Transcript of The Man Who Thinks AI Could Surpass Humanity New

Meditieren, Yoga, Joggen— nichts entspannt mich. Echt?

Mich entspannt meine Steuer total.

Steuer? Wie Finanzamt? Die Steuererklärung?

Ja, ich hab ganz locker über 1000 Euro zurückbekommen.

Hast du geheime Connections?

Nö, nur die WISO Steuer App.

Wow! Und das ist einfach?

Klar, die macht fast alles automatisch.

Plötzlich fühle ich mich so entspannt. Hol dir dein Geld zurück, tiefenentspannt mit WISO Steuer.

Kaffee in seiner besten Form mit der neuen Qubo One Kapselmaschine von Tchibo. In jeder Qubo-Kapsel steckt Spitzenkaffee aus besonderen Anbaugebieten. Für Espresso, Kaffee-Crema oder Kaffee-Grande auf Knopfdruck. Die neue Qubo One überzeugt mit Premium-Design, kompakter Größe und kleinem Einstiegspreis ab 29 €. Dank innovativer PressBrew-Technologie wird jede Tasse besonders aromatisch mitsamt dicker Crema. Entdecke jetzt die Qubo-Kapselmaschinen in deiner Tchibo-Fiale und auf tchibo.de.

For years, artificial intelligence was treated like science fiction or a Silicon Valley party trick. Then almost overnight, it began writing code, designing proteins, and raising a question that suddenly feels less than theoretical. Can machines actually think?

In his new book, The Infinity Machine, journalist Sebastian Malaby traces how AI moved from academic obscurity to the center of science, business, and geopolitics.

And at the heart of that story is a man named Demis Hassabis, the chess prodigy and Google DeepMind co-founder whose work helped launch the modern AI race.

Malaby joins us today to explain the human story about how Hassabis got us here, how AI is already changing, and why the next breakthroughs may be far bigger than chatbots.

I'm Daily Wire Executive Editor John Bickley with Georgia Howell. This is a weekend edition of Morning Wire.

Spring is here and your yard probably needs some love after winter. Our sponsor FastGrowingTrees is America's largest and most trusted online nursery with thousands of trees and plants and over 2 million happy customers. Think of it as your local nursery available anywhere with a bigger selection than you'll find anywhere else. From fruit and privacy trees to flowering varieties, shrubs, and houseplants, every plant is grown with care and guaranteed to arrive healthy. Simply click, order, and grow to create your dream yard with options that fit your climate, space, and lifestyle. I ordered two Bambino Fiddle Leaf Figs, so they're small, but they are so adorable and they arrived in beautiful condition, healthy, happy, and my living room is forever changed. Their Alive and Thrive Guarantee ensures your plants arrive happy and healthy, backed by ongoing support from trained plant experts to help you plan, choose, and care for your new greenery. Right now they have great deals on spring planting essentials, up to half off on select plants. And listeners to our show get 20% off their first purchase when using the code WIRE at checkout. That's an additional 20% off better plants and better growing at FastGrowingTrees.com using the code WIRE at checkout.

FastGrowingTrees.com, code WIRE. Now is the the perfect time to plant. Let's grow together. Use WIRE to save today. Offer is valid for a limited time. Terms and conditions may apply. There's a reason so many Americans feel called to stand up for children right now, because in a culture that moves incredibly fast, kids deserve protection, guidance, and the chance to grow up without being pushed toward irreversible decisions before they're ready to fully understand the consequences. That's why organizations like Alliance Defending Freedom are stepping in. ADF is working to encourage major corporations to reconsider policies that provide coverage for so-called gender transition procedures for minors through employee healthcare plans. This conversation is bigger than politics. It's about whether our culture is willing to slow down, protect childhood, and support families navigating difficult decisions with care and wisdom. And the reality is, public pressure matters. When everyday Americans speak up together, companies listen. That's why Alliance Defending Freedom is inviting people across the country to make their voices heard by signing their petition to corporate America. If you believe children deserve compassion, protection, and thoughtful care, this is an opportunity to take action. Visit joinadf.com/wire to sign the petition today.

Again, that's joinadf.com/wire.

Joining us now is Sebastian Mallaby, an acclaimed journalist and author, a contributing columnist at The Washington Post and senior fellow at the Council on Foreign Relations. He's also a two-time Pulitzer Prize finalist, and he's got a new book, The Infinity Machine, which takes us inside the race to build artificial intelligence and the people that are trying to shape its future. You got a lot going on, Sebastian. Great to have you on.

