren’t well-understood. “In that data, I think we're going to have insight about some of these neurodegenerative diseases that have eluded us for decades,” Meek says.

In the longer term, Meek imagines BCIs changing how people interact with digital technology. Instead of relying on screens and constant notifications, people might eventually receive information in ways that feel more natural to the brain. There could also be applications in improving sleep, focus, and anxiety, as well as creating more productive ways to interact with AI.

Protecting Our Most Personal Data

Moving from animal experiments to human applications will require more than demonstrating that the technology works. Subsense will have to demonstrate that its nanoparticles can be delivered safely, be effective for a predictable period of time, then clear from the brain without accumulating. The company is designing the particles to clear naturally, with a target usable lifetime of about three months. Testing in mice has shown no adverse effects thus far.

There are also big questions around data security, such as who can access patients’ neural data, how it might be used, and how it’s kept safe. “Brain data is going to have to be as secure or more secure than financial data,” Meek says. “We need to have that level of trust if any of these systems are going to work.” He adds that the broader BCI industry will need to establish shared standards and work with regulators to gain peoples’ trust.

For now, the next challenge facing Subsense is to demonstrate that its technology works in larger animals before ultimately moving to human trials. Meek believes they’ll get there, and not a moment too soon. “There are a lot of conditions that we've made very little progress on using traditional methods,” he says. “I worked on Alzheimer's in my PhD, and we still don't understand the fundamental mechanisms of that disease 20 years and billions of dollars later. We need another level of technology to understand the brain and help manipulate it.”

‍

Get access

Singularity

Singularity's team of internal thought leadership works to develop interesting resources, articles and insights about our core areas of expertise, programs and global community.

Transcript

Singularity Discussions | November 2026 | 60 min

Reading the Brain Without Surgery: Dr. Scott Meek on the Future of Non-Surgical BCI

Brain-computer interfaces have mostly forced a trade-off: accurate data that requires brain surgery, or safe, non-invasive tools that only hear the brain "through the wall." Subsense is betting there's a third path. In this session, co-founder and head of R&D Dr. Scott Meek joins Singularity's Cody Hanson to explain how nanoparticles delivered by nasal spray, guided by magnetic fields, and activated by a headset the size of a baseball cap could bring high-fidelity neurotechnology to millions of people, starting with Parkinson's disease.

Key Takeaways

  • A third option for BCI. Subsense aims to combine the accuracy of implanted devices with the scalability of non-invasive ones by placing nanoparticles directly in brain tissue, no surgery required.
  • Proof of concept is done in mice. Neural reading, neural stimulation, and nanoparticle navigation have all been demonstrated in mice. Large-animal studies in pigs and sheep are next.
  • Parkinson's first, everyday life later. The first target is a non-surgical alternative to deep brain stimulation. Epilepsy, treatment-resistant depression, and OCD could follow, with non-medical uses like sleep, focus, and screen-free data access further out.
  • Built-in safety by design. Nanoparticles are engineered to clear naturally after about three months, and they do nothing without the wearable headset. Take it off and the system is effectively air-gapped.
  • AI is accelerating the science. AI simulations that once took more than a year now let the team screen hundreds to thousands of nanoparticle designs, and AI helps decode neural signals no one has seen before.
  • A 5 to 10 year horizon. Scott expects a major ramp-up in medical BCI over the next five years, with everyday applications emerging in the five to ten year window.

About the Speakers

Dr. Scott Meek is co-founder and head of R&D at Subsense, which is developing non-surgical brain-computer interfaces. His background spans materials science, chemistry, and sensor science. He earned his PhD at MIT, developing near-infrared dyes to detect beta-amyloid plaques for Alzheimer's diagnosis, then worked as a postdoctoral fellow at Sandia National Labs on gas and radiation sensing nanomaterials. He spent the following 12 years leading R&D at early-stage biosensor startups.

Cody Hanson hosts the Singularity Discussions series.

This transcript has been lightly edited for clarity and length.

Meet the Expert

Cody Hanson: Welcome to the November 2026 Singularity Discussions. We do these every month with Singularity experts and alumni, and this is our first time featuring someone who isn't a Singularity expert but is doing really interesting work that fits squarely within the Singularity taxonomy. We're with Dr. Scott Meek, co-founder and head of R&D at Subsense. They're working on non-surgical brain-computer interfaces. Scott, can you give us a quick intro?

