| Not checkable |
Not checkable |
23 Jul 2025 |
Yeah, exactly. But what you’d like is a little bit better than just sort of a random generation.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And we’re still seeing some fantastic progress on each different version of Gemini.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
But then afterwards they’ll figure out with their intuition why this works.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
AI will create new jobs that cannot be imagined today, as the internet did.
Prediction · 23 Jul 2025 |
AI, the economy and work |
| Not checkable |
Not checkable |
23 Jul 2025 |
But yeah, you can’t do that without having that abundance first.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So for example, the water access problem goes away because you can just use desalination.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And of course we have a lot of great compute as well.
Not checkable · 23 Jul 2025 |
|
| Own records only |
Factual |
23 Jul 2025 |
Hassabis said it maybe takes roughly six months or something to do a new full run and the full productization of a new version of the model.
Factual · 23 Jul 2025 |
AI capabilities |
| Not checkable |
Not checkable |
23 Jul 2025 |
And I agree with that.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Demis Hassabis used to talk with Paul Nurse, whom he describes as a mentor in biology.
Factual · 23 Jul 2025 |
Their own work or organisation |
| Not checkable |
Not checkable |
23 Jul 2025 |
So in a way that’s what I want to build AGI for is to help us as scientists answer these questions like P equals NP.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So combined with Elon’s, amazing self landing rockets, then it could be you sort of like a bus service to space.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
But I still operate and I was always operating with old DeepMind as a start-up still.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I would say I’m not sure it’s even desirable because that’s a kind of hard takeoff scenario.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
Not to the level that you can do it, Lex, I don’t think.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think we need to approach it with whatever you want to call it, a spiritual dimension or humanist dimension.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And I encourage top economists in the world and philosophers to start thinking about how is society going to be affected by this and what…
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So I try to maintain good relations with all the leading lab people.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Over the last decade or 15 years, maybe 80-90% of the breakthroughs underpinning the modern AI field came originally from Google Brain, Google Research and DeepMind; he prefaced the figure with "I don't know".
Factual · 23 Jul 2025 |
AI, the economy and work |
| Not checkable |
Not checkable |
23 Jul 2025 |
Now those could be difficult to model.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
One of the best definitions I like of consciousness is it’s the way information feels when we process it, right?
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And then the other thing is maybe it’s a sabbatical after AGI has been safely stewarded into the world and delivered into the world.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
It’s pretty tricky problem that I’ve not heard a clear solution to.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
Science is a collaborative endeavor.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
the trick is how do you break it down into manageable, achievable, interim steps that are meaningful and useful in their own right.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Hassabis has experience working with cutting-edge technology in products, as games were in the 1990s.
Factual · 23 Jul 2025 |
Their own work or organisation |
| Not checkable |
Not checkable |
23 Jul 2025 |
I think all these things are related, at least in my mind.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And of course that’s what you would need for a true AGI system.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
Well, games were my first love really.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
you can’t do it without the best talent.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think that’s one good thing about the chatbot era and the product era of AI is that everyday person can actually feel and interact with…
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And I think you could build absolutely mind-blowing games.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I mean, this is my whole reason why I worked on AI and AGI my whole life, because I think it can be the ultimate tool to help us answer…
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So you’ve got to continually monitor this increasingly larger and larger suite of benchmarks.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
But in my world, they’d be related because it would be an open world simulated game as realistic as possible.
Not checkable · 23 Jul 2025 |
|
| Resolved true |
Factual |
23 Jul 2025 |
Demis Hassabis worked on the early stages of the game Black & White in the 1990s, which he says had probably the best learning AI in a game.
Factual · 23 Jul 2025 |
Their own work or organisation |
| Not checkable |
Not checkable |
23 Jul 2025 |
I think they operate over different timescales and they’re equally important to address.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Demis Hassabis owned a Spectrum and then a Commodore Amiga 500 computer while growing up in the UK in the late 1980s and 1990s.
Factual · 23 Jul 2025 |
Their own work or organisation |
| Not checkable |
Not checkable |
23 Jul 2025 |
Of course, you can’t create infinite amounts of game assets.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
But of course it’s very hard to program open world games.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And we’re not that nobody’s there yet, in my opinion.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
You need them, but it’s important that you don’t overfit to them.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
AI's impact will be at least 10 times that of the Industrial Revolution.
