Unlocking The Brain Through Gaming
For this Deep Tech Dive, I had the pleasure of sitting down with Maciej Rudziński, CEO of Brainattach. A brain-computer interfaces (BCIs) company from Poland focused on creating an EEG headband enhancing gaming by measuring emotions, engagement, and flow. Their proprietary algorithm has been tested on more than 400 people with US patent and patent-pending approval in Europe.
Key Takeaways:
- Gaming could be the future of education and well being of future generations
- The brain is a very plastic device, it’s plug-and-play. If it’s fed information then it will use it. It is possible that if we could connect our brains with passive BCIs, we could teach ourselves new senses i.e. hearing color.
- The ethical implications of any BCI are a major concern. With proper regulations and mitigation, BCIs could make The Social Dilemma seem like a minor problem.
- The myth that human beings only use 10% of their brains isn’t true, but there are unlocked parts of our brain that can be unlocked through rewiring and training certain channels.
“Your mind is a key, and it unlocks the greatest treasure of all.”
― Michael Bassey Johnson
This interview was edited and condensed for clarity.
Tell me about yourself and your background?
I’m from Poland and in 2011 I studied in my college’s first-ever Masters’ program in Neuroscience. I was 1 of the first 10 students in the country, so the program was very strenuous. More comparable to a 5-year Ph.D. I ended up specializing in Dynamical Social Psychology, a niche discipline combining dynamical systems from physics with social psychology. It was very rare as around a few 100 researchers are in this domain.
After my 1st year, I decided to create a publication on the intersection of dynamic social psychology and affective neuroscience. In psychology and usually neuroscience, you assume that our emotions are static, that we jump from one mental state to the next. However, in dynamical psychology, everything is constantly changing. We see similar dynamic outputs from neurological devices like an electroencephalogram (EEG) or functional magnetic resonance imaging (fMRI) as they change from one state to the other. My research suggests that our emotions are in constant flux on many timescales from part of a second to years. Even if someone believes they can feel the same emotion, their actual feelings are always evolving.
How did your research with emotions bring you to creating Brainattach?
I’ve spent the last 12 years watching how people react from moment to moment, to games, ads, websites, etc. My father funded my first EEG and eventually got sucked into psychology. We’re both fascinated by the animalistic emotions confined within people. Since this study is so unique, we found it possible to file a patent on our distinctive way to measure emotions.
Some time ago, I started a neuromarketing company, named Braintracking. We were testing the emotional reaction of customers to ads and websites, so I could finance our scientific endeavors. After a few years, we learned a lot about how people interact and feel, which led us to games because they evoke the highest emotions that are easiest to interpret.
I'm testing today
audio-bio-feedback
based on my emotions and arousalIt's fascinating and creepy!
for start, I know when I drift off
I'm starting to catch my involuntary reactions to eg unexpected sounds
and I've learned that my wife test of my reactions ruined my scale 😅 pic.twitter.com/jPWm4pLzsE— Rudzinski Maciej (@rudzinskimaciej) March 3, 2021
With all this data we conducted a comprehensive study on games with a game developer. Then, we recorded avoid-approach reactions (BIS-BAS) at ~2X per second. These reactions are similar to positive-negative emotions but from a perspective of animal survival. As we perceived the dynamics of these situations we were able to witness flow, the “Holy Grail” of game development. A mental state in which a gamer is fully immersed in what they are doing.
From all the data collected with internal game logs and interviews executed, we realized how hard it is to properly research games, as we drifted toward Big Data methods, but also learned of all the possibilities. This gave us the conviction that we have something new and useful. After we secured a patent application, we have started searching for a partner to license our application and create a wearable EEG band. We started reaching out to EEG companies. But being from Poland we were unable to even receive a response from anybody. Thus, we founded Brainattach to build our own EEG device.
What is the mission of BrainAttach and where are you with development?
It was initially just for fun. An opportunity to continue learning about the brain, how emotions and attention works. But when Google launched Google Glass, it struck us that we could make a product from our studies. Also, since I have a struggling memory (which created my interest in neuroscience), the idea that I could make a device that could improve memory was enticing. With few internal breakthroughs in game research, large improvements in the quality and robustness of our indicators, we were able to make a BCI device that would allow us to steer games with the mind. This device can create a new market of fully adaptive games that would understand the user in real-time and work toward maintaining his flow in the game.
