Live data from Hacker News

Google DeepMind CEO says some form of AGI possible in a few years

wsj.com

311–320 of 388 posts

Re: Google DeepMind CEO says some form of AGI possible in a few years

#311

As someone who has worked in the field of AI/ML for quite awhile now, the problem with current AGI predictions is ML hasn't done anything new since the 80s (or arguably earlier). At the end of the day all ML is using gradient descent to do some sort of non-linear projection of the data on to a latent space, then doing some relatively simple math in this latent space to perform some task. Personally I think the limits…

I disagree with you. > Essentially all we've done is pushed the basic model proposed by linear regression to it's absolutely limits No, we haven't pushed linear regression to its limits. If it was only linear regression, it wouldn't work. Neural networks need a non-linearity to model complex things. The beauty is that given an infinite series of nonlinearities, one can model any mathematical function. In practice we…

> I think ChatGPT is good evidence of this. What evidence do we have that this isn't how intelligence works?

The fact that ChatGPT doesn’t behave like an intelligence. (Though it converses like one with radical deficiencies in certain areas, which highlights the narrowness of the Turing Test, which itself is a big step.)

OTOH, it gets less bad at this when you wrap it in a loop that provides recall and other capacities with carefully constructed prompting as to how (“thought” process-wise, not just interaction mechanics) to integrate those capacities, so there’s maybe a decent argument that it models an important component of general intelligence.

Re: Google DeepMind CEO says some form of AGI possible in a few years

#312

Earlier quoted context omitted.

It’s terrible and math and logic, but ChatGPT is amazing at concepts—that’s why it is so powerful. It doesn’t work programmatically—that’s why it fails at logic. But it can reason inductively very very well. Do you have an example besides logic/math where it doesn’t understand simple concepts?

No, it's amazing at words . Humans use words to encode concepts, but ChatGPT doesn't get the concepts at all - just words and their relationships. To the extent that humans have encoded the concepts into words, and that text is in the training set, to that degree ChatGPT can work with the words in a way that is at least somewhat true to the concepts encoded in them. But it doesn't actually understand any of the conce…

I disagree. If you play around with ChatGPT4 enough you can see it understands the concept. That is, it's able to model it and draw inferences from the model in a way that's impossible through just words and their relationships. "Sparks of AGI" paper gives some good examples, for instance where it's asked to balance a book, 9 eggs, a laptop, a bottle and a nail. I recently asked it to design a plan for a bathroom. It wrote and SVG mockup and got most of details right. For instance, it understood that sink and tub required both cold and hot water lines but toilet only requires a cold water line. These things are not possible with just words, you can see it's able to create an underlying world model.

Re: Google DeepMind CEO says some form of AGI possible in a few years

#313

As someone who has worked in the field of AI/ML for quite awhile now, the problem with current AGI predictions is ML hasn't done anything new since the 80s (or arguably earlier). At the end of the day all ML is using gradient descent to do some sort of non-linear projection of the data on to a latent space, then doing some relatively simple math in this latent space to perform some task. Personally I think the limits…

It's all moot until we can agree on a definition of intelligence. I don't necessarily think it's relevant to the discussion whether an AI is truly intelligent or not. What matters is whether the technology is good enough to fool an "intelligent" human. It's good enough to do that today. I'd say we would have to train a "dumber" model that can respond with a simple "im not sure" instead of the typical responses we get today would make it even more believable, not less.

It's obvious it's an AI because it's too smart. It's too intelligent. Dumb it down, remove the AI branding, and it would fool the majority of the world.

Re: Google DeepMind CEO says some form of AGI possible in a few years

#314

Earlier quoted context omitted.

This is the debate, isn’t it? I think if we create tests for understanding and deliver them to people, we will find variations in what people understand. I think we will find the same for chatGPT. But I suspect your notion of understanding is not measurable, is it? For you, chatGPT lacks something essential such that it is incapable of understanding, no matter the test. Or do you have a way to measure this without ap…

Well, consciousness is part of the question, isn't it? We know that we are conscious (even if we can't precisely define what that means). Is ChatGPT conscious? I'm pretty sure the answer is no, but how do you prove it? Does understanding require consciousness? Maybe yes, for the kind of understanding I'm thinking of, but I'm not certain of that. How do you measure understanding? You step a bit outside the training se…

Give an example where ChatGPT4 falls down horribly on a novel concept where human would do better?

Re: Google DeepMind CEO says some form of AGI possible in a few years

#315

Earlier quoted context omitted.

