How Brain Organoids Work (And Why They Could Replace Silicon)

Junaid Kalia, MD:

Good morning everyone and welcome to Signals and Symptoms podcast. I'm extremely grateful today for Edmund White to join us from Germany. He has a very unique perspective and in all honesty I'm actually above my pay grade what he's going to discuss.

So I'm going to let him introduce himself and then talk about AI today or OI tomorrow, which is Organoid intelligence and one of his thesis and of course he has done some major work in it. He's essentially a biomedical innovator and author.

Edmund White:

Wow, thank you very much Junaid and Ed and very honored to be participating in your podcast today. So and greetings from Romania. I live in Germany but I'm in transit for some meetings this week in, in Romania. So yeah, just a little intro so about about myself for people who don't know who I am.

So I'm basically been working in sort of medical innovation up to 30 years now. And primarily really most, most of my life has been working with technologies that really when patients are sick and coming in for arrhythmias, heart failure and helping them with devices to improve their outcomes and been doing that for various different guises in large corporate startup companies.

And then about a few years ago started reading and thinking about how can we get prevention more into the equation. So for example can we actually stop these patients coming in so late and get earlier on intervention. And I saw one of the other podcasts someone's talking about prevention I think Philip Chairwell and kind of triggered. So this kind of triggered my mind. So about 20, 24, 23, brain organoids came. I was thinking about brain organoids and I was looking really about.

So I was working for an organization where we could take the echoes actually of children's hearts in conjunction these and made them into 3D models so we could actually look at twinning technology.

And it was just so many different dimensions coming out of that from a medical point of view. First of all, there are many other aspects there in terms of personalized, personalized medicine, looking into energy efficiencies for example, and looking into computation and whether models could.

You know, we live in a very much a silicon world today. But always for me it's always about patients got to come first, can we build technologies that help the patient outcome.

So this is what kind of, kind of got me thinking, researching and actually eventually to write a book because I just felt this really needed, the message really needed to go out there about organoid intelligence. There. And organoid intelligence is really the combination of. So brain organoids were developed from induced pluripotent stem cells back over a decade ago, there was a lady called Madeleine Lancaster from Cambridge and she was doing some research into this sort of microencephaly.

There was some experiments done to really test the input and output of these organoids, in a closed loop system to see if you could actually train them.

So a group of scientists came together in Baltimore and created this new kind of discipline. And for me it was fascinating because there was just so many things that will impact our lives out of this, both today and also we look science fiction future as well.

So that's kind of what was driving me. And I, I work in the medical space, tech space, but I also have a Very strong interest in this brain organoid side as well. Well so, yeah,

Junaid Kalia, MD:

fascinating.

So number one, learning itself in terms of human versus machine and number two big statement is how do we go back to the basics. How humans learn and human stem cells work and protein and how, how to optimize our process.

So at the end of the day I have a very efficient system here as compared energy efficient system. So that's number two layer that you properly suggested. And then of course the last one, does it matter or we should just put some more solar farms.

So I'm going to let Ed, side with the conversation now.

Edward Marx:

If you think AI is the future, think again. This is your quote. The next frontier emerges. Biology with tech building intelligence from living neurons rather than silicon. So that's pretty profound. Can you, we've talked, we've touched on this on the periphery a little bit.

But can you kind of expand on that?

Edmund White:

Yeah, no, for sure. So we all want things predictable in the world. Our silicon, our chips, everything we use is all very predictable.

And as well, you know, in all our devices we use it day in, day out and we don't really think about it, we just, just have it there in the background. So it's very good. I mean nothing wrong with silicon but one of the big issues is that it's energy, takes a lot of energy.

And you know, with the learning language models now the data warehouses are growing, there's more and more energy requirements and it's also not just the energy but the water as well involved in cooling these big data centers. So there's a lot of energy demands exponenting in the world today.

And that's one of the drivers. And the second piece is. But in biology, biology is not really predictable. It's adaptable and it's like, it's very it's very clever in terms of how it works. It doesn't use a lot of energy. But one of the challenges with biology is that you've got to nurture and grow that cell, whatever it needs, a nutrient system or something there.

And there's been work today in building biological computers. They need nutrients, they need things to support them to grow. And I was thinking in the world today that you cannot, I don't think we could today build an infrastructure around biology and replace it with silicon.

That's not going to happen today. But you could take the best elements of those worlds. So the silicon reliability, the kind of consistency of that with the energy of the biology and adaptability, because it's very deterministic model in silicon but the biological, but it's so adaptable that you combine those two.

Bit like a hybrid car where you got the petrol and electric, you know, you've got that hybrid approach.

