Grandmaster Garry Kasparov on Artificial Intelligence • Hacker News Board

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#21

Post by lodi » Tue, Sep 26, 2017, 4:36 PM UTC

colorint wrote: post #13

ilaksh wrote: post #9

He was right a few years ago. But now there are several groups who have developed efficient capable online learning systems that don't require much data or iteration. When these and other existing types of cutting edge neural network advances such as techniques for avoiding catastrophic forgetting are combined with incremental training in diverse environments with general inputs and outputs, I believe we will see gen…

Do you believe that artificial intelligence will be capable of deciding, given an algorithm and a set of inputs, whether the algorithm will finish running?

Do you believe that human intelligence, given an algorithm and a set of inputs, is capable of deciding whether the algorithm halts or not?

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#22

Post by ilaksh » Tue, Sep 26, 2017, 4:57 PM UTC

colorint wrote: post #17

ilaksh wrote: post #16

Earlier quoted context omitted.

Lol. Ok so you are taking this halting thing and think that it means generally that no computer program can predict what a computer program will do, and therefore that proves that we will never have computers writing programs, and therefore never have general intelligence. You are really misinterpreting that stuff and not thinking it through. For starters take a look at the field of program synthesis. It's not AGI bu…

It turns out that the problem of deciding whether a predicate is universally valid in an axiomatic system is also undecidable (and, appropriately enough, called the decision problem). That is to say, declarative systems "corresponding to" general computation are also undecidable. Which isn't really surprising, since logical recurrence is isomorphic to functional recursion. Hence also why the examples I could readily…

Our brains and NNs don't work like that and don't need to decide things that are undecidable.

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#23

Post by xxxdarrenxxx » Tue, Sep 26, 2017, 5:45 PM UTC

lodi wrote: post #21

colorint wrote: post #13

Earlier quoted context omitted.

Do you believe that artificial intelligence will be capable of deciding, given an algorithm and a set of inputs, whether the algorithm will finish running?

Do you believe that human intelligence, given an algorithm and a set of inputs, is capable of deciding whether the algorithm halts or not?

I feel humans work on logic and where logic falls short, the emotion pseudo rng machine fills in the gaps.

I also feel an AI system should be event loop based, as humans whether consciously or subconsciously, always are "open" and become "notified" to async information from the outside world, (our senses)

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#24

Post by cgmg » Tue, Sep 26, 2017, 5:49 PM UTC

colorint wrote: post #15

cgmg wrote: post #14

Earlier quoted context omitted.

No, the halting problem is undecidable.

Yes, so given this, do you (i.e., people optimistic about AI) believe that computer programs will be able to generate meaningful, novel computer programs, given that even the most cursory subproblem is impossible? Obviously I don't just mean metaprogramming, but the sorts of things people want artificial intelligence to be able to do, the singularity and so forth.

> do you believe that computer programs will be able to generate meaningful, novel computer programs

Yes. There's a whole field dedicated to this called program synthesis. The undecidability of the halting problem does not preclude program synthesis.

> given that even the most cursory subproblem is impossible

What makes you think the halting problem is a 'cursory subproblem'?

Moreover, what makes you think humans can solve the halting problem?

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#25

Post by cgmg » Tue, Sep 26, 2017, 5:53 PM UTC

colorint wrote: post #17

ilaksh wrote: post #16

Earlier quoted context omitted.

Lol. Ok so you are taking this halting thing and think that it means generally that no computer program can predict what a computer program will do, and therefore that proves that we will never have computers writing programs, and therefore never have general intelligence. You are really misinterpreting that stuff and not thinking it through. For starters take a look at the field of program synthesis. It's not AGI bu…

It turns out that the problem of deciding whether a predicate is universally valid in an axiomatic system is also undecidable (and, appropriately enough, called the decision problem). That is to say, declarative systems "corresponding to" general computation are also undecidable. Which isn't really surprising, since logical recurrence is isomorphic to functional recursion. Hence also why the examples I could readily…

> the examples I could readily find of program synthesis are decidable problems, like deciding the maximum of two numbers or deciding membership in a list.

Can you find examples of human program synthesis for undecidable problems?

Re: Grandmaster Garry Kasparov on Artificial Intelligence

#26

Post by xxxdarrenxxx » Tue, Sep 26, 2017, 5:55 PM UTC

quantdev wrote: post #12

pawn wrote: post #8

One thing I'd be interested to learn is, how much of what makes the difference between an above average chess player and a Master or a Grandmaster can be tied to better decision making after looking 3 or 5 moves ahead, and how much is the Master/Grandmaster's ability to look 10+ moves ahead?

The looking 10 or even just 5 moves ahead thing is overstated and this is not actually how it works most of the time. Most GMs only calculate that far in the endgame. Before that, often looking 2 or 3 moves ahead is sufficient based on strategic elements or opening theory (which can't easily be understood by 'looking moves ahead'; they're things like, "this pawn is passed" or "my light squares will become very weak"…

It depends on how complex this is. If they played million chess games and saw so many moves then their brain linking memories is relatively trivial. This is in fact how computers do it.

Whats more interesting is how the moves came to be. How did one evaluate somewhere wothout reference evaluate a move or set of moves.

How does one "see" before calculating (checking for sensible conclusions).

More off these mysteries are found in the arts, since logic is common ground for the ration human mind and a computer. Emotions however are not.