Great to be with you.

So first, are we really at the beginning of a scientific revolution with AI? That's what a lot of people believe. Is this true, or are we just in a commercial boom that looks bigger than it is?

No, I think it's a real scientific revolution. I think it's probably bigger than the Industrial Revolution, although that was enormous. It's more like the invention of, you know, consciousness. I mean, last time we had a new form of intelligence that came into the world, it was human intelligence, and now we have machine intelligence.

So bigger than the Industrial Revolution. That's obviously a huge claim. And as you highlight in your book, the scientific arena appears to be poised for perhaps the most important advancements with the use of AI. Before we go further, you've written about AlphaFold, which has a major scientific application. What exactly is AlphaFold and why does it matter so much?

Sure. AlphaFold was a system invented in 2020 by DeepMind, the Google AI subsidiary in London. And what it does is it unravels all of the shapes of proteins in nature, like 200 million different intricate protein shapes so that you can see how they, how they look, and therefore you can figure out what kind of compounds might bind onto them. This is useful for building new medicines, for example.

And you believe this is going to be something that's a major player in the advancement of science, correct?

Absolutely. I mean, I think all kinds of scientific frontiers will be more crackable in the future than they have been in the past. We already see that in medicine with, you know, faster drug discovery, although the clinical trial process is still very long. So we haven't seen actual drugs in human beings making them healthy yet, but the discovery process is faster than it used to be. Material sciences is gonna be affected. Progress in nuclear fusion is gonna be affected. It's going to be a big deal.

All right, so we're seeing AlphaFold being used in drug development. What would you consider proof that it really can make a difference in the real world?

Well, when I say it's working, what I mean is I've spoken to researchers who are running pharmaceutical companies, and they use this tool all the time. It used to be that you wanted to find the shape of the target protein onto which your medicine was going to bind, and it could take a PhD, biologist, you know, 3 or 4 years of work to find that shape because these proteins are very complicated and intricate in their shape. And now it's just like Googling something. You just look it up in this database where all of the shapes have been stored for you by DeepMind. So it saves you a lot of time.

Now you've mentioned DeepMind a couple of times. What is DeepMind? What, what sets DeepMind apart from other AI projects?

Yeah, so it's interesting. I mean, DeepMind was the original AI discovery lab set up in 2010. And then when OpenAI, which is maybe a bit more famous, especially in the US, was created in 2015, it was basically a copy of DeepMind. And then Anthropic, also famous, came 5 years after that. And so DeepMind was the original, and it was set up by this guy Demis Hassabis, who is a British of Silicon Valley folks like to say they've got the melting pot. This guy had a Chinese-Singaporean mother and a Greek-Cypriot father, which makes him a classic Londoner. And he believed in AI. He believed it was going to work. And he set up DeepMind in London in 2010.

I wanted to ask you exactly about him. So Demis Hassabis, a brilliant guy. Can you give us a little background on him so we got a sense of who he is and what his mission is?

Yeah. So when he was about 4 or 5 years old, he climbed up on a chair and watched his dad playing chess and instantly understood the game. And then within a year or two, he was playing in tournaments., at chess. And then a few years later, he was the best youth player in Britain and was the captain of the team and the second best player worldwide. Um, and then he decided that chess wasn't enough for him and he branched into coding and he wrote a video game design which sold like 5 million, uh, copies before he was 18. Uh, and then the boss of the video game design studio said, hey, you're so good at this, I'll give you more than a million bucks. If you just stay with me and work with me and you don't go to college. And he said, no, I'm going to college. And he went and did computer science.

So how is the development of AI tied so intimately with this particular figure?

Well, the extraordinary thing is that Demis Hassabis had this vision that you could build powerful AI really, really early. And just to set the context here, in 2010, when he created DeepMind, AI could not recognize the photograph of a cat. Like AI could do nothing. It was deep, deep AI winter. And it was only in 2012 that anything began to work. And so part of that making it work was Demis and his team at DeepMind building an agent that could play all kinds of different simple video games. There's this suite called the Atari suite, you know, the games like Pong, Breakout, Seaquest, and so forth. And they built an agent that without being told the rules of these games, could figure out how to play it by itself and then figure out strategies for winning and, uh, get really strong at them. And so this was an early proof that AI was gonna work.

All right. So his gaming background really ended up being key in the development of AI.