Prediction · 23 Jul 2025 |
AI, the economy and work |
| Unresolvable |
Factual |
23 Jul 2025 |
Current AI systems cannot come up with a conjecture worthy of study by a mathematician such as Terence Tao.
Factual · 23 Jul 2025 |
AI capabilities |
| Not researched |
Factual |
23 Jul 2025 |
Hassabis said he thinks his own lab is really the only AI lab using AI for science, while other labs talk about science.
Factual · 23 Jul 2025 |
Science and research methods |
| Not checkable |
Not checkable |
23 Jul 2025 |
And we’re continually fighting and cutting away bureaucracy to allow the research culture and the relentless shipping culture to flourish.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think he would’ve loved where we are today
Not checkable · 23 Jul 2025 |
|
| Not yet due |
Prediction |
23 Jul 2025 |
There is a 50% chance that AGI is achieved by 2030.
Prediction · 23 Jul 2025 |
AI timelines and forecasting |
| Not checkable |
Not checkable |
23 Jul 2025 |
Well look, of course there’s a strategy that Meta is taking right now.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So I think political philosophy and political science is going to be key to that.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And probably a system being able to do several of those things for it to be very general, not just one domain.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Factual |
23 Jul 2025 |
Hassabis's bet is that the brain is mostly doing classical computing, which he said would suggest its phenomena can be modelled or mimicked by a classical computer.
Factual · 23 Jul 2025 |
Science and research methods |
| Not checkable |
Not checkable |
23 Jul 2025 |
Basically polynomial times, so tractable by a classical system, which a neural network is.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
generally I’m in huge favor of open science and open source
Not checkable · 23 Jul 2025 |
|
| Mostly true |
Factual |
23 Jul 2025 |
Video models such as Veo 3 can pretty accurately generate about eight seconds of video that, by eye and at a glance, looks consistent.
Factual · 23 Jul 2025 |
AI capabilities |
| Not checkable |
Not checkable |
23 Jul 2025 |
technology, I’ve always seen technology as the enabler, the tools that enable us to flourish and to understand more about the world.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think there probably needs to be 10 times more effort of that than there is now as we are getting closer and closer to the AGI line.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think they’re most likely in the past, if you read that book of how unlikely to go have any life at all.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So actually I think it’s an example of very promising directions where you combine LLMs or foundation models with other computational…
Not checkable · 23 Jul 2025 |
|
| Own records only |
Factual |
23 Jul 2025 |
Interim Gemini versions such as 2.5 are often patches or post-training ideas applied to the same base architecture.
Factual · 23 Jul 2025 |
AI capabilities |
| Not yet due |
Prediction |
23 Jul 2025 |
Over the next five to ten years, people who embrace AI technologies will become superhumanly productive in creative and technical industries.
Prediction · 23 Jul 2025 |
AI, the economy and work |
| Not checkable |
Not checkable |
23 Jul 2025 |
And if they can’t, then I think you can be pretty confident we have a fully general system.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And I think we haven’t really even sort of scratched the surface yet of what classical systems so-called could do.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
It’s almost like cutting edge part of psychology research or personality research.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
No, it won’t solve all problems because there’s still other human foibles that will still exist, but it will at least remove one, I think…
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Companies such as Isomorphic Labs were spun out on the back of AlphaFold to do drug discovery, and it is going really well.
Factual · 23 Jul 2025 |
Science and research methods |
| Not checkable |
Not checkable |
23 Jul 2025 |
Yes, the mastery. There’s nothing more satisfying in a way.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
Google DeepMind has models that define the Pareto frontier of performance versus cost and latency.
Factual · 23 Jul 2025 |
AI capabilities |
| Unresolvable |
Prediction |
23 Jul 2025 |
AI will have probably about 10 times the impact of the Industrial Revolution, occurring about 10 times faster.
