The Chaser Report - EXTRA : Welcome To the Future - Organoid Intelligence

Episode Date: May 20, 2023

Follow Welcome To The Future in all your podcast apps!What is "organoid intelligence" and will flesh computers take over the world? Charles sure thinks so. Hosted on Acast. See acast.com/privacy for more information.

The Chaser Report is recorded on Gatigal Land. Striving for mediocrity in a world of excellence, this is The Chaser Report. Hello and welcome to The Chaser Report with Charles and Dom and sometimes Lachlan, but always you, the dear listener. I hope you're having a lovely little day at the moment and enjoying yourself very much. I don't want to take anything away from it, but look, I think it's time. that you and I had a talk. So let's put on some serious but stern music. Thank you, Lachlan, in the editing suite.

I just want to talk to you, the listener, for a moment. Are you, are you free? Cool, okay. So you might want to take a seat. What the hell's going on? Hey, I give you podcasts to listen to, and you just don't. I thought we had a deal. I thought we were friends.

I thought that this parasocial relationship we developed was working for both of us, and here I am. making episodes of awesome podcasts like the shot podcast like welcome to the future like the chaser podcast and you're just not listening to them huh what's going on there i thought we had a deal and you just backstabbed me in my ears i suppose so you know what we're going to do we're going to make this series you're going to go to your room i don't care what you're doing right now whether you're running you're probably not running whether you're in your car that is statistically what you're most likely doing right now.

I want you to just sit down and listen to an episode of Welcome to the Future, where Charles and Dom talk about organoid intelligence, and it is good, and you're going to enjoy it, and you're going to follow, welcome to the future, in the appropriate podcast feeds on your favorite podcast listening app, and then we're not going to have this discussion again, okay? I shouldn't have lashed out. That was made of me. I hope this doesn't cause a rift between us.

I love you Enjoy this episode Hello and welcome to another episode of Welcome to the Future Charles you sound so upbeat about the future Which is odd given every single news article I read about the future is terrifying Well the thing is

Today we're not going to talk about Bluetooth Are you sure No No Bluetooth We're at the bottom of that well at last No, because actually this is, this is actually about the real future. Oh, yeah. And the point is, out official intelligence, AI, that's old news, that's yesterday's news.

Because every episode of this podcast so far has been either about Bluetooth or AI. Are you telling me we're coming up with a third topic? This is called, and this is actually suggested by a listener of this podcast, Kelly, who I have to thank for sending this in. organoid intelligence. Oh, good. Let's find out what that is after this. Organoid is not a word that I've heard before, Charles.

And frankly, if you'd said it to me, I would have expected it to involve porn. Nope. Nope. It's the future. Right. So, a bit of a tale, I'll just have, I'll start at the beginning, 2006. So scientists, computer scientists, facing a bit of a problem, which is that microchips are not small enough. Oh.

Right? So you know Moore's law how, you know, microchips halve in size every 18 months and thus their energy consumption halves every 18 months and thus they speed up. And yeah, they get twice as powerful every 18 months, right? It's a sort of relationship sort of thing. The point is by 2006 scientists could go, okay, there is a physical limit. There is literally a point at which electrons get too big for. the size of the microchips that we are making.

Yeah, aren't they going down to something like three nanometer chips at the moment? Which is just insanely small. Yeah, and it's going to get to the point where, like, you literally, it's just a physical limit where suddenly, instead of talking about sort of Newtonian physics where things operate, I mean, I think they already operate at the sort of quantum scale, but those quantum effects become so massive in that thing. because in such a small space

that suddenly it all becomes very unpredictable, right? So they have to solve that problem. So somebody comes up with the idea of, instead of using, like, physical things, like silicon and metal, right, to make these chips, why don't we use the same foundational things as organic life uses? Oh, wow.

So we have chemicals in organic chemistry to do it, right? And so what they did is they scraped some skin, right? And the first test was, can we create brain cells from skin? Can we just take some skin cells and make them into brain cells, basically? Okay, can we kind of make neural pathways exist within skin cells? Because then what we can do is we can sort of own a substance that is able to transmit information, electrons, all that sort of stuff that we need for computer chips.

But it's doing it at a chemical level rather than at a physical level. So not using... So basically what the brain can do in a very, very small amount of space relative to even a computer chip. So these scientists, they did this, they got it working, made brain cells from skin,

they won a Nobel Prize for it in 2012, right? Okay. So at that point, they were just trying to solve the simple problem of, you know, can we create brain cells, right? But then this one guy kept on saying, you know, like, what are some applications for this? Do these brain cells have some sort of consciousness? Because we've got complete power over these brain cells here. There's no ethical issue.