Absolutely. So gaming was important partly because, um, the games developers back in the day realized that if they had AI, they could make the games more compelling and interesting and fun. 'Cause you could have characters in the game that like learnt new behaviors based on the feedback they got from the player. So that kind of set the seed of the aspiration to build AI. But then also games turned out to be a fantastic environment in which to train an AI agent because, you know, you could play the game over and over and over, and the machine could learn by playing many times. And so after that Atari system I told you about, the next thing was to build an AI system that could win the game of Go, this ancient Chinese game. And in 2016, they went to Korea, played this big show match against a Korean champion, and the AI system defeated him.

Can the AI beat Hassabis in chess?

Absolutely. Let's do that. You know, ever since 1997, when the Deep Blue system that IBM had defeated Garry Kasparov in chess, it's been game over for humans, I'm afraid.

So after spending years with Hassabis, what did you find as the underlying thread for him? What is really driving him? There's a lot of emphasis on this with Elon Musk and others in the AI race. What do you think he's ultimately aiming for here? What is he trying to really accomplish?

Yeah, so I think, you know, different leaders in this space have different fundamental motivations. I think for Elon, he wants to be the biggest, baddest industrialist in the whole world, and he just wants to, if he sees the next fantastic technology, he wants to do that too. So space and electric vehicles wasn't enough for him. I think with Sam Altman, he wants power. You know, he thought of running for governor of California. Allegedly, he thought of running for president. And AI is kind of a way to get to prominence and influence. And for Demis Hassabis, it's different. What drives him is scientific curiosity. He's possessed, I mean possessed, by his desire to understand what he calls the fabric of reality. And he believes this so strongly that he expresses it in quasi-spiritual terms. He says to me, I see reality staring at me in the face, waiting to be discovered. And if I can understand it better, I'm getting closer to the intelligent being that may perhaps have created all of the order around us. Maybe that's God. Maybe if I can do AI, understand the fabric of reality, I will be approaching God.

I think when a lot of average people like me engage with AI, all we see is that it's imitating human behavior. It looks at and sort of aggregates and calculates predicts what is the most likely next thing or most likely response that a human would do, et cetera. A whole other level of consciousness is something else. I mean, do you actually agree with that?

Yeah, I think the consciousness word gets into tricky arguments because we don't really have a great definition for what consciousness is. So I'm going to sidestep that bit of it. But I think if you just don't worry about conscious or not, you just look at the outward behavior of these machine intelligences, you see that, for example, in a domain like the game of Go, it not only beats humans, it invents new strategies that humans had never thought of. And so that's more than just imitating humans. It's actually inventing something completely novel. And so I do think that this is a creative intelligence, um, that will come up with new knowledge, new science, new breakthroughs. And you know, it's both exciting and scary.

Yeah, it is. And speaking of that, AI is becoming as strategically important as nuclear weapons or energy in many ways. Are we already in kind of a Cold War dynamic between the US and China in particular?

Absolutely. I mean, AI is like the new nuclear material. You know, you can use AI to make your weapons more destructive. You can have swarms of drones controlled by AI. You can use AI for cyberhacking, obviously for intelligence interpretation and collection. So this is a super powerful weapon.

What were some things that maybe surprised you in your research on this when it comes to AI's capacity? What are some things that you feel that the average person doesn't realize AI can play a key role in?

Well, maybe the best way to answer is to describe what protein folding really entails. So you have a strand of amino acid, imagine this string, and it's got some DNA code on it, and the sequence of the DNA on that string predicts the way that the string will fold itself up like a self-executing origami model into a super complicated intricate shape with tentacles hanging off it and so forth. And to predict how that string will tie itself up into a knot and a strange shape, it's like it, it could, it could fold itself in trillions of different ways. How the heck do you predict that shape? And AI can do that. It can just look at the sequence of the code on the string and say, ah, here's what it's gonna look like. I mean, it's way, way beyond what humans could imagine doing. The biggest example at the moment is the industry of writing code. I mean, software developers, um, almost all of them now, uh, do their jobs with kind of a coworker, which is an AI, which writes a bunch of the code for them. And, you know, you hear different estimates from different labs, but let's say around half of the code.

These days is being written by code.

Well, really amazing stuff. It's hard to wrap your mind around, but as you've highlighted, there's also a very human story behind all of this. Thank you so much for joining us.

Absolutely. Thank you very much for having me.

That was Sebastian Mallaby, author of The Infinity Machine, and this has been a weekend episode of Morning Wire.