On the hardware side, so far we have completed construction and design for the headband, due to a European Union (EU) grant. We have internally developed the software, firmware, and encryption. We are at the last electronics revision and plan to start selling the device in a few months.
What types of BCIs are there and how do they work?
There are many types of BCIs but the most common ones are:
- EEG devices and have been around for 90 years. They can distinguish between the electrical levels on an electrode, check which side of a brain is more active, see what type of activity is the brain undertaking, measure how well memories are working, how good someone is sleeping, etc.
- Vestibulo-collic reflexes (VCR) devices are usually used for meditation. They help can help with concentration and providing instructions to a computer.
- Specific absorption rate (SAR) devices are used to help people write on a computer but they aren’t very efficient right now. They are very noisy and difficult to use, but they can give someone access to new kinds of information. For people that can focus, they’re very useful, since even the smallest amount of information is enough for them to use such a device.
- fMRI devices are more powerful than EEG’s but slow and size of a room. In more colorful research, they can distinguish what someone is dreaming [as long as it’s something recorded before].
- Functional near-infrared spectroscopy (fNIRS), sibling to the fMRI, can only read from the cortex in a less precise manner but can be worn on the head.
- Magnetoencephalography (MEG) devices can provide signals with higher resolution than EEGs. A neurotech company named Kernel has built these kinds of BCIs, named Kernal Flux and Kernal Flow.
BCIs aren’t commercial yet as they are really expensive but prices are beginning to drop as this industry gains traction. A lot more BCIs are on the horizon as we see neurotech startups such as Elon Musk’s Neuralink and Muse continue to make progress.
So what did you think of Neuralink’s demonstration of the monkey playing pong with his mind?
Neuralink wasn’t introducing new neuroscientific advancements. Some of the neuroscientist community wasn’t as impressed because these scientific advancements took place 20 years ago but that wasn’t the point. What was impressive, is the engineering and technological progress demonstrated. They pushed BCIs to a new scale that can encourage new advancements in this space. They didn’t show the extent of data that can be pushed into the brain, but it was novel.
It’s important to realize that the smallest amount of new information from their device can change what humans are capable of. The brain is a very plastic device, it’s plug-and-play, this can be shown from brain experiments that didn’t include any BCIs. Neuroscientist, David Eagleman, used a Versatile Extra-Sensory Transducer (VEST) to show that human sensory perception can go beyond sight, sound, touch, smell, and taste. For instance, being able to “hear” color. If BCIs could feed electrical signals through unusual channels of our brain, we could be learning new senses in days. The nice part about our brain is that if you give it some useful information, it will integrate it with other sensors and make use of that. Even with no new research, this is could open corridors to our brain that have never been used before.
About Elon Musks’ talk on using Neuralink to read consciousness or changing our memories, I’m not sure what that is supposed to mean because we don’t know what consciousness really is.
Do you believe we can only use a small portion of our brains?
No, that’s so not true. That myth comes from the fact that our neurons fire only once every time. There are hidden potential/areas within our brains being unused that can be rewired and trained in a better way, but it’s not an overwhelming gap of 90%.
Some time ago, me and my friends studied young and old people with low to high attention spans. We found that faster people have better performance when signals are clear and the brain wiring is better organized. Your brain has a higher synchronization, lower noise, and this can be improved.
How do you use Brainattach to train people’s brains with games?
We use our EEG device to record how someone’s brain is behaving when engaging in an activity. From my experience, people with short attention spans play much better in first-person shooter (FPS) games while those with longer attention spans, like me, do well with strategy. When attention spans oscillate, we can modulate the task being performed, so it correlates with brain activity. This way, people can learn much faster. Also, the software gets better at checking how hard a task is for someone, so it can push them further if necessary.
We began to build games because only in games do people have full information about what they are doing. People can also anticipate what their opponent will do. Brainattach can automatically change the difficulty of the game in a way that suits somebody’s attention span. If the game has a short attention span, the game will attack in the moments that the gamer can pay attention to. We can also show new content when somebody is bored, adjust the difficulty, remove obstacles if needed, and guide generated content for endless fully personalized gaming. We also can do more trivial things like adjust accuracy or spell power based on focus and emotions. This way we can keep the brain at the highest possible level of activity to optimize human performance.