We haven't taken someone's brain and are now using it in the lab. But if we start sort of building out some of these skin cells that are now brain cells, could we actually start communicating with those brain cells? Oh, so designing a brain cell? that is patented and has no free will. Exactly. Right. Yes.

Anyway, so, and this is all like in the realms of, like it sounds science fictiony. These are all like, you know, well respected scientists. Nobel Prize winner. And Nobel Prize winning scientists across the academy, right? And so they did that. They started being able to communicate and sort of train these brain cells to sort of receive reliably receive and they're able to sort of train the brain cells to sort of have simple responses. There's nowhere near level of consciousness yet, right?

And that is where we're at the moment, right? Okay. So we're at the moment, you know, there's reassuring articles that Kelly sent me going, look, it'll be 100 years before that becomes a problem. Let's not worry about it. Now, this other guy has now come out and said, well, hang on, back in 2015, The whole point about AI was everyone said, well, we don't have to worry about the consciousness of AI because it'll be like 100 years before you can chat to a robot and then be convinced that it's a robot.

Fast forward to eight years later, we are now all just chatting to Chatbot GPT and just convinced that we're talking to. Like, artificial intelligence has largely been solved by these large language models. And the Turing test has been... People have married them, which is, you know, a far more sincere form of connection. Wouldn't be a wonderful thing to be able to marry a skin cell? So the point is that if this organic for organoid intelligence, which they've now started mastering, can be sort of matched with the same type of technology

and sort of thinking and training, because they've started to be able to train organoid intelligence, then instead of it being 100 years away that we have to worry about whether these models will achieve consciousness, We're talking about, like, maybe in the next eight years, we're talking about something where it could actually be a real problem, where these brain cells start achieving some level of consciousness. Do you think it's possible, Charles, that we, ourselves, humanity, are a scraping from the

knee of God. Oh, I think that's definitely true. A scab that instead of flicking in the bin, God just went, no, let's see if I can make this into something a little bit better. Well, this is one of the funny things is, so say I scraped you. your skin, all the problems that you face would then express themselves in the brain cells even though the cells didn't come from your brain. Oh my God. So if you made a brain out of my skin cells, would have all of my problems? Nobody wants that. That's not organoid trauma.

No, because this is the thing. So they scraped Alzheimer's, people who had Alzheimer's, and developed their skin cells into brain cells, hoping that, oh, maybe we can revoke live their brains by developing brain cells that don't have Alzheimer's. But instead, all their brains, the sort of skin-based brain cells, also had Alzheimer's. So they couldn't get around the problem. And actually they're going, well, actually now, you know, we can actually research Alzheimer's brains in a petri dish. Because we now have access to live brain cells. Without having to lobotomize.

Yeah. Charles, if only, if only some sort of. physical specimens free of any defect had been have been found if only humans I don't know in some form of master race

had been identified in a way that wouldn't be problematic at all you could just get those people and turn their skin their beautiful airian skin

into some sort of a brain bot I know you're joking but you know I am willing to proffer up my skin for science if it you have

if it involves you know creating a master race of Charles's that are easily sunburn Right. Yeah, well, you and I could both give an amazing male pattern baldness computer, couldn't we?

Do you think that maybe, so does that mean that essentially what will happen is it'll be Brad Peter or something? It'll end up being run by the best-looking person. It'd be much worse. It'll be Mark Zuckerberg. He's already part computer. No, it'll be Taylor's. It won't be Mark Zuckerberg. But he'll be the one's going to invent the technology.

No one's going to get an O-I-intelligence boss. and go, oh, I want the Mark Zuckerberg. You're going to have won the Taylor Swift model. But the computer that runs the world, Charles, will have been designed by them. Because this is what every Silicon Valley billionaire wants to achieve immortality. Yeah.

This is how they're going to do it. Zuckerberg's going to scrape it a little bit of his putrid skin. You can get a Musk bot. And Elon Musk, of course, will do that. Oh, a Bezos. They'll be a perpetual. Oh. I mean, he's already ruined Twitter.

Imagine what an organoid intelligence Elon Muskbot would do to the world. Hey, Jeff, that would be the Jeff Bezos, sort of. No, the worst, it'll only be the worst people who are turning to computers, Charles. It won't be kind people. It's certainly, there's not going to be a charity, like, you know, someone who devotes their life to good works. They're not going to be an organoid intelligence. They'll be destroyed by the organoid intelligence.