Prediction · 23 Jul 2025 |
AI, the economy and work |
| Not checkable |
Not checkable |
23 Jul 2025 |
I think to the extent that it can predict the next frames in a coherent way, that is a form of understanding
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
But at the moment these are more incremental improvements, sort of small iterations.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
It’s all good for humanity in the end if you cure terrible diseases and you come up with an incredible cure, this is net win for humanity.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I hope not. I think that would be very dangerous to do.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And that research base means that if some new breakthrough is required, like an AlphaGo or Transformers, I would back us to be the place…
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And I think that shows that it has some notion of at least intuitive physics, how things are supposed to work intuitively
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
There is roughly a 50/50 chance that new breakthroughs are needed for AGI rather than scaling existing methods being enough.
Prediction · 23 Jul 2025 |
AI timelines and forecasting |
| Not yet due |
Prediction |
23 Jul 2025 |
Within the next five years we will probably have AI systems that can materially help with climate and energy problems.
Prediction · 23 Jul 2025 |
AI capabilities |
| Not researched |
Factual |
23 Jul 2025 |
Google has its own TPU line and is looking at inference-only chips to make inference more efficient.
Factual · 23 Jul 2025 |
AI compute, hardware and data |
| Not checkable |
Not checkable |
23 Jul 2025 |
But with an AI that’s on silicon, we won’t be able to rely on the second part, even if it exhibits the first part
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think it would imply a level of precision that is not there.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
AI technology could be used to provide early warning of bad-actor uses such as bio or nuclear threats.
Prediction · 23 Jul 2025 |
Biosecurity |
| Unresolvable |
Prediction |
23 Jul 2025 |
Taste or judgment in science will be one of the hardest things for AI to mimic or model.
Prediction · 23 Jul 2025 |
Science and research methods |
| Mostly true |
Factual |
23 Jul 2025 |
Veo, Google DeepMind's video generation model, can model liquids surprisingly well.
Factual · 23 Jul 2025 |
AI capabilities |
| Unresolvable |
Prediction |
23 Jul 2025 |
Demis Hassabis said he does not really see the demand for AI compute slowing down, because as AI systems get better they will become more useful and there will be more demand for them.
Prediction · 23 Jul 2025 |
AI compute, hardware and data |
| Not checkable |
Not checkable |
23 Jul 2025 |
I think information is primary, information is the most sort of fundamental unit of the universe, more fundamental than energy and matter.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
Our brains are general systems, intelligent systems, so I think it’s almost limitless what we can potentially do with them.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
I think that’s what separates the great scientists from the good scientists.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
It’s difficult with the current geopolitical climate, I think, to see cooperation, but things can change.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So I think most emergent systems, cellular automata, things like that could be model-able by a classical system.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So I often say it’s harder to come up with a conjecture, a really good conjecture than it is to solve it.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
We’re trying to build general systems that are good across the board, and you try and make no-regret improvements.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
AI could be amazingly transformative for good.
Prediction · 23 Jul 2025 |
Philosophy, ethics and society |
| Unresolvable |
Prediction |
23 Jul 2025 |
Training may become the smaller part of overall AI compute required, compared with inference.
Prediction · 23 Jul 2025 |
AI compute, hardware and data |
| Not checkable |
Not checkable |
23 Jul 2025 |
But yeah, I’m on good terms I hope with pretty much all of them.
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
So whatever your trade-off is that you want as an individual user or as a developer, you should find one of our models satisfies that…
Not checkable · 23 Jul 2025 |
|
| Not checkable |
Not checkable |
23 Jul 2025 |
And that’s why projects like AlphaFold are so important to me.
Not checkable · 23 Jul 2025 |
|
| Unresolvable |
Prediction |
23 Jul 2025 |
Asteroid mining will become a real activity.
Prediction · 23 Jul 2025 |
Technology |
| Not checkable |
Not checkable |
23 Jul 2025 |
So they can be efficiently rediscovered or recovered because nature’s not random.
Not checkable · 23 Jul 2025 |
|
| Not researched |
Factual |
23 Jul 2025 |
AlphaFold 3 is a first step towards modeling biological interactions, such as proteins with proteins, RNA and DNA.
Factual · 23 Jul 2025 |
Science and research methods |
| Not researched |
Factual |
23 Jul 2025 |
Classical computing systems can model protein structures and play Go better than world champion level.
Factual · 23 Jul 2025 |
AI capabilities |