State of the Art Biological Computing: An Interview with Dr. Ewelina Kurtys

Introduction

Charles Humble: Hello and welcome to this mini series for Goto Unscripted called The State of the Art, where we are exploring emerging trends in computing. I'm your host, Charles Humble. I have 30 years experience as a programmer, architect, and CTO, and I'm currently working mainly as a freelance consultant and advisor, as well as a tech journalist, editor and podcaster.

Interest and investment in AI research has never been higher, but current artificial intelligence systems have adopted a narrow approach which relies on extrinsic learning mechanisms, gradient descent and its variants serving as the primary means of model optimization. Intelligent systems derived from biological brains, however, display intrinsic learning through integrated processes. Since biological brains also operate with power and data efficiencies which are otherwise unparalleled and have the capacity for rapid real time adaptation, it may be beneficial to reconsider neural systems in the current search for AI.

I'm joined today by someone who's been exploring this area in detail. Dr. Ewelina Kurtys is a scientist turned entrepreneur pushing the frontier of bio-inspired computing. She graduated in pharmacy and biotechnology in Poland, pursued a PhD in neuroscience in the Netherlands, where she specialized in brain imaging. She continued to work in this field as a postdoc at King's College in London. She has since left academia for industry and is currently strategy advisor to FinalSpark. FinalSpark is one of a small handful of companies looking to move AI from digital processes to biological processes and design systems built from living neurons. I'm very much looking forward to this conversation. I'd like to welcome you to the show.

Ewelina Kurtys: Thank you so much. Very happy to be here.

From Academia to Biological Computing

Charles Humble: It's wonderful to have you on. Maybe let's start a little bit with your career journey. What prompted you to move from academia to business and this focus on small startups?

Ewelina Kurtys: That was trial and error. I always wanted to leave academia because I was curious what is outside. What else can you do in life except research? I was also compelled because I think that in industry you have more options. In academia, I believe there is only one way to be successful, which means to become a professor. But in industry, I think there are many ways to become successful depending on what you want and what your personality is.

I was always intrigued by that. Initially, I thought the best way is to work in a big company. Even then, I realized that to work in a big company, you have to fit. Then I realized that's not for me. By coincidence, by networking in London at the time, which is a perfect place for exploration, some people advised me to apply to a small company which was doing medical imaging. That was also something I was doing during my research, mainly on brain imaging.

I thought it's a good starting point because there would be at least one thing which I understand very well. That indeed gave me an advantage to start because everything else was new about working in industry. On the technical side, I had some edge in that I understood medical imaging, which is a highly technical service.

This is how I started. But then I realized that I prefer much more sales and business development and marketing than being a scientist in industry. I think a scientist in industry is totally different than a scientist in academia because in academia, you have a lot of freedom to do what you want more or less. You have freedom, maybe not a lot of money, but you always have freedom. In industry, you have to stick to the agenda, and you cannot follow your curiosity so much. You have to do a lot of stuff which is sometimes highly technical.

I knew I wouldn't want to be a scientist. Then, trial and error, trying different things. I also tried some project management. I realized that for me, the best job is when I talk to people, when I have interaction, when I can explain something, and when I can explore. That was very important for me. I think for everyone it's different. But that was my move from academia to industry.

Charles Humble: Is there anything that links the two areas for you? Are there lessons from being an academic researcher that apply to what you're doing now?

Ewelina Kurtys: Absolutely. There are a lot of common links. There are general skills which you get in any job, like reasoning or problem solving or resilience. In both cases, when you work in a small company and when you work in academia, you have a lot of uncertainty. It's not sure what you do and a lot of things can change. I think there are a lot of things in common.

I think in general when you do a PhD, after when you move to industry, you can do practically any job you want. The rest is just your character. More or less, you can do everything. Maybe except some highly specific technical jobs. You can do anything. The rest is just your character and your personal preferences.

Charles Humble: What drew you specifically to FinalSpark and this idea of biological computing?

Ewelina Kurtys: The first thing, I always look at the people, not so much the ideas because ideas can be good or bad. It doesn't really matter, but it's important the people. What I discovered is that in deep tech, you have nice people to work with. I have the impression there is such a tendency.

I had the chance to meet the founders of FinalSpark at one of the conferences in London, which at the time was a real eye opener for me. That London AI Summit, I think now it's getting a little bit boring because every AI discussion is large language models which are being sold as a solution for everything. But two years ago before LLMs, we had big variety. I came from the imaging industry, so image analysis is something what you can do with AI. But then at the summit, I realized that there are many other things you can do and it doesn't have to be medical imaging. It doesn't have to be imaging at all.

That was a big revolution for me. I became interested in FinalSpark because the founders have a nice culture, and indeed they have a very nice team, very nice atmosphere. Everyone is happy there. I think it's a good place to work, although I'm always remote, which I also enjoy.

That happened to be the perfect match, remote working for my personality. I really enjoy working from home. I can discipline myself. I can do very well by myself. It's not for everyone. Some people feel better in the office. I think it's very important to remember there is not one solution for everyone, but at least for me, it was perfect because when I met them, I was in the UK, in England. I'm still here and they are in Switzerland. It seems to be perfect because I can work remotely. I never have to go to the office.

I think it's mostly about the people and the ideas. Every idea you can execute good or bad, but a lot depends on the daily basis and the atmosphere you have in the lab, in the team.

The Vision: Building Computers from Living Neurons

Charles Humble: I said in the introduction that the goal for FinalSpark is to build computers using living neurons for AI. Is this a specialized device like a GPU or is this something more ambitious, like a general purpose computing device?

Ewelina Kurtys: It's definitely general purpose computing and is a very ambitious goal. We expect to be ready with a computer in ten years, which is an estimation, because with such a project, you can never have a precise timeline. But we estimate it will take ten years. Maybe it could be faster because of LLMs.

We come back to the same topic like in every AI discussion today. But it's true that they can accelerate research a lot. We can already see this today in our R&D, that we can set up some automated agents who can write Python scripts for us. We see this already in real life that it can accelerate R&D a lot. Who knows, maybe our timeline will be faster, but for now we estimate ten years.

It's very ambitious, very challenging. We are only six people, a very small team, because we are self-funded by the founders, Fred Jordan and Martin Kutter. These are two Swiss entrepreneurs, and they already have another company. This is why they have money for FinalSpark.

To really reach the goal in a reasonable time, like ten years, we need to extend our investments. We are seeking 50 million Swiss francs, which is a lot of money. But this is the money which would make a difference for our budget today. The main objective for the investment is to increase the team, because, which is not a surprise in such a project in deep tech, the success depends on the people and on how many people work on the difficult problem.

We are very ambitious, but it's also challenging and a long term idea.

Charles Humble: When I was thinking about this, it seems to me there are two mutually inclusive possibilities. You could use biological cells as the substrate for information processing and intelligence. The other option is you could develop new algorithms which are inspired by those biological systems. Why specifically are you interested in the first option? Why are you specifically interested in using biological cells for this problem?

Ewelina Kurtys: Because we believe this will be a real revolution. You are right, a lot of algorithms can be optimized and that's already happening. We see in the industry people are doing this all the time. You can also optimize the hardware. People try to optimize digital hardware. They even sometimes try to make analog hardware, so that's a bit closer to the brain. This can lead to even bigger gains of energy.

But we believe that nothing could be so revolutionary like living neurons, because living neurons are one million times more energy efficient than digital software and hardware we have today. All these improvements which we are doing today are little steps, but living neurons would be a huge step forward to energy efficiency.

Charles Humble: That's really interesting. I mentioned it in the introduction because much of my consulting work is focused on ways that we might reduce computing's carbon footprint. The energy efficiency of neurons is astonishing. Can you give us a sense? I appreciate it's very early days, but can you give us a sense of what you're expecting to see or what you can see in terms of that efficiency?

Ewelina Kurtys: Well, one million times more efficient. That will mean we don't want to change anything for the end user. That's our dream, to just replace the hardware. Basically today you use GPT on digital hardware. Tomorrow you can use it on living neurons. The only difference for you should be the price.