Scott Meek: Hi everybody. My background is in materials science, chemistry, and sensor science. I'm not a neuroscientist. I began my scientific journey with a PhD at MIT, where I developed near-infrared dyes to help detect beta-amyloid plaques and diagnose Alzheimer's disease. I then worked at Sandia National Labs as a postdoctoral fellow developing gas and radiation sensing nanomaterials. For the last 12 years, I've been leading R&D at early-stage biosensor startups.

As the last of those startups ended, I met Tetiana, my co-founder, and we discussed this project. It struck me as the combination of all of these things, in service of getting data from the brain, which is the final frontier of biodata. It's a very exciting project to work on, and I love it.

How the Subsense Interface Works

Cody Hanson: As I understand it, the Subsense interface is a nasal spray that sends nanoparticles through the nasal passage into the brain, plus a headset that activates them. Can you explain at a high level how that works?

Scott Meek: The nanoparticles reach the brain through the olfactory bulb and the olfactory nerve, and they're guided there with magnetic fields. Nanoparticles are so small that they can move in the intercellular space and be positioned exactly where we want them in the brain. Once they're in place, we have one set of nanoparticles we can use to stimulate neural activity and another we can use to read neural activity.

Cody Hanson: Where do they end up in the brain?

Scott Meek: Wherever we most want to stimulate or read activity. For example, to treat the symptoms of Parkinson's disease as an analog to deep brain stimulation (DBS), we could place nanoparticles in the subthalamic nucleus, run a stimulation protocol, and have a non-surgical version of DBS. DBS is a well-known treatment for people with Parkinson's, but it involves a craniotomy and surgery.

Cody Hanson: Is there anything you can share about how you get the nanoparticles to particular parts of the brain?

Scott Meek: We're using a magnetic navigation system. Systems like this are otherwise used for remote endovascular surgery. You can move nanoparticles through the vasculature and then into the tissue where you want them to act.

Cody Hanson: In simple terms, what is the nanoparticle? How small are they, and what are they made of?

Scott Meek: Nanoparticles are typically defined as anything less than 100 nanometers in diameter, and ours are a fair bit smaller than that. That's about the size of a virus, much smaller than most human cells. These are synthetic, not found in nature, and predominantly made of metals.

We have two types. The first is magnetoelectric. These have one component that changes shape in a magnetic field and another that is piezoelectric, meaning it creates an electric field from mechanical motion. When this nanoparticle is placed in a magnetic field, it compresses and creates a local electric field that can stimulate neurons.

For neural reading, we use plasmonic nanoparticles. These scatter light at different wavelengths depending on their local environment. If a neuron nearby is firing versus not firing, the wavelength of the scattered light changes. We tune those wavelengths into the near-infrared, where the body is most transparent to light. You can try a version of this yourself: shine a blue light through your hand and you won't see anything. Do the same with a red light and you'll see it coming through.

Cody Hanson: The mRNA vaccines used during COVID relied on nanoparticles to deliver messenger RNA into cells. How are yours similar or different?

Scott Meek: Some of the outer coatings are similar. We have to coat the surface of these particles with various chemical constituents to help with biocompatibility and dispersion, which is very similar to the polymer nanoparticles used in vaccines. Otherwise, the makeup is different. Nanoparticles sit in a wonderful sweet spot between materials science, biology, and chemistry. So, they're similar in how they interact with their environment and different in their mechanism of action.

Where the Research Stands

Cody Hanson: You've demonstrated state-of-the-art imaging of neuron activity using these nanoparticles. What stage of development are you in?

Scott Meek: We've now done neural reading in vitro, in cells in a dish. We've done it in simple living creatures like the C. elegans worm. And now we've done it in mice. We've been able to scale up the level of complexity, which has been very exciting.

Cody Hanson: Any major breakthroughs or setbacks in the last 6 to 12 months? And what's the most important next step?

Scott Meek: The major breakthrough in the last 12 months has been showing proof of concept of every key component of the system in mice: neural reading, neural stimulation, and nanoparticle navigation. No particular setbacks. It's been going according to the R&D plan, which is a real testament to how strong our team is. Next is large animals. We'll be starting work in pigs and sheep very soon, with the goal of proving the concept from a biological and physics perspective in those models.