What are the ethical implications of BCIs and how do you think we should navigate through this?
We are a young company without a product in the market, yet we already have an ethics commission. There’s just so much that the government won’t care about until something bad happen and past experiments worried us of our BCI’s capabilities.
The problem with synthetic data, like genetics, is that data can be extracted even after the recording is complete. For example, let’s say that there was a company that measured EEG to recommend music to a person. After 10- 20 years, data extraction gets easier for the company. They would be able to learn exactly how that person reacts to rewards, when and what ads to show them, what do they hate, what sex do they prefer, etc. We don’t want anything like this to be possible so we try our best to anonymize or destroy parts of the data recorded. But managing risk is expensive. Even though we don’t have money for it we still spend some time mitigating risks.
The difference between what we do and others is that we measure 2D emotions. Just from a 5-minute recording, we can determine personalities. We can know how people react to pleasant/unpleasant things, how much of a gambler they are, and that’s just the beginning. With more data recorded, we can see how their personality is changing, and when it will be more vulnerable, things people don’t like to share.
When we started our investor’s presentations we asked somebody to demonstrate our device during the meeting. We could see how he was reacting when different people spoke or when different ideas were proposed. When he spoke, we could see if what he said came from his memory, without thinking even, because when you are speaking out of memory, your brain is not as activated as when you are creating a new idea. With more data and computing power, information extraction would become easy. Afterward, we decided that we would never do anything like that again.
Now we only have 2 large indicators, for notions, tensions, and reason. We have 27 other candidates for indicators that we know are stable and useful but we didn’t map them out yet because our customers were struggling to integrate just 2 indicators. We moved to games because in comparison to other BCI use cases they’re a lot more fun. Still, this kind of tech is can be addictive so we need to lower the precision of the indicators as we scale.
What do you think about the future of education, and where does gaming fit in this?
Usually, if you implement any kind of fun activity into something, kids are more bound to learn it regardless of complexity. Which would make one wonder if we can teach difficult concepts at a much younger age? It would be very beneficial to learn complex concepts such as math and programming in adolescence. Our brains are more plastic at that age as compared to post-adolescence (college and beyond). Brainattach can help with constantly keeping your brain in high gear to learn these ideas at a young age.
Also, when people are that young, they are bad at describing and controlling their own emotions. If you tie the experience of games to their emotions, like shooting more accurately when feeling concentrated, children will react faster and learn how to control their emotions better. This is helpful because if they can control their emotions, they can be better learners.
Designs for future iterations of the EEG Headband.
Catch the latest in my Deep Tech Dive series where I interview:
- Nicholas Nadeau, Project Director & General Manager at Halodi Robotics. A Robotics-as-a-Service (RAAS) company working on bringing humanoid robots to everyone.
- Alishba Imran, Co-Founder of Voltx and Machine Learning (ML) Developer at Hanson Robotics. Voltx is an ML company developing software to accelerate research and development testing of energy storage devices, including batteries and supercapacitors (SC). And Hanson Robotics is an AI and robotics company dedicated to creating socially intelligent machines that enrich the quality of our lives.
- Jai Malik, General Partner at Countdown Capital. A Deep Tech VC firm focused on investing and providing in-house support for teams committed to deep philosophical and ethical thinking on top of breakthrough engineering.
- Norris Tie, Co-Founder & CEO of Exosonic, Inc. A supersonic aviation company developing a new standard for travel with their low-boom supersonic jet capable of cruising at Mach 1.8 overland and comfortably carrying up to 70 passengers.
- Drew Bellcock, CTO of Pipedream Labs. A hyperlogistics robotic company building a network of underground tubes that offers near-instantaneous delivery of objects to and from homes and businesses.
Hey, thanks for reading! I’m Andrew, CTO of Sencha Credit, and a Hard Tech Analyst at Cantos Ventures.
I understand that innovation is growing exponentially, but I am worried it takes too long for organizations to accept frontier technology. I aim to demystify the complexity of cutting-edge technologies for the average person. I encourage entrepreneurship while exhibiting my excitement for accelerating technology, hoping that you may feel empowered to engage & build the future.
Follow me on Twitter and/or LinkedIn and please feel free to reach out to me anytime.
