“In the future you’ll be able to save and replay memories”
“Just by thinking, you can type and control a phone”
“It’s kind of like Fitbit in your skull with tiny wires”
These quotes are from Elon Musk during the recent Neuralink update presentation. If you missed it, here’s a stream. The gist of his pitch was that big innovation in neurotech was coming.
We started Eno four years ago with the goal of bringing neurotech to the masses, to help you understand your work habits and improve your focus. We were excited by what we had seen: neurotechnology was a nascent technology with proven medical benefits and had recently been made more accessible by improvements in electronics and data science.
Shortly after, we wrote our first blog post, titled “A History of Modern EEG” There, we explained Eno’s core thesis: brain-sensing technologies are moving out of the lab and into consumer devices.
Since then, the neurotech industry has changed dramatically. Dozens of consumer neurotech startups across a variety of applications have popped up all over the world. Nerve-sensing has reached a tipping point, as CTRL labs was acquired by Facebook for $500 million. Medical neurotech companies have found a niche in neuromarketing and have grown dramatically. Lastly, this new category of invasive neurotech has come to light, led by Neuralink’s incredible marketing.
As we’re shipping our first product, I think it’s time to reflect: where are we headed, and what is the future of neurotech?
The history of neurotech, going back as far as 50 years, has been segmented in roughly 3 categories: medical, invasive, and consumer. Let’s take a look at each.
Medical neurotech remains the largest category to this day. Herein lies the full-cap, 64-electrode EEGs and million dollar fMRI machines that researchers, neurologists, and clinicians use to measure and influence the brain under controlled conditions.
While it doesn’t get as much media attention, the medical neurotechnology sector is a $9 billion industry. It powers every lab, every neurological clinic, and every neurofeedback center in the world with the technology they need. It wouldn’t be overstated to claim that this sector is responsible for the majority of innovation in understanding the brain in the last few decades.
This industry is concerned with one thing: accuracy. Every year, they strive to build more accurate systems than the year before. Reducing the noise, increasing the resolution, sampling faster, measuring longer – whatever it takes to improve the accuracy of the readings. This is why medical neurotech usually operates in labs and clinics: controlled conditions, where extraneous effects can be eliminated, mitigated, and quantified alongside your readings.
Essentially: medical neurotech wants to better understand the brain.
So, what changes have the last few years brought to medical neurotech? Frankly, not many.
Medical neurotech has been pushing the boundaries of science for decades – the mechanical, electrical, and algorithmic innovations always start there. Even the most recent innovations in AI and deep learning were being used on medical EEG data in academia prior to us starting Eno.
This is not to say there’s been no progress. Every year, the tools get slightly better, and the insights slightly clearer. But short of discovering an entirely new brain-sensing modality (fNIRS anyone?), we are unlikely to see a dramatic shift in this arena very quickly.
Instead, medical neurotech will continue to do what it does best: serve the massive community of neuroscientists, neurologists, doctors, and clinicians that use these tools every day to measure and understand the patients they serve.
Invasive neurotech is a branch of the medical neurotech tree. It follows the simple principle that if you want to measure something, you should get as close as possible to it.
Invasive neurotech involves drilling a hole into a patient’s skull in order to implant a series of electrodes or stimulating probes (or both). These are then used to replicate the same experiments as medical neurotech, but are dramatically more accurate..
This improvement is because it overcomes the biggest issue when measuring the brain: your skull is in the way. By implanting small electrodes directly on the brain, you can measure the activity of ~1,000 neurons firing together in an isolated area versus >1,000,000 neurons firing in a cluster.
The same is true with stimulation technologies. While tDCS and TMS have shown promise in some specific applications (depression being a major one), they lack the precision necessary to do the kind of things that Elon Musk brags about. For that kind of resolution, you need to bypass the skull and directly interface with the tissue.
This has been a prominent area of study for a while. Deep brain stimulation is frequently used to treat Parkinson’s Disease, among others. Invasive EEG is also frequently used in labs to help understand the impact of very specific areas of the brain. All those experiments you hear about where monkeys control computers with their brain – this is where they come from.
What’s most interesting is that the limitations in this field are wide-ranging, and are primarily related to the engineering, not the neuroscience.
At the top of the list of issues to overcome is that it’s incredibly difficult to put objects inside the body without infection or rejection. Brain surgery is also lengthy, expensive, and not without risk. Next comes the short life of batteries, the unreliability of wireless communication, and the necessity of a removal process. Not to mention, convincing people to get the surgery in the first place.
Unlike medical neurotech, invasive neurotech is not yet pushing the limits of science – only the limits of engineering. This is why Elon Musk feels that Neuralink is possible: engineering problems can often be overcome.
This will likely take years (read: decades). But it will come, one step at a time, starting with high-value medical applications. Expect to hear more about this in the news over the next years, as the factors limiting its potential start to crumble.
But don’t expect to wear one for quite a while.
This leads us to the consumer category. Consumer is the newest branch of neurotech, which evolved directly from medical applications. As electronics became cheaper and data science more powerful, it became possible to replicate medical neurotech’s results in consumer-friendly form factors.
The consumer neurotech space is currently experiencing a massive wave of innovation. A recent decrease in the cost of ultra-high impedance front end amplifiers has made it possible to measure clinical-grade signals with dry electrodes. Deep learning applied to a massive scale of data has enabled applications to make measurements that used to require $10,000 equipment. Last, form factors like headphones, glasses, headbands, and helmets have enabled niche applications across industries.
As of today, the consumer market is generally segmented into 4 categories:
First is the OG consumer niche: stress, anxiety, and neurofeedback. Pioneered by Muse, they realized that they can measure indicators of stress and calm using a few sensors on the front of the head. Muse’s meditation application – and the numerous subsequent spin-offs – continues to add value to hundreds of thousands of customers.
Second is sleep. Since EEG is the go-to technology in sleep-tracking, this was a natural fit. Companies like Dreem, and even Philips, have released products that help track and influence your sleep more accurately than a wrist-worn wearable. This industry will likely continue to thrive as these companies combine other sleep-influencing technologies with their biosensing.
Third is attention and focus. Attention research is predominantly done via EEG, so the consumerization of this technology was quite appealing. This is where Eno plays: we measure indicators of focus (motivation, fatigue, mind wandering, etc) in order to help you understand your work habits and improve your focus while you work. The massive amount of data you can measure during the day (how often do you wear headphones?) allows us to extract more value than was previously possible with consumer neurotech.
Fourth is what I call pro-sumer. These are cost-accessible research tools, used by hobbyists and academics to study the brain without requiring the investment of a $20,000 EEG kit. Emotiv pioneered this space, releasing a variety of products across a range of prices. As a result of these tools – and the community using them – neurotech has become a commonplace hobby project for developers and hackers across the world.
In the last few years, the number of applications in consumer neurotech has skyrocketed as companies try to learn what is possible and valuable when using these technologies. I believe that each of the four categories above will continue to thrive, as tracking your brain becomes as commonplace as tracking your heart, exercise, or nutrition.
I also believe that the consumer space is where the biggest innovation will occur because it incorporates what invasive and medical neurotech can’t: massive scale.
Because, in the end, neurotech is simply another biosensing technology. What truly adds value is the understanding that data generates and the applications that it enables. And, unlike in medical and invasive, this is an area that engineers understand very, very well.
This is why Eno exists: to use these new technologies to enable you to understand your work habits and improve your focus.
So keep an eye out – some big innovation in neurotech is coming.