Closing the Gap Between Surgical and Non-Surgical BCI

Cody Hanson: Your co-founder, Tetiana, has talked about the trade-offs between surgical and non-surgical BCI, especially the difficulty of getting close to neurons. How are you closing that gap?

Scott Meek: This is why the company was founded the way it was. In BCI, you have non-invasive approaches, like EEG, and implanted approaches, like Neuralink or the Utah array from Blackrock. Both are really powerful tools.

Non-invasive BCI gives you an easy source of biomarkers that help neuroscience and medicine, but it's limited in accuracy. People describe it as listening to someone talking in the next room. You're never quite going to get good data from it. Advanced AI models will make it better, but not all the way.

Implanted BCI gives you really great data, but it requires invasive surgery, which isn't scalable. It's appropriate for only a very small subset of patients, even among those with profound medical need. And it can only cover a small part of the brain.

If we want BCI to expand beyond niche applications, we need a technology that is accurate, which as we see it requires connection with the tissue. It also needs to be non-surgical so it can scale, and it needs whole-brain coverage so you can look at effects beyond a small area.

From Parkinson's to Everyday Life

Cody Hanson: Breakthrough medical technologies usually start with a very specific need, and for you that's Parkinson's. Is there a plan to expand into less serious conditions, or even non-medical uses?

Scott Meek: Absolutely. The plan is to start with high-acuity medical conditions. That's driven by the FDA and by wanting to create products for people who really need them. We're starting with Parkinson's, and then we could move to conditions like treatment-resistant depression and OCD. Beyond that, products for people who are mostly healthy, things that could help with everyday life.

People talk about human enhancement. For me, I would love a world without screens. It's a dream of mine to not spend all day on a screen. There are almost limitless applications for this technology, and I think it's going to be the breakthrough technology of the next several decades.

Cody Hanson: If we're talking about a world without screens, does that mean seeing your email or data in your field of vision?

Scott Meek: Maybe not even that. Maybe data as a normal function of your brain. We live in a world with constant streams of data coming at us. As humans, our senses bring us data in a way that's a far more natural fit with our consciousness and thinking than emails, Slack, and everything else that bombards us. Personally, that's the application I'd love to see someday: data coming in naturally. For others, think about improving sleep, focus, or reducing anxiety. There's a lot this kind of technology could do to benefit everyday life.

Cody Hanson: The idea of one system replacing antidepressants and whatever Slack does at the same time is a weird concept.

Scott Meek: It is a little weird, but that's why neurotechnology is such an amazing field.

How AI Is Accelerating the Work

Cody Hanson: You're using AI to explore nanoparticle design. What does that research process look like?

Scott Meek: Our R&D department has an amazing nanoscience program doing traditional materials science and chemistry, which has helped take these nanoparticles from an academic setting into something feasible for a medical device. We're complementing that with an AI program that simulates nanoparticle properties, so we can screen through hundreds or thousands of candidates as we build up these techniques.

So far, we've created a suite of simulations and validated them against actual nanoparticle properties in the lab. That all looks good, and it's new work. No one else is really doing this. Now we're using those techniques to explore the design space and find next-generation nanoparticles no one has even thought of before.

Cody Hanson: Chemical properties, physical properties, or both?

Scott Meek: Both. Ideally, we'll look at biological properties as well. That's another level of simulation.

Cody Hanson: Has it only become feasible in the last few years to use AI this way?

Scott Meek: It really has. About 15 years ago, I did some work predicting materials properties. It took us over a year to run a simulation for just a couple of materials. Now it's much, much faster, and it's accelerating. We've done a lot of the initial work and are now making it as efficient and computationally tractable as possible.

Cody Hanson: Does AI play a role in the product itself?

Scott Meek: Absolutely. We use AI to simulate magnetic fields, which helps us develop new versions of the hardware. And we use AI to decode signals. Some of the signals we're getting from this system have never been seen before, because this is the first time anyone has done this kind of neural reading. We've used AI extensively to comb through that data and confirm the signals correlate with known sources.

Cody Hanson: Are you working with electrical signals, chemical signals, or both?

Scott Meek: Predominantly electrical. You can think of these nanoparticles as wireless electrodes. So far, we've been able to accomplish everything we want. One of the really cool things about our neural reading technology is that it could be expanded to look at chemical signaling in the brain, and there really isn't any other technology that can do that. That's future work.