Surely Jay-Z would have a rhyming intelligence. Well, that'd be good. That wouldn't be too bad. No, but he's quite entrepreneur. Yeah, yeah, that's true. Yeah, I think he'll. The first billionaire, OI, can be Jay-Z. Everyone would buy that.

Well, you'd get a Beyonce computer, wouldn't you? The two of them. So the thing is, it sounds all very dystopic, right? Dystopian. It certainly does. But, look, there is actually a sort of good news angle to it. Well, is it less likely to be horrifying than chat GPT? I think I'd rather an intelligence that in some way resembled a human brain

rather than simply the absolute mess of ethics that is anything completely artificial. Oh, yeah, no, I mean, there's huge, like I got to send all these. ethical issues by Kelly. I didn't actually read me. It's so boring. The best way to deal with that is don't worry about it. Just let opinion writers, you know, wrestle with that. And we were talking in yesterday's podcast about how 70 years ago someone had warned about

climate change. That's the sort of thing you ignore. Yeah, no, you just, yeah, just out of mind, out of sight. I mean, that sort of thing really slows you down on the breakthroughs. What are the implications of this thing I've invented? So, but the, no, but the good news is, aside from, all the ethical problems. There is a sort of good news, which is that organoid intelligence operates at a scale

far smaller than microchips, right? So we're talking minimum 10,000-fold reduction in size, right? So think about it. So at the moment, they actually have developed a supercomputer that matches the human brain in terms of its number of connections. Thank goodness. It's a supercomputer. It costs $600 million to make.

It's the size. of two tennis cords, and it is unlikely that you'd be able to get it down much smaller than that using microchips. With a organoid intelligence, we're talking about the size of your fist. Like, that is the size of your brain. The mild fist that will destroy us all. And so that means that, so that means that, like, things like our phones are going to be either vastly more intelligent, vastly sort of thinner and slimmer and things of that. And And the amount of energy required also goes down to 10,000 fold. Imagine not having to carry a phone.

Well, the thing is, no, but you'll also have to, it'll be a different form of energy. It won't be battery. It'll be like you give it a snack. You have to feed your phone celery. It's fantastic. So our phone is like a toddler, basically. Yeah, yeah. Yeah, right.

A toddler with all the world's information at their resources. Actually, my five-year-old's pretty confident you know all the world's information. But, I mean, Charles, we won't go into this in detail. But the other version of this, I interviewed actually the other day, a world pioneer in quantum computing, which is a similarly ridiculously vast improvement in computing power. It can do, basically, it can perform calculations so complicated that conventional computers couldn't possibly perform them given available time. There's simply not enough time in the universe.

It would take millions of years in order to go and get them done. But the problem with this child... It seems like cracking my password. Yes, that's right. You can basically trivially crack any encryption screen, which is great news. And I asked this expert, what would happen if you use those quantum computers to develop scams on the internet? She said, it would be very hard to tell that they were fake.

The only problem with quantum computing is that, as I understand it, genuinely, some of the absolute leaders in the world are Australian, which makes you think it's not going to work. Yeah. I just can't, I can't conceive of a technology where Australia is actually leading the world. No, well, no, but actually, don't worry too much. Don, because if it's anything like the Black Box or the Hillshoist or the Victor Moe, you know, those iconic Australian inventions, the patent will be sold off to some American well before it gets commercialised.

Excellent. All right. Well, we might talk more about quantum computing. Some other time, I don't know whether I'm reassured or terrified by this. I love the idea of a tiny phone that has snacks, though. So go on, tech, go on. And, you know, the important thing is it doesn't matter what we think because it's invented anyway, never stop.

And the good news is probably the best way. to communicate because the whole thing is how do you actually train your phone or whatever that's organoid intelligence and the answer is I think we all know Bluetooth

Bluetooth. You'll invent the world's most amazing device and then it will have Bluetooth and it won't work there you go so the feature looks familiar. You just won't be able to connect to it yeah they'll be about to take over the world's

you know communication system and destroyed like a sky net. But it'll all be run by Bluetooth. It's kind of ensuring that if a real sky net was developed, it would simply break. That's something that's a terminated movie never explored. It's kind of liquid metal T-1000 just going

connection error. Failure to connect. Fabia to connect. All right. That's a good night to leave it on. Thank you, Charles. Our gears from Roe, we're part of the Iconoclass Network. Back next week.