Why general artificial intelligence will not be realized • Hacker News Board

Re: Why general artificial intelligence will not be realized

#331

Post by vidarh » Sat, Jul 11, 2020, 10:53 AM UTC

lurquer wrote: post #79

vidarh wrote: post #50

Earlier quoted context omitted.

I contend the brain by definition is a computer, and so that any claim that we can not produce an artificial one implies that we'll forever be unable to replicate a process that is repeated over and over by nature through simple biochemical processes.

'>I contend the brain by definition is a computer, I contend the brain is by definition a duck. It's easy to redefine words.

I'm not redefining anything. I'm applying reasonable definitions of a "computer" to the brain.

Indeed, the term was originally used about people - our electronic computers today have the name because they were taking over functions carried out by human brains.

Re: Why general artificial intelligence will not be realized

#332

Post by codr7 » Sat, Jul 11, 2020, 10:56 AM UTC

simonh wrote: post #299

Noos wrote: post #249

Earlier quoted context omitted.

That's not teaching a computer to walk, that's building a walking machine. Subtle difference, but it's easy to prove; no matter how well you build a Boston dynamics robot, it will never like walking in the same way your TV will never like entertaining people. There's no "I" there to learn.

Way to change the goal posts. We built a machine than learned how to walk. Mission accomplished. The assertion in the article was either wrong or irrelevant, pick one. Then a bald assertion, computers will never X. Says you. Just because we’re not there yet is no proof it’s impossible.

My son learned how to stand up at 8 months, pretty soon he was walking and even running. No one had to teach him anything, he did this by observing the world around him and drawing his own conclusions.

This is not about definitions, since we don't even know what exactly we're chasing there are no ways to express the difference unambigously. What we get instead is one side trying to (unsuccessfully) define the difference and the other pretending it doesn't exist.

We're a long, long way from an AI learning how to walk by itself. Neural networks and machine learning is one piece of a puzzle, expert systems are probably in there somewhere as well. Perhaps one day we will identify all the pieces but we're definitely not even close.

Re: Why general artificial intelligence will not be realized

#333

Post by cygx » Sat, Jul 11, 2020, 11:01 AM UTC

engineeringwoke wrote: post #324

krisoft wrote: post #313

Earlier quoted context omitted.

That quoted sentence argues that we do not have AGI now. I have no counterargument against that. We do not have AGI now. That sentence on the other hand fails to argue it is impossible to develop AGI. Someone before the invention of the aeroplane could have said: Our only known good model for flying is birds and insects. It took billions of years of the universe churning at random to accidentally generate birds and i…

Flight and consciousness are not in any way compareable concepts. We literally cannot perceive the constraints of the latter's underlying physical system. Sometimes I feel computer scientists choose to misunderstand physicists because it would make them feel stupid if they did understand.

We literally cannot perceive the constraints of the latter's underlying physical system.

Well, for one, I'm reasonably certain that such a system fits inside a skull...

Re: Why general artificial intelligence will not be realized

#334

Post by vidarh » Sat, Jul 11, 2020, 11:02 AM UTC

Balgair wrote: post #74

vidarh wrote: post #68

Earlier quoted context omitted.

It's not very well understood, but the process is known to exist and work. To assume it can't even in principle be replicated would be an extraordinary claim. Yet the article suggests artificial intelligence can not exist even in principle.

I mean, from a biochem perspective, it's flabbergasting that intelligence exists at all. The brain is so noisy. To me, it's not far fetched to say that you can't make it happen again. Though I think it's technically possible. We're missing something big with intelligence. We know neurons and synapses a little bit. We know a few dozen neurons a little, though it's complexity is crazy big. But that gap between a normal…

We know it happens again millions of times a year just for human brains. As such the complexity of the brain is irrelevant - what is relevant is the complexity and reliability of the machinery that constructs it.

We know the volume of the machinery that constructs it, which allows us to compute an upper bound on the informational complexity of a system capable of constructing human brains.

To me it is totally unreasonable to suggest we will never be able to at the bare minimum mimic that process and grow whole brains.

To me most of the opposition to the idea of artificial intelligence seems to come down to people assuming we're bound to only try to do this with software on a digital computer. But if that proves fruitless, there's no reason to assume we won't try analog systems, or if all else fails try biochemical systems, all the way down to genetic manipulation and tricking cells into growing into brains for us to hook up to computers.

Re: Why general artificial intelligence will not be realized

#335

Post by codr7 » Sat, Jul 11, 2020, 11:03 AM UTC

baddox wrote: post #158

empath75 wrote: post #151

Earlier quoted context omitted.

So build a body.

What is "a body" other than inputs and outputs to interface with the environment, and perhaps a mechanism of perceiving the environment and remembering what was perceived?

The capital T-truth is that no one really knows since we're all observing the system subjectively from the inside.

Many who practice meditation and/or experiment with psychedelic drugs will tell you that there's something in there that's not a computer.

We could assume, as many do, they're fooling themselves; since we can't measure it. Or, we could trust our authentic experience of the real world even when it can't (yet) be measured.

Re: Why general artificial intelligence will not be realized

#336

Post by codr7 » Sat, Jul 11, 2020, 11:09 AM UTC

PaulDavisThe1st wrote: post #167

This is possibly the worst article I have ever seen in Nature. It is barely at the level of an undergraduate essay on AI (or AGI or ANI, as the author prefers). A hand-waving argument about "computers not being in the world" that relies on not bothering to define "computers", "being in" or "the world". Convenient avoidance of almost every more subtle anti-Searle/anti-Dreyfus critic (notably Dennett). Almost wilfull n…

You're demanding that they exactly define what can't be exactly defined, and it's not going to happen ever. We're probably going to get closer and closer, but in the end we're still observing ourselves and reality from the inside.

I believe the goal of the article was to describe the differences between computers and humans, from the viewpoint of humans. What seems to be more popular these days is to start from the viewpoint of a computer and pretend humans are the same thing.

Re: Why general artificial intelligence will not be realized

#337

Post by codr7 » Sat, Jul 11, 2020, 11:16 AM UTC

Strilanc wrote: post #194

TheOtherHobbes wrote: post #181

Earlier quoted context omitted.

The point isn't bike riding. Bike riding is an example of a class of problems, and your solution begs the question. The point is that humans and computers operate differently. The human approach is based on adaptive experiential heuristics. The computer approach is based on explicit formalism. (Even in neural networks, there's still a formal model. It's just made of weightings instead of logic paths.) The epistemolog…

> The problem isn't getting a computer to ride a bike, it's getting a computer to learn to ride a bike how a human learns. This is just a more sophisticated way of assuming that there's something human-intrinsic about bike riding. Human-like is not the only way to approach doing or learning. Whatever works, works.

But humans learn by themselves, from observing the world around them and drawing conclusions. Which is a major advantage and a core problem in strong AI.

The only reason it works is that a lot of humans put a lot of effort into more or less figuring out exactly how to ride a bike. It doesn't generalize at all, teaching the same robot to ride a skateboard would mean doing it all over again.

Re: Why general artificial intelligence will not be realized

#338

Post by codr7 » Sat, Jul 11, 2020, 11:32 AM UTC

michaelmrose wrote: post #96

lurquer wrote: post #87

Earlier quoted context omitted.

I'm not arguing that it can't. I was merely pointing out the proper 'burden of proof.' The article was criticized for failing to demonstrate that something can't be done... that's not fair. The burden would be on the proponent of the proposition that a machine can attain general AI. That's all. Perhaps there could be general AI... I'm not saying it can't be done. I would point out, though, that IF it is to be done, i…

Maybe you are stuck on the notion of a computer as a silicon chip. Biological entities are just a special case of machine ergo it is already proved that a machine can attain general AI.

Says who? You are subjectively observing yourself and reality from the inside. How can you be so sure?

There is no proof, or we would not be having this discussion over and over again.

Re: Why general artificial intelligence will not be realized

#339

Post by vidarh » Sat, Jul 11, 2020, 11:36 AM UTC

bena wrote: post #90

vidarh wrote: post #69

Earlier quoted context omitted.

The strong indication that it can be done is that brains exists and grow guided by machinery that is computationally limited by the fact it takes up exceedingly little space, and we by extension can put an upper bound on the computational capacity of that machinery. That there is a tough engineering job involved in reproducing such machinery so we can grown brain matter at will, and that we'd prefer our computers to…

"It exists, therefore we can make it." I don't find that line of thought compelling. Although you can argue that we already do it because we, as a species, reproduce. It could be quite possible that our level of intelligence is not reproducible by the means we have chosen. It could be that there is something inherent in our entire composition that allows for our particular expression of intelligence and consciousness…

It's indeed quite possible it is not reproducible by building digital computers out of silicon, sure, but that does not prevent us from picking another way of building them.

I'm very much sympathetic to the idea that there could potentially be aspects about the specific structure of a brain that is necessary to produce intelligence, or at least necessary for sentience. But does not inherently make it impossible to reproduce. For it to be impossible to reproduce a brain, there would need to be something in the cellular machinery used to construct a brain that we can not pick a party and replicate.

As a reminder, the argument made in the article is that it is impossible even in principle, not that it is impossible to do it using a digital computer of typical current architecture.

Re: Why general artificial intelligence will not be realized

#340

Post by d3nigma » Sat, Jul 11, 2020, 11:49 AM UTC

I disagree with Dreyfus's view about strong ai. It's like saying that all lifeforms in the universe are carbon based, because we are carbon based lifeforms. But that's not necessarily true. I am a huge fan of bio-inspired engineering, however, and I think imitating/simulating "growing up" could be a reasonable approach to human-like intelligence. This phase of life is key for our understanding of the world. We define and refine our values and evolve from a blob of cells to a conscious being by observing our surroundings and acquiring huge amounts of knowledge. If this concept (refined by evolution over the last million years) is good enough for humankind, it is certainly good enough for human made intelligence. The biggest problem here is efficiency and complexity. We still don't know all mechanisms of our brain and we are not capable (yet) to simulate 100 billion neurons, each connected to up to 30.000 other neurons. That's the main problem here!