Cody Hanson: For Parkinson's, what do you hope this system can do that current treatments can't?

Scott Meek: Be accessible. That's what I really hope for. With current DBS technology, something like 10 to 15% of Parkinson's patients are eligible, but only about 1% end up getting it. By the time people are eligible, they've already moved through drug treatment, they're often medically frail, and they don't want another surgery. It's too risky for many patients.

With a non-surgical version, patients could access it earlier in their disease progression, and that opens up new types of therapy. Not only could more patients use it, but using it earlier could help maintain the efficacy of drugs longer. A lot of this is about accessibility, taking neurotech and making it scalable beyond very few people.

Safety, Regulation, and Ethics

Cody Hanson: We're seeing a lot of audience questions about risks and ethics. What are the biggest differences between demonstrating this in mice and making it work in humans?

Scott Meek: It's always a problem of scale, from cells to mice to large animals to humans. There's hardware development that has to happen as part of that scaling, which is what we're working on now. There's also extensive safety testing. We've done both in-brain and in-body safety testing in mice with no adverse effects, and we'll run more traditional biosafety studies as we move into larger animals.

We're doing all of this in collaboration with the FDA. We've already had initial conversations with them, and we're going through the pre-submission process and submitting our biosafety testing protocols so we have their buy-in. That way patients can be assured this is being done at the highest level of safety.

Cody Hanson: How do the nanoparticles exit the brain?

Scott Meek: We're engineering them to clear out of the brain naturally, with a usable device lifetime of about three months. That could also be accelerated with navigation. This has some interesting product features. If you don't like the system, it simply stops working after a while. It's also easily upgradable: as we improve the nanoparticles, you can install the next version with more features.

The other key safety feature is that the system relies on wearable hardware, like a hat or headset. You can always take it off, and the nanoparticles become inert. They don't do anything. You can't connect to them with Wi-Fi or any other device. It's effectively air-gapped.

Cody Hanson: Could there be long-term risks of accumulation with repeated use over months or years?

Scott Meek: That's part of the testing protocol. We dose animals and examine various organs for bioaccumulation. It's also a core design principle. We've gone beyond previous academic work to make sure the nanoparticles are designed to flush out, through a combination of design features. Surface chemistry is one of the more obvious ones.

Cody Hanson: What does the regulatory landscape look like for BCI right now?

Scott Meek: The FDA has been cautious about implanted solutions. When we talked to them earlier this year, they were very excited to hear from someone developing a non-surgical version. Many BCI devices have qualified for breakthrough designation, which shows the FDA's excitement about patient need, but there's been reticence about surgical platforms. We were very happy to see how collaborative the FDA has been. They'd love to see these technologies get out into the world. They're just hesitant about the surgery.

Cody Hanson: It's wild to think of a technology that can read what's happening in my brain. Has any thought gone into the next frontier of privacy and keeping neural data secure?

Scott Meek: Neural data has to be very secure. We've seen with social media what happens when we broadcast our data into the ether. It can be used for manipulation, including political manipulation, as we've seen over the last two decades. Brain data will need to be as secure or more secure than financial data. We need that level of trust for any of these systems to work.

That's something we as an entire industry will have to work on together, setting standards and working with regulators so there's public trust. For a technology like this, it has to be done in the open. There are already industry groups working on this, and we're part of that conversation.

Cody Hanson: Are there applications that are technically feasible but that you'd choose not to pursue on ethical grounds?

Scott Meek: Military applications. For us, that's a hard line.

The Road Ahead for BCI

Cody Hanson: Where do you see the BCI space going? Will it always be a mix of surgical and non-surgical?

Scott Meek: First, we're going to see an expansion of medical uses. There's already been great work with TMS and EEG in non-invasive stimulation to help treat treatment-resistant depression and OCD, and we'll see a lot more of that. We'll see the first scaling of implanted BCI for people with ALS and spinal cord injury, which is going to be game-changing for those patients. Then the industry will expand into larger patient populations, like mental health, and from there into applications for daily life. Non-medical applications will require really profound results in patients with medical need first, and I think we'll see those results.

There are many conditions where we've made very little progress with traditional methods. I worked on Alzheimer's 20 years ago in my PhD, and 20 years and billions of dollars later, we still don't understand the fundamental mechanisms of that disease. We need another level of technology to understand the brain and help manipulate it, and I think neurotech is going to unlock that.

