4 ms·

To me impressive would be if an AI could be told a story, such as The 3 little pigs and then be able to reason about it speculatively, e.g. answering questions

by lottin 10y ago

To me impressive would be if an AI could be told a story, such as The 3 little pigs and then be able to reason about it speculatively, e.g. answering questions such as "Why do you think this character did this and that?". This would should that the machine has developed an intelligence. So far, all AI can do is pattern recognition, which can be very useful for many tasks, but in my opinion it shouldn't be called "intelligence" because it isn't. Far from it.

dasboth 10y ago
I agree about the story. Text generation is already well within our capabilities, but it would be much more "intelligent" if the algorithm understood concepts of plot and characters to come up with a coherent piece of writing. There are already algorithms for summarising sporting and news events, but that's pretty formulaic, not creative like writing a story.
- dmreedy 10y ago
  I want to be careful here; yes, semi-coherent text generation is possible under the current state of the art, but there is no evidence right now that it 'on the right track', as it were, to solving the harder problem of binding semantics and language that you allude to; it may be the case that our current techniques for text generation, inverses of our current techniques for text 'understanding' (and I use the term loosely) are just sufficiently powerful to pick up whatever signal is available from the surface-forms of language, but the information is too scarce, the entropy too high for that next big step of 'why'
summerdown2 10y ago
This is actually one