Junaid Kalia, MD:

So Ed, here's the thing. We have democratized and demonetized intelligence. And Ed is saying, and rightly so, that it may take three years, five years. What do you think we're gonna do when.

When artificial intelligence, which is demonetized and democratized intelligence that is cheap and does not require that much energy. What are we going to do? As human beings in general, 8 billion people.

I'm gonna let you answer.

Edward Marx:

Oh, okay. So this reminds me when I was at Cleveland Clinic and there was this group of physicians who started Brain X AI.

But I don't have a profound answer.

Junaid, to this. Would really like to hear more from, Edmund. But what gets me excited about all this is that we run into limitations as we're talking about that have extreme environmental impacts.

And so I'm just super excited. The fact that, yes. With 8 billion people. And we can't, you know, with, with what we know about quantum computing today. And you know, we would run out a resource. We'd have to build things in space like Harvey's working on, you know, but that's like taking old ways and putting it in space.

but all we're doing is taking today's mindset like this is the only way we know how to do it.

So naturally it's put it in space. That makes sense. But as Edmund's pointing out, there are other opportunities. And so that's kind of my general reaction to your question.

Junaid Kalia, MD:

Love it. I mean, I'm just saying that I have, I'm living it with the demonetization and democratization of intelligence. And then you know how when you're in Germany, Germany is, the highest tax and the highest benefits.

We live in US which is somewhere in the middle. And then of course, there are countries that are not taxed. So talk about the social impact of this as well, especially your experience living in Germany. We are Texans, by the way, so we don't like some of the social. So. I'm just kidding. So overall thoughts about it?

Edmund White:

So give everybody a uniform, you know, income and then let them get on with their creative pursuits and what they want to do, their hobbies and so on.

So there's been ideas like that. So time, we're, you know, traditionally we, we sat at a computer, we given up or we've given our time, for many years, decades to work. And maybe that might be in 10 years time. We look back and think, oh, that's a quaint, that was a quaint area.

But the problem is at the same time, obviously we've got, in parallel, we've got all these humanoids are going to be produced and going to do all the tasks and do the production. So I actually have a very controversial view about where things are going and I don't think many people are going to like this.

So I'll, I'll explain it for the first time to you guys. I actually think the cost of everything, we would think inflation, everything's going up in price today. You know, you know, petrol, fuel, you know, Germany, we feel it a lot. You know, I actually think prices of things eventually will go down in price so things will drop because the reason is that we're going to be producing robotic systems to produce infrastructure, our world around us.

And that cost won't be so labor intensive and therefore the costs may eventually drop for everything that we have.

But eventually if there's no jobs because we AI is taking all the jobs, where does the money come? Well, we get a universal income, but then who's doing all that? It was all roboticized and the costs just come down. So maybe eventually we'll have a complete opposite. Prices will go down to a ceiling or something like that.

You know, Nate, there's a lot of organizations out there are native AI or coming in and doing, you know, tasks with particularly repetitive tasks at the moment of being replaced.

So I think you're right. There's going to be a lot of thinking around how we're going to have a different type of society in the future, given what we're seeing playing out.

Junaid Kalia, MD:

The point was that we have seen these and people keep saying that I think we will find ways to do something. That's number one.

Number two, we need to reinvent how we think things. And that's a new reality we're going to live in. So I think, as you said, it may not be ubi. It may not.

It may be something else. You know, are you courteous? Are you taking care of your elders? That's how you get paid. Do you understand? Like there are ways to making sure, sure that social impact is there. But then that's the real question here. Is it really not that important? I think it might be important to actually bring those, important subjects in line with, with what we're doing. So that, that's my thought.

Edward Marx:

And I'll tell you, they will say that you and I, Junaid Harvey, and maybe a lot of our listeners are in the, let's just say 5% that are out there.

We make a living, we leverage AI and we see a lot of things. But how much of it is actually like translating back down?

So are we in a bubble? And if, yes, you know, when does that bubble return to normal? And then we see an uptake of actual AI enabled, capabilities. Or are we not in a bubble and like this is people better. You know it's going to hit everyone like a tsunami.

Edmund White:

Sure. Yeah. So I, I think if we look back about 25 years ago there was the Internet dot com. You remember back in the late 90s, everything was about Internet. You know having.com was the big thing to do at that time.

I remember using the Google browser search engine when it first came out the first day. And it was really nice interface and it was nice but didn't realize how, how that was going to progress. And then there was a, a bubble and a burst.

So around 2000 I think the dot com and a lot of companies went bankrupt but there were a few which kind of survived and, and became sort of Today obviously Amazon came out of that and Google some of the big titans there and maybe the same thing is happening today but I think they're slightly different in terms of that. And so we may be actually, we actually going through a revolution but there may be in, in the long term definitely 100% but in the short term there may be a bubble.