I disagree with you. > Essentially all we've done is pushed the basic model proposed by linear regression to it's absolutely limits No, we haven't pushed linear regression to its limits. If it was only linear regression, it wouldn't work. Neural networks need a non-linearity to model complex things. The beauty is that given an infinite series of nonlinearities, one can model any mathematical function. In practice we…

I like what you are saying but not the last line. Angels exist - what evidence do we have that they don't? ChatGPT points to us possibly going in the right direction. It is so good that people need to write long articles about why it is not as good as a human. Contrast to older efforts, which were pretty clearly not so great. I find this pretty compelling, and GOTCHA examples that show that ChatGPT isn't as good as h…

You inverted the argument accidentally with your angel analogy, I think something like "we have labradors, what evidence do we have that they are not angels?" matches much more closely to the thrust of their point while also highlighting what you find lacking about the argument.

Edit: the important bit is that the subject already has some qualities in common with the proposed classification which is a significantly different proposition than a purely hypothetical orbital teapot

Re: Google DeepMind CEO says some form of AGI possible in a few years

#316

As someone who has worked in the field of AI/ML for quite awhile now, the problem with current AGI predictions is ML hasn't done anything new since the 80s (or arguably earlier). At the end of the day all ML is using gradient descent to do some sort of non-linear projection of the data on to a latent space, then doing some relatively simple math in this latent space to perform some task. Personally I think the limits…

Unfortunately this argument, while imho entirely valid and also frequently seen in discussion, is unable to stop the massive train that has been set in motion. There are various types of retorts: i) the brain is also doing gradient descent; ii) what the brain does does not matter (if you can fake intelligence you have intellegence) iii) the pace is now so fast that has not happened in decades will happen in the next…

The brain is definitely not doing gradient descent. That is the biggest issue with using artificial neural networks as a neuroscience modeling tool. The fact other learning algorithms have not come close to gradient descent performance in most cases has therefore largely kept neuro theory as a niche in modern academic research -- despite its entanglement with the early development of artificial neural networks.

Not taking a stance on whether "intelligence" can emerge out of gradient descent but it's certainly not biological.

Re: Google DeepMind CEO says some form of AGI possible in a few years

#317

Earlier quoted context omitted.

If all your transistors update instantly at any distance, you have less of a problem, but you still have one.

You don't have to simulate in real-time

No, you don't, but if you're taking all the data at a "subatomic level" you will have so much memory required than unless you have some kind of memory I haven't heard of, it will need to take that into account because the memory alone will have a volume of significant light-time dimensions.

You need something like 3 petabytes just to model the neurons. There are 100 trillion atoms in a neuron. And he said subatomic, so you need some significant factor of that, plus you have to save some state.

So 300 trillion petabytes+ might be a memory the size of the moon. I think relativity does matter at that scale. Also, you will probably have to mine silicon asteroids and invent an AI factory to make the memory, and power it with what, fusion?

So yeah, just model the brain at the subatomic level, bro, it's easy bro, and you're just a naysayer for saying it can't be done, bro.

But, please. If you have a solution to this, don't post it on HN, file a patent. Not just for the model, but for the revolutionary new dense memory, the location of silicon asteroids, the planetary memory factory, the power source for it, and so on.

Re: Google DeepMind CEO says some form of AGI possible in a few years

#318
post #174

Earlier quoted context omitted.

Prove such a simulation can be created with actual hardware, taking relativity into account.

If you were to try to build such a simulation, given the natural limits of real hardware, wouldn’t some form of relativity be a necessary feature of the simulation to avoid segfaults, rather than something simply to take into account while constructing the hardware?

In what way is that not "taking it into account?"

Re: Google DeepMind CEO says some form of AGI possible in a few years

#319
post #309

Earlier quoted context omitted.

They probably can't overcome the universal speed limit. Space is big.

A nearly omnipotent AI could probably get close to light speed, which would allow it to reach all edges of the galaxy in 100,000 or so years.

That's assuming the AI really wants to make a lot of paperclips though

Re: Google DeepMind CEO says some form of AGI possible in a few years

#320
post #13

The other few things possible in next few years for past few years are: 1. Fusion 2. Self driving cars. 3. Ubiquitous, generally available and stable 8192 bit quantum computers. 4. High energy density grid scale batteries. Just few more "this maybe the... what could revolutionize..." articles away. What we forget is that it is in these people's business interest (executives making such statements) to state the exagge…

I'll add another one: all trucks and cars will be electric in 10 years.

I'd reformulate it like this: "Most new trucks and cars will be electric in 10 years"

The keywords being "most" and "new"

Post reply on HN