This is something which is very important, although many people don't appreciate this yet. Today we don't realize how expensive AI is. Why is this? Because running large language models or any other sophisticated AI systems is really expensive. Today, a lot of companies are selling this cheaper than it really costs because they want to increase adoption.

For example, OpenAI, who created GPT, they lost already 19 billion dollars. What does it mean? That means that it is not as cheap as it should be to make this company able to survive. This is a long term strategy because they want to work for adoption. They want us first to learn how to use it. Then later, once we appreciate the value of systems like GPT or other LLMs, then they can increase the price because we will be ready to pay more.

This is a really important problem which we don't see because of these lower prices. But in the future it can become even more serious. Why? Because every better model is consuming more energy. Also, if we use more and more AI, that means the overall demand will be higher.

Today, maybe you just need a little subscription for $20, which is around the price for now for GPT. But maybe tomorrow you will automate much more things, much more tasks, because we will learn how to do this. Then even this will increase the price. So the usage is growing and the real price of AI is much higher than in reality.

That means that in the future we can expect very high prices unless some solution would be found to how to make it cheaper. I'm sure that there will be a lot of small incremental changes on the way because we see this already. There's a lot of papers on software and hardware optimization. We will see a lot of small percentage optimization, but living neurons will bring a revolution that will be one million times more energy efficient. That is a huge difference.

The difference will be in the price. That's what we want people to experience in the future. We don't expect it to be some different models, because our dream is to match the performance of digital in some tasks. Not everything. Because living neurons are good for some tasks, like maybe generative AI. We conclude this based on how a human brain is working.

We don't expect that neurons would replace digital in everything. But in some things it will replace and will be much cheaper. That will be the result because everyone is asking for use cases. We imagine generative AI being much cheaper.

Charles Humble: I think it's worth saying the ways that we currently run large language models are incredibly inefficient. There are absolutely ways that we can use to bring the efficiency down and the cost down. There are a lot of other things that we're not doing that we can do. But I also completely agree that what you're talking about, if it all works out, could be a huge change in terms of AI capabilities and efficiency, which I do think is really interesting.

I appreciate you're at very early stages, but can you give us a sense of what your performance expectations are? I would imagine biological systems are going to be an awful lot slower than either silicon or quantum based systems. Can you give us a sense of what the performance is like? Do we know yet?

Ewelina Kurtys: Absolutely. You are right. The performance will be different. We can make some estimation based on how we see the human brain. We know that the human brain is very efficient in complex tasks, but we know also it's not as fast as you said. It also has less memory.

Your brain will never remember all the books you have read by heart, every page. But a computer can do this, even much more books. There is significant difference. That's why we believe living neurons are good for generative AI, for creating ideas, for solving complex problems very efficiently. But not necessarily for high speed computing. For sure not.

We can expect for sure it will be slower, but how slow? It's difficult to say, because today we have only human brains, which are quite small, and what we want to build as a computer would actually be a huge bio server, which maybe will be even 100 meters long, all from living neurons. It's hard to predict how fast this structure will be. But we know for sure it would be very efficient.

Charles Humble: I had a related question, which is a bit of a mystery question, but do our brains work in parallel, or would you expect biological computers to be, in a sense, like a single threaded process or handling one task in a linear way? How does that even work?

Ewelina Kurtys: We definitely have parallel computing in the brain. I actually wrote a blog article about this, about the difference between brain and digital. It's on my LinkedIn. If someone enters my profile, it's one of the featured articles, so it's easy to find. Definitely parallel computing is also one of the reasons why brain and living neurons are so energy efficient.

Technical Challenges and Breakthroughs

Charles Humble: How much do we understand about how neurons encode information? I thought that was something we didn't really understand at all.

Ewelina Kurtys: That's the challenge of this project, I would say. We know quite a lot about how neurons process information and about how the signals are processed by living neurons. But we don't really know a lot about the encoding. That's really the challenge, because for sure, encoding is different.

We know that in computers you have zeros and ones. In the brain you have time and space. It matters when and at which point a neuron is active, electrically active. This is totally different encoding, but we don't really know what it means.

In practice, nobody can read our mind today. We can detect the signals from the head, but we don't really know what they mean. We can make some correlations, but we cannot really decode perfectly the signals. That's actually the biggest challenge.

Charles Humble: I think brains are also plastic, but basically they change over time. I'm presuming when you're running experiments in your lab, those experiments can produce different results. Identical experiments with the same set of cells and neurons can produce different results on different days.

Ewelina Kurtys: You're right. Plasticity is a good thing because it allows us adaptive learning, which is fantastic. We can learn by experience step by step every day. But it's also a challenge for experiments. Because the system in general, neurons are not a stable system in general. That means, as you said, that today experiments can give different results. The model can do something different. This is just another variability in our research, which increases the challenge.

Charles Humble: I'm presuming that when you're running your experiments, you have to have an awful lot of automation because you've got so many unknowns and you need the experiments themselves to be repeatable, right?

Ewelina Kurtys: Absolutely. We never report any results if we haven't repeated this several times because we get a lot of spectacular results which were just one time or maybe two times. That's not enough.

For example, we managed to store one bit of information. That's something what we have done many times. We know it's reproducible. We were basically able to change the state of the neurons in a manner of two states, two different states, zero and one, which can be mathematically calculated.

If we say we've done something, that means we've done this really many times, reproducibly. That's actually the biggest challenge. That's why we don't report many of our results, because they are hard to reproduce.

Charles Humble: I had a conversation a while back with Dr. Barbara Oakley. A lot of neuroscience research papers are not actually reproducible at all. Outside of the fact that you have something that is reproducible, it is itself continuously changing. So is that the main way that you validate your results or are there other forms of validation that you do?

Ewelina Kurtys: Reproducibility, I would say, is important. We also have a lot of methods on how to process the signals. We should make sure that what we get is real. For example, in biological signals, you have a lot of noise. That's well known. That's why you have to filter out the noise. We do a lot of things, and repeatability is, I think, the most important criteria to claim that we have done something right.

Charles Humble: Is there any consensus among neuroscientists as to whether our brains are deterministic or non-deterministic?

Ewelina Kurtys: This is a very interesting question, which is definitely, I think, beyond the expertise of FinalSpark. But I read recently a very nice book about this, which is called "Determined." I think it was written by Robert Sapolsky. I have it on my shelf, and I think it's... I have to say, this book helped me to change my way of thinking, which is rare because I'm somewhat stubborn and I hate that I cannot find the reason to refute the arguments in this book.

So far I have nothing to say. I have no way to justify why it's not true. Basically what the book says, and what is consistent with current neuroscience, is that everything is actually deterministic. It's just that it's so complex that we don't see the cause and effect, which is absolutely normal.

When you look at the human brain, there are a lot of obvious things, like, for example, our intelligence is shaped also by genetics. I mean, if you are born, for example, with Down syndrome, you can never have an IQ of 150. This is maybe the most blunt and obvious example, but there are many other subtle examples of how our intelligence or our brain capacity is determined by genetics.

Of course, upbringing also matters because there is this vivid discussion always in biology, whether it's more environment or the genes. I think currently it's quite agreed that it's an interplay between genetics and environment. You know, one cannot go without another.

Actually, environment we cannot choose either. You don't choose where you're born and what is your environment and you don't choose your genetics. If you really think about this deeply, everything goes from every moment. There is some molecule. If you have this molecule, it's because of your expression of genes. If you have these genes, it's because of the genetic material you got. Actually everything is determined.

I have to say that I believe... no, maybe I actually believe is not a good word in science. But I would say I'm convinced today that indeed everything is determined, definitely determined. Although it is an open discussion and many people... it's connected to the discussion about free will, which we don't want to accept that we don't have it.

But actually, if you follow the reasoning of our current neuroscience, there is no free will. We are just determined by the circumstances which I think is very interesting. I think, unfortunately, it's true. I cannot find the reason why it wouldn't be true.

I think also in the brain, everything is determined, definitely. I don't think it's random. It's just very complex. The same also with what we have in our lab. That's why I believe we can build a computer, because it's all controllable. We just have to find a way how to do this.

I think today, it's determined. It's actually not only neurons, but you could say the same about every living cell because what I said about neurons, you can actually say about any kind of living cell, even one cell. That's what I believe.

Charles Humble: I want to ask one more background question. My understanding, and this is not my topic at all, but my understanding is that neurons in adult brains live more or less for our lifetimes or even more, because we don't replace them, right?

Ewelina Kurtys: Absolutely true. Although we have some new neurons growing also.

Charles Humble: Right. So with your neurons, how long can you keep them alive currently?

Ewelina Kurtys: This is a very good question. Very important, practical. Currently around three months, although we recently had a success of seven months in our lab. But our client generally is around three months, which is very long for lab conditions. However, it's not very long for a neuron in what is possible because as you mentioned, they can live even a hundred years in our brains.

It's all about how to make the condition of neurons as close as possible to the natural. This is still a challenge because everything in the lab... for example, we put neurons on electrodes so that we can send them electrical signals and we can receive responses. That's already stressful for them because it's a contact with artificial material.

The same is about any kind of glass you use, any kind of plastic material used to keep your neurons. These are all the challenges of why neurons don't live as long as in our bodies. But we believe this is also an engineering challenge. We believe we can solve it. Today, the biggest challenge is learning in vitro. How to control the neurons in vitro, how to make them perform some tasks.

For this, three months is enough for experimenting. Then once we solve this problem of the algorithm, which is very long and challenging, we can think about scaling. We believe it's possible to solve.

Charles Humble: Presumably then you're keeping the cells in conditions that are very similar to a human brain. Same physiological conditions, pH, temperature, nutrients, fluids, whatever.

Ewelina Kurtys: Absolutely. That's actually what everyone is doing in the lab everywhere. When you keep living neurons or living cells of any kind, any kind of cells, anything you want to do is try to keep the conditions as close as possible to nature. pH and temperature. They always have to be in liquid. They have to get some nutrients to survive, like every living cell.

Charles Humble: Can you tell us about the prototypes you've built and how do they actually work? How are they constructed?

Ewelina Kurtys: Currently we've built a prototype. We have little 3D structures of neurons, around 10,000 neurons each. We put them on electrodes and we can send them electrical signals and we can measure responses. This is the prototype of the computer.

What is very special is that this prototype is multi-user. We have universities from all over the world who are using it. We also have industry clients, people who pay us for access to the lab so that they can do fundamental research on signal processing on neurons. This is very fundamental. We cannot yet sell computing power, but we have a very robust and very unique tool for studying biocomputing.

Charles Humble: You're using human stem cells to do this?

Ewelina Kurtys: Yes. Basically we are using stem cells which are derived from human skin. This is the procedure which was awarded the Nobel Prize around 15 years ago. Basically you can sample skin cells, you can produce stem cells from that, and then you can produce any cells you want from this, including neurons.

Charles Humble: When I was doing background reading for this podcast, I was reading on the FinalSpark site about brain organoids. Can you explain what those are?

Ewelina Kurtys: Organoids is actually a technical term for how we call these 3D structures which we keep on electrodes. Organoids is something what is meant to be complex and simply resemble a human brain. However, it is very important it's not the same because we are just using building blocks of the brain. We have nothing to do with the brain.

Some people call organoids "brain on a dish," but I think it's unnecessary marketing which is actually hurting the field because when policy makers start to believe that we really have a brain on a dish, they will start to invent some regulations for that, which I think is wrong because organoids, of course, they are complex. They are 3D structures.

They can have a few types of neurons, different cells, but that doesn't mean they are a brain. It's far, far away because the brain is much more complex. We would never be able to build something so complex as a human brain.

What is an organoid? It's a 3D structure. What is specific in organoids is that you have a few cells, usually a few types of cells, and it somehow can resemble some kind of circuits which we have also in the brain. You can, for example, try to combine different types of cells which you believe could be good for learning.

Charles Humble: And how do they form connections?

Ewelina Kurtys: This is just biology. You put neurons together and they will just connect if they can. This is just pure nature. Of course, totally random. This is the same actually as what our brains started from.

Charles Humble: You said you can basically store one bit of information on a neuron currently, right?

Ewelina Kurtys: Yes. But this is just an example. I wouldn't say this is an indicator of memory. I think it's too far. It's just that we were able to consistently keep two states of the neurons, and in this way we use something called the center of activity.