And you know so is, is, is this all going to pop and there's going to be something there could likely. But actually longer term directories for sure. You know I think we're going to have to be working in a new, new world, new environment.

So I, I, I would say there's probably maybe a little bit of short term frothiness and then a bit of a cut. And then long term it will kind of go a bit like the dot com. That's how I take a crystal ball.

Edward marx:

I think that's pretty. I tend to agree with that, philosophy.

Junaid Kalia

But what do you think in your future given your, you know, purview of very confidential information that you're going to share with us? Of course. What's happening in space computing? How do you think this progresses in your mind? Yeah.

Harvey Castro, MD, MBA:

So I. Great, seeing everybody. here's the way I see it. We're going to be efficient with everything we have down to, like, you guys just gave some examples.

But think about all the compute power. The United States has more compute power on our cell phones than most other places. So why not leverage our cell phones? Why not leverage this compute power that I have on this iPhone? I'm not using it all the time. And where we're on a network now, extrapolate that concept.

Now, to your point, there's a weird inflection point. And there's many laws to if we make it cheaper and cheaper, we're going to use it more and more, and then we actually make it more expensive.

But then you can also argue that with time, we use it so much and it's everywhere, that we're actually at one point, how much do we really need for our daily living?

Junaid Kalia, MD:

So. Absolutely true. So again I agree with the second one with an example that there are many laws.

And I, I don't know the word law even applies anymore because it's just one engineer's dream away from going through this. So, going back to Ed,

So what are your thoughts on what Ed, was saying the AI bubble which is. You're absolutely right. You're going to see a hockey stick like a dot com bubble.

But how this adoption can be controlled or should it be controlled in the first place?

Edmund White:

So being in Germany, we sit under a lot of regulations in many different industries. So there are kind of mechanisms to use to kind of put this a little bit and keep the moat high. I think and particularly for companies and who can get and meet those requirements, it brings them a lot of I think, I think validation.

If they can get hit these hurdles in the regulation side, in the brain space where in the brain organoid space. Just going on this topic again it's a bit of a gray zone because the regulations aren't really caught up in this area. So it takes I think vision from policies to see well where this is actually going to help the population. You know. And I always feel like if we can help get the best patient outcome done it ethically in a good way then we should encourage that innovation, that policy.

But at the same time policy's also got to, you know, so it's kind of, I think it's a balancing act. I mean we love to see innovation, but we need to keep it just checked and tempered and make sure no one's doing anything, you know, incorrectly.

Junaid Kalia, MD:

Yes. Regulation is, is so insane that as a. Well, it's not worth it because I mean you, you do have, I mean I'm not you know, being facetious here,

Edward Marx:

But we're spending so much money in all these individual hospitals and in the payers and, and for sure in our partners, vendor partners.

Everyone's doing their own little experimentation, right? And it's, we're getting limited impact. Imagine if we took a fraction of that collective, took that money collectively and then experimented in a safe environment and just accelerated the pace of change.

So anyways, I, I would love to see Us somehow figure this one out where we could do some serious RD and rapid cycle.

Because otherwise we're going to be, we're gonna get beat, as Edmund was pointing out, we're gonna get beat by nefarious actors.

Junaid Kalia, MD:

Yeah, Yep. And that's so true. Edmund, you're suddenly the Chancellor of Germany. This is your last question for the day

So what would be your first, the top three things in terms of number one research dollars going into AI versus augmented AI, quantum AI, etc. The second of course is the policy that you know. And then the third one would be.

Pick a topic.

Edmund White:

Yeah, yeah, good question. Yeah, I think it's a, yeah it's about prioritization. In, in this, in this role you'd have to really look at what, what are the segments which are going to have the most impact for the dollar.

here's a thing called a black book and you can see every year where the funding has, has been impacted and what it will, the outcomes have been. And it's not always very good but you need to look into the areas where you're going to have the most innovation, create an environment, with that funding that's going to make an ecosystem, ecosystem flourish.

And I think we need to do a lot more to encourage individuals or a system which can encourage innovation to be reiterated and, and really get out there quickly. And it's happening but it needs more injection in there.

And also from a competitive point, you know, in a global market, you know, you've got to build and train that workforce, to, to, to operate in this new environment in a way where they're going to feel comfortable, and have at their hand tools and technologies that can help, that society, the society to, to do better.

There, that's a little bit of a generic answer, but I think I, I do think the system needs to be more freer, and have a little bit less pleasing, as we said earlier, but encourage that in that, in that capacity.

So that's, that's how I would. It's needed today, actually. They're not, speaking on behalf of anybody, but I do see there's a critical need in Germany at the moment. And let's bring less risk aversion into that environment as well.

Junaid Kalia, MD: