Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI

blog.samaltman.com

Post reply on HN
Print view View original

281–290 of 317 posts

Re: AI

#281

Post by voyou »

twoodfin wrote: post #112
Houshalter wrote: post #80

Earlier quoted context omitted.

Not really. Just pick your favorite AI problem and see how well it does on that. Pick a bunch of AI problems and see how well it does on all of them. Weight the algorithms by simplicity if you are worried about it over-fitting.

With this hypothetical infinite speed computer you will get solutions that are perfect matches for your test cases but essentially random for all other inputs.

Exactly. Saying that with infinite CPU power you could brute-force a solution is like saying the Library of Babel[1] contains every book ever written. True, in a sense, but not as useful as you might think.

[1]: http://jubal.westnet.com/hyperdiscordia/library_of_babel.htm...

Re: AI

#282

Post by tawan »

kevinalexbrown wrote: post #132

If the only goal is an 'artificial' consciousness, it might be more prudent to consider a functional definition of what consciousness is and try to build that. We didn't make computers by modeling how individual neurons perform mathematical calculations. On the other hand, if you want to go the biological route, there's some awesome work to be done. If I were to study consciousness, here's the question I would ask: h…

To question what consciousness is and how we define it is a good starting point. And because there is so much uncertainty and no overall consensus makes it seem to be very hard.

Somehow I fear the only satisfying implementation of AGI has to end in a skynet scenario. We as human will only accept intelligence as general when it is at least as intelligent as humans. But that would mean that we have to accept that we can't control it for sure. In fact only a machine that is able to rebel can be considered to be general intelligent. So I am not surprised that consciousness is not defined that way, because pursuing to build a machine capable of consciousness would mean to build a possible enemy.

Re: AI

#283

Post by loup-vaillant »

crassus wrote: post #256
edanm wrote: post #201

Earlier quoted context omitted.

One of the major realisations from reading Less Wrong is taht Intelligence and Goals/Desires have nothing to do with each other. An AI might be much more intelligent, but it has a goal system that basically says: "Make as many paperclips as you can". Everything it does will be with the singular purpose of making more paperclips. Not music, math, sport, culture or anything else that we think of as good. Not "help save…

If our approach to AI is to model the human brain (i.e. passing the Turing test), this probably won't happen. Friendly AI is a silly research project at this stage of AI research. It's like trying to figure out how to make horseless carriages safe before you have an internal combustion engine, or even know what one is.

Before the first Nuke (or the first H-bomb, I don't remember) was fired, there was a study about the risk of burning up the entire atmosphere. See, at the time, they were quite confident the Nuke would just be a huge bomb, but there was this little uncertainty they needed to sort out. In the end, they concluded that firing the Nuke would not burn up the whole atmosphere. It didn't.

The lesson is this: we had only one try. If nukes did cause the atmosphere to burn up in a giant blaze, we would all be dead by now. If you do something, anything, you better make sure it won't kill us all.

Horseless carriages? Sure, these might kill a few people, here and then[1], but we're pretty sure they won't kill us all in one blow.

Intelligence on the other hand is way more dangerous. Human intelligence designed Nukes in the first place remember? AI can do way worse. Even if we model it after the human brain, if it's smart enough to do the same as we did, then it will be able to model another such AI, only slightly better, and so on until it takes over the world. "Taking over the world" may sound enormous, but it really isn't. Imagine for a minute a small group of cavemen vs an army of chimps. Well, if you give the cavemen a chance to prepare, the chimps are toast: the cavemen have spears, fire, better communication… Now imagine an AI imagine the AI is smarter than us by the same margin we're smarter than chimps. Same thing: if it's not safe, we're toast.

[1]: http://www.statisticbrain.com/car-crash-fatality-statistics-...

Re: AI

#284

Post by voyou »

Houshalter wrote: post #80
twoodfin wrote: post #50

Earlier quoted context omitted.

[automated test] is the hard part.

Not really. Just pick your favorite AI problem and see how well it does on that. Pick a bunch of AI problems and see how well it does on all of them. Weight the algorithms by simplicity if you are worried about it over-fitting.

"Pick a bunch of AI problems and see how well it does on all of them."

The problem is, you have infinite potential algorithms and a finite number of tests. This means you'll necessarily get algorithms that pass all your tests but fail at least one other test of intelligence. Because of this, your tests won't actually let you discover which of the generated algorithms is intelligent.

Or, you could have an infinite number of tests, but if you have infinite tests for intelligence, you effectively already have an algorithm for intelligence (for any problem, just look up the answer in your list of tests), so, again, brute-forcing a solution isn't helpful.

Re: AI

#285

Post by loup-vaillant »

membrain wrote: post #231
edanm wrote: post #86

If anyone is interested in AI, I highly recommend joining Less Wrong, a community started by AI researcher Eliezer Yudkowsky. He started the community to convince people to focus on the "friendly AI problem". [1] I actually recommend that everyone read LW, but especially if you're interested in AI. [1] In a nutshell, the friendly AI problem is: assume we create an AI. It may rapidly become more intelligent than us, i…

> friendly AI problem nooo, please!! not any more of this kurzweil crap. guys wake up! This world is not some asimov sci-fi story. friendly ai I always get a headache when I read that term online. ppl seem to go crazy about the machines taking over earth idea but this whole debate is so utterly useless! If all the effort fapping to conscious AI and friendly AI would be put into concrete ai research (agi as well as ap…

> nooo, please!! not any more of this kurzweil crap.

Ray Kurzweil has little to do with this. When he talks about "singularity", he thinks about "accelerating change", with exponential growth everywhere, even beyond human intelligence scales.

Those who work on Friendly AI have another scenario: "intelligence explosion". Typically a self-improving AI that would grow itself into a super intelligence. You should read this introduction on the subject: http://intelligenceexplosion.com/

> This world is not some asimov sci-fi story.

Indeed. In Asimov's stories, the robots mostly behave, though in every story there is some problem with the way the laws of robotics apply.

In the real world, the laws of robotics don't work, and we would at best be put "safe" into cushioned rooms so we can't hurt ourselves. As for emotional hurt, drugs and lobotomy are extremely efficient remedies.

Re: AI

#286

Post by loup-vaillant »

querious wrote: post #218
electrograv wrote: post #190

Earlier quoted context omitted.

I'm sorry, but AGI vs Alchemy is a poor analogy with almost no parallels. I see this as a rather blatant strawman argument. Flying machines vs AGI makes a much better analogy. Natural flying machines (birds) existed in nature, before artificial ones (planes) were invented. Artificial manned flight was speculated as a possibility for centuries with no macroscopic working examples, and heavily criticized as a transport…

The question that comes up for me at this point is whether there is much that is dispensable about humans when it comes to exhibiting intelligent behavior (running on 100 watts, no less). It turned out that for abstracting useful flight dynamics based on birds, there was a rather simple rule: lift > weight. Sure, reducing it to that simple formula may not help you build a machine that maneuvers as well as birds and i…

You're falling into the same naturalistic fallacy the likes of Clement Ader fell into: thinking you had to imitate nature to get flight (or AI) done. I trust the underlying principles of intelligence are much simpler than current implementations.

It's probably not going to be easy, tough.

Re: AI

#287

Post by marvin »

nnq wrote: post #207
hooande wrote: post #173

Many people see Artificial General Intelligence as a panacea. The idea is, "We'll create artificially intelligent scientists who will solve all of our other problems for us!". I think that future generations will look back on this as a modern version of alchemy. If blogs had existed centuries ago I'm sure that people would have called the transmutation of metal the most important trend of their time. This isn't to de…

It's funny that most people that are actually working on AGI have stopped even bothering to provide arguments for why they think AGI it's possible to the general public, because they know they're close and that "they'll just prove that it's possible to build it by fucking building it" instead of doing armchair thinking, while AGI detractors don't lose any opportunity to launch pseudo-arguments like missiles :) ...jus…

>It's funny that most people that are actually working on AGI have stopped even bothering to provide arguments for why they think AGI it's possible to the general public, because they know they're close and that "they'll just prove that it's possible to build it by fucking building it" instead of doing armchair thinking, while AGI detractors don't lose any opportunity to launch pseudo-arguments like missiles :)

If you've read some history, it's clear that this is exactly what the Wright brothers did when they pioneered powered flight. Three things makes Kitty Hawk a good place to test airplanes: You'll survive if you crash in the sand dunes, there is enough steady wind and there are no newspaper reporters around to make fun of your failures.

Re: AI

#288

Post by bencollier49 »

Does anyone have a list of public companies (especially small ones) which are involved in AI? There don't seem to be many besides the big players.

Re: AI

#289

Post by axilmar »

> Andrew Ng, who worked or works on Google’s AI, has said that he believes learning comes from a single algorithm - the part of your brain that processes input from your ears is also capable of learning to process input from your eyes. If we can just figure out this one general-purpose algorithm, programs may be able to learn general-purpose things.

As I wrote a few days ago here:

https://news.ycombinator.com/item?id=7217967

"The way intelligence works, in my opinion, is this:

1) experiences are stored in the brain. Experiences contain inputs from the 5 senses as well as the sense of danger/satisfaction at that point.

2) at each given moment, the brain takes the current input and matches it against the stored experiences. If there is a match (up to a threshold), then the sense of danger/satisfaction is recalled. Thus the entity is able to 'predict', up to a specific point, if the outcome of the current situation is bad or good for it, and react accordingly.

The key thing to the above is that the whole process is fused together: the steps for adding new experiences, matching new experiences and recalling reactions is fused together in big pile of neurons."

Re: AI

#290

Post by mannykannot »

rdl wrote: post #36
elwell wrote: post #10

Critically speaking, this article doesn't add anything. I don't know why I read it. Can anyone explain why they upvoted?

For me, "There are certainly some reasons to be optimistic. Andrew Ng, who worked or works on Google’s AI, has said that he believes learning comes from a single algorithm - the part of your brain that processes input from your ears is also capable of learning to process input from your eyes. If we can just figure out this one general-purpose algorithm, programs may be able to learn general-purpose things." is what w…

"Andrew Ng, who worked or works on Google’s AI, has said that he believes learning comes from a single algorithm."

As algorithms can be combined, the existence of any set of algorithms satisfying this goal would automatically imply the existence of a single algorithm incorporating all of them.

If a sufficiently-detailed physical simulation of a human's brain satisfied this goal, then that would be one such algorithm.

Post reply on HN