Basically, neurons are connected to eight electrodes and you can mathematically calculate the center based on the activity which is measured from each electrode. It works the same way like gravity. You have more activity, the center is going towards this point. We were able to move this point, the center of activity. That's what we did.

That's more like a proof of concept that we can do something. But that's not really how I think, I believe it's not how the computer will work.

Charles Humble: Making them active is not neurotransmitters or is it an electrical signal? How does that work?

Ewelina Kurtys: Actually in a neuron, you can divide two types of signals: chemical and electrical signal. In biology it's all chemical. But we just make this distinction, it's easier for us because in the end, in a neuron, everything happens because of chemistry. Also, when you have electrical signals from neurons, in the end it's the ions which are plus and minus and they change location. That's why you have something electrical, what is actually changing the voltage, which you can measure as an electrical signal.

But for simplifying, you can say you have electrical and chemical signals. We use voltage. We have developed... we're always improving. We started with electrical signals. Neurons are always on electrodes. We send them signals and we receive a response.

But now we can also send chemical signals. That means we can, for example, release dopamine, which we can use as a reward during some experiments. That's an example of something we did.

Charles Humble: Can you communicate with neurons? I mean, do we know how to program neurons?

Ewelina Kurtys: Well, when we store one bit of information, that is communication because we send them signals, we receive responses. We always have bidirectional communication. It just doesn't always work as we wanted.

Ethical Considerations and Future Implications

Charles Humble: I'd like to shift and talk a bit about ethics. There are a lot of ethical issues that come out of this for me. The first thing that I was thinking about is how do you get informed consent from donors? It's very difficult, I think, for most of us outside of the field to even imagine what you're doing. How do you go about getting informed consent? Can you get consent?

Ewelina Kurtys: Actually, it's very easy, because this is a process which is already commercialized. There are companies which are selling stem cells which are derived from human skin. That means that everything is already sorted by someone else, and you just buy the stem cells. They are a commercial product, so they are compliant with all the requirements.

Charles Humble: Then I guess there's something else that comes up. It comes up a lot actually in the context of generative AI, which is the lack of arrangements in relation to the ownership and commercialization of results. Is that something that you think about?

Ewelina Kurtys: I'm not sure what you mean by arrangement?

Charles Humble: Certainly in generative AI, a lot of it is built by essentially ingesting the internet without permission to do that. People are... I mean, no one's really making money off it, but in theory, you could be making money off it. There are interesting questions that come up about whether that is okay.

In this case, if you're building something that is profitable from donated stem cells, I'm just wondering what the ownership and commercialization of your results looks like in that context.

Ewelina Kurtys: That's very interesting. I never thought about this. I don't think that people who gave the cells should own the results. I don't think it makes sense because they have no contribution. You could always take from someone else. I don't think so.

I honestly think that sometimes ownership is exaggerated. I understand when you take art pieces, but in this case... this is a question about ownership, not... if you just take biological samples, I don't think that donors should have the revenue because they already sold them.

Basically, if you agree to give your skin, you agree to give also the outcome. I generally think this is fair because it's an honest deal. You give your cells, you get the money, and it's... so I don't think they should claim any ownership on this, but it's a very interesting question. Nobody asked this, so I don't know, maybe one day it will arise, but I think it would be relatively easy to solve.

Charles Humble: You talked about... you used the word stress in the context of cells in lab trials. Can you unpack what you mean by stress? I guess my question really is, can biological systems suffer?

Ewelina Kurtys: Suffer is not... I think suffer is something more complex because we have brains. But stress, biological stress can be observed in anything from a single cell to a complex organism. Stress means any aberration from homeostasis.

For example, if you have living neurons in experiments I did during my PhD, you have living neurons in the lab. You put them in something called LPS, which is lipopolysaccharide, which is a complex name for something which causes inflammation. Then you induce stress. That means they will start to produce pro-inflammatory molecules.

There's a lot of... anything can stress the cells. I wouldn't say it's related to suffering. I think suffering in this case would be... you know, like anthropomorphization, that you believe... but it doesn't mean that it has feelings and they are suffering. But yes, definitely in every biology, in every biological item or organism, stress occurs. Also in plants, in anything what is living. Actually you could call stress also fo a dead matter, but that's outside my field.