Cody Hanson: What's the timeline?

Scott Meek: I think five years to much greater medical use. Over the next five years, we'll see that ramp up. In the five to ten year timeframe, I think we'll start to see this coming into everyday life. It will depend a lot on regulation, on how well the technologies perform in medical applications, and on our industry's ability to build solutions that scale. That's why we're doing what we're doing. We believe this is the path to scalability.

Audience Q&A

Cody Hanson: What might be some unforeseen side effects?

Scott Meek: Here's one side effect of traditional neurotechnology: with implanted electrodes, a biofilm forms around them, so the power output has to increase over time, the surrounding cells eventually burn out, and the system stops working. Our system gets around that by being removable and refreshable. That's how we think about it: what are the potential side effects of neurotech, and how can we design around them?

Cody Hanson: Several people asked how you're ensuring the technology works effectively across genders.

Scott Meek: We haven't seen any gender differences in our work so far. That gets into a broader neuroscience question about differences across types of brains, which is a bit beyond the scope of what we've done. The good news is our technology is very flexible. Even with deep brain stimulation, everyone's brain is a little different, and where you place the nanoparticles will vary from person to person. Our system is tunable in a way that's harder for other systems, so it should be tunable across different populations as well.

Cody Hanson: Do you need focused magnetic fields to move the particles into place? And would that require a medical facility?

Scott Meek: Focused magnetic fields. I can't go into detail on the types of fields because that's proprietary. The initial version will use a separate machine for installation and navigation, distinct from the wearable hardware, so it would be done in a clinic. Think of something smaller and less complicated than an MRI.

Cody Hanson: How do you see the data being visualized? Could it be rendered as a 3D brain map?

Scott Meek: I like to think of it as that 3D brain map image. We've actually done this in our C. elegans experiment with our collaborators in the UC system. It's very cool when you can visualize that signal in a living creature. We're doing that mapping in mice now, and we plan to publish, with the proviso that it makes sense from an IP perspective. We've been trying to publish as much as we can because it also helps build trust.

Cody Hanson: Where can people find your research?

Scott Meek: Our C. elegans work hasn't been published yet, but it should be soon. It will be our first major publication. In the meantime, our website has a lot of background information.

Cody Hanson: What does the receiver or control module look like? What's the hat?

Scott Meek: It looks like a baseball cap. That's our initial design. It has optics and field generators in it. We've built a non-functional prototype and will be integrating all the features into it.

Cody Hanson: How long has Subsense been around?

Scott Meek: We were founded about three years ago. Tetiana and I spent the first year reviewing all the technology in the world that could meet our three product requirements. We came up with a plan, got it funded, and raised our seed round about two years ago. We've been in the research phase ever since.

Cody Hanson: Could this technology help treat epilepsy or ADHD?

Scott Meek: 100%. Epilepsy is another great application and will likely be one of our primary targets after Parkinson's. ADHD is definitely a possibility too, though it will be a later application.

Cody Hanson: Your PhD focused on Alzheimer's. Could this apply there as well?

Scott Meek: Yes. One of the most exciting things about this system is that it could create a stream of longitudinal neural activity data that we simply don't have today. I think that data will give us insight into neurodegenerative diseases that have eluded us for decades. It's not BCI exactly, but it's an effect of the BCI industry: it unlocks a massive neural dataset. That's what's missing. We don't have a mechanistic understanding of these diseases, so we're treating them in the dark and creating therapies that are hard to even evaluate. There's a huge opportunity here, and it's something I really want to see.

Cody Hanson: Is there a future for athletic performance enhancement?

Scott Meek: Maybe. It probably falls along the same lines as performance-enhancing drugs, and as an enhancement device for competition it would likely be banned. But there's a lot of potential for therapy: injury recovery, chronic pain, and support for aging athletes. We've seen exoskeletons and e-bikes come to market, and after some initial skepticism, they've allowed people to stay active longer, which is incredible.

Cody Hanson: For healthy users, how would you set up protocols to study efficacy?

Scott Meek: This is an area where we'll need to collaborate very strongly with the FDA. There's never been a technology quite like this that would be over-the-counter, which is ultimately what we're all shooting for as an industry. Setting those guidelines will require industry collaboration with regulators so we're all engineering toward the same goal. Doing it together is ultimately what will make it come out faster.