Charles Humble: It's interesting to me because instinctively... I mean, I use the word suffer because there's an English philosopher, Jeremy Bentham, who argued with respect to the moral status of animals, the question is not can they reason or can they talk, but can they suffer?

Ewelina Kurtys: Absolutely. This is very interesting. This line of thinking is, I think, causing a lot of troubles because people associate human feeling with everything. You know, we have to respect philosophers. They spend a lot of time thinking about this. But yes, you had ideas, but actually it causes a lot of problems, I think, sometimes in, especially in the perception of bioethics, because people see human rights in our living neurons, which I don't think is right.

Charles Humble: Maybe there's another question. Given our limited understanding of consciousness, how do you quantify when a neural system is displaying a trait that maybe requires moral attention?

Ewelina Kurtys: You absolutely cannot. I cannot do this because there is no scientific way to measure consciousness. Actually, philosophically, I heard from philosophers that the only consciousness you can be sure about is your own, because you cannot measure this even in other people. But it's an open question. For philosophy it's a very difficult question.

This is why I should mention this. When we talk about philosophy, we actually reach out to philosophers because we ourselves are not competent. We can elaborate, but it's not really competent elaboration about philosophy. We reach out to real philosophers to help us to answer all these questions.

Although I can tell you, consciousness is more for the general public. For real philosophers, the more interesting topic, I think, is about the relationship between technology and us, how it's changing. But consciousness is also an open topic. When you look at the literature, you can see that there are a hundred years of research and there are all spectrums of opinions from that it doesn't exist to that it is everywhere.

It's very open and there's no way to measure scientifically. That's why it's so controversial. But we hope philosophers can help us with that.

Charles Humble: That's super interesting. Thank you. This is slightly off topic, but I was just thinking... obviously our understanding of our own brain and how it works is pretty limited still. I mean, we understand bits of it, but there's a lot we don't understand. Do you think the work that you're doing might inform greater understanding of how our own brains operate and might lead to more treatments for mental health type conditions and things like that?

Ewelina Kurtys: Absolutely. That in many ways. I believe our research can contribute, although it's not our objective. It's a side effect. Like in every deep tech project, you have a kind of side effect which you didn't aim for, but you get some technologies, some solutions which... like, for example, understanding learning in vitro would be a huge revolution also for medicine.

Charles Humble: Is there any question that I have not asked you that you would have liked me to have asked you?

Ewelina Kurtys: Well, what I could add is where people can find us, which I would recommend highly. Check our website, finalSpark.com, which is a great repository of information about biocomputing. I think people can decide how they want to interact with us. They can send us a message, they can subscribe to a newsletter, they can join Discord.

At least on our website, we give a choice. It depends. Even if you are a journalist or you are a general public person or a scientist or investor, there are different ways you can engage with us and we are open for discussion with everyone. I would highly encourage checking our website.

Charles Humble: I will include a link in the show notes. Are there any books or papers or anything that you recommend if we've sparked anyone's interest in the subject?

Ewelina Kurtys: I would definitely recommend our own paper, which you can find on our website. This is the only paper we've published because we are a private company, so we don't really do academic career, let's say. But we have published a paper about our lab.

Everyone who is interested in the field of biocomputing, maybe to work on this in the future or even now, I think that's a good starting point because it gives you the foundation of what you need to know to work in the field. You need some biology, of course you need some engineering. But in the paper, you have exact examples of what you need to be able to work in the field.

I think it's very useful. I always recommend this to our future interns, for example. But to anyone who is technical. For books, I don't know any book so far. I think this is too new a topic. I could recommend our blog articles on the website. I think in such a new field, it's more likely you find articles, not really full books on the topic. At least I'm not aware of any. But if anyone knows, they're welcome to send us anything. But I've never heard about a book specifically on biocomputing.

Charles Humble: That's fantastic. I don't have any more questions. Thank you very much indeed for joining me today on the podcast.

Ewelina Kurtys: It was a pleasure. Thank you.