Cody Hanson: What does fundamental neuroscience research look like on your team? Do you work with external labs?

Scott Meek: It's extremely multidisciplinary. Our research group includes nanoscientists, neuroscientists, electrical, mechanical, and optical engineers, physicists, and computer scientists all working together. We've also collaborated with universities, including our work with UC Santa Cruz. Universities are great at basic science, so we can each focus on what we're strongest at, and that helps move things faster.

Cody Hanson: In a future state, would users be able to monitor their own brain activity, maybe through an app?

Scott Meek: I think of it as the ultimate Fitbit. Those devices became popular because they offered a source of data people didn't have before, and this would be orders of magnitude beyond that. One eventual feature could be health monitoring. Once we understand what the early development of Parkinson's, Alzheimer's, or Huntington's looks like, those signals could show up early and direct people to care.

Cody Hanson: A more philosophical question: could data delivered directly into our experience quietly shape how someone perceives reality, toward flourishing, but also toward whatever the source wants them to believe? How are you thinking about agency, transparency, and safeguards?

Scott Meek: This is the opportunity and the risk of all information technology. It's true of books, movies, and the internet. It's something we need to think about. As I mentioned, one key design feature that helps with safety is that the system is air-gapped. You can take it off, and your reality returns to the reality your senses give you.

Cody Hanson: Do you have a favorite sci-fi story about BCI, and which comes closest to being right?

Scott Meek: Most sci-fi I've seen is far too focused on the dystopian vision. But I think that's actually good. That's a great role for sci-fi to play: calling out potential risks so we can all see them and think about how to mitigate them. As for a favorite, in a strange way, The Matrix. Even though it's entirely dystopian, the idea that you could learn new things quickly and enhance yourself is pretty cool. All the warnings that come with it are important too.

Cody Hanson: Coming from biosensors and nanomaterials, what surprised you most about working on the brain?

Scott Meek: Our ability to collect data from such a complicated system, and how much progress we've made in a short time. Neuroscience is a huge field, and a lot of that existing work has been translatable to our technology, so we're not starting from scratch. Coming from other types of sensors, that was a little surprising, and very encouraging.

Cody Hanson: How do you determine where each patient's nanoparticles are delivered?

Scott Meek: It depends on the application, but you'd start with MRI mapping to get structural data, and potentially functional data with fMRI, then go from there.

Cody Hanson: With all the coverage of AI's exponential progress and potential risks, do you see BCI as a way for humans to integrate with AI so we aren't left behind?

Scott Meek: Absolutely. When I talked about interfacing with the digital world, part of that is AI. It's hard to speculate what our relationship with AI will look like, but AI is the transformative technology of now, and if we're going to have a productive relationship with it, BCI will play a role. I also think that as we collect more brain data and understand our brains better, it will help us build better, more human-centric AI. That's a big challenge for the field: how do we create this powerful technology in a way that serves all of us? Integrating data from people and their thought patterns is one way to do that.

Cody Hanson: Is there anything preventing nanoparticle delivery through the bloodstream to collect data from systems outside the brain?

Scott Meek: No. Particularly for the peripheral nervous system and other sensor platforms, there's application there as well.

Cody Hanson: Last question: many breakthrough technologies show scientific promise but struggle commercially. What milestones should investors and industry observers watch over the next three to five years to separate genuine BCI progress from hype?

Scott Meek: Look at two things: the plan for how the technology scales, and how it performs in animal models. Not 100 patients, not 1,000 patients. What does a million patients look like, and is the technology feasible for that? And how does it perform in animal models, including safety? That's where investors and the general public should look. There can be hype in this field, as in any field. But when I first started looking into this, a lot of the hype actually turned out to be quite true. Much of the work already done is revolutionary, and I'm eager to see how the field grows.

Closing Thoughts

Cody Hanson: Between AI and its convergent impacts across spaces like biotech and BCI, we're living in a very interesting time. As we always say at Singularity, technology seems to move almost like a force of nature. For those of us who aren't doing the research, our job is leadership: helping guide the applications of these technologies, and the people using them, in the best possible direction. Scott, thank you for being here, and thank you to everyone who joined and asked questions.

Scott Meek: Thank you so much, Cody, and thank you, everyone, for joining.

Learn more about Subsense: https://subsense-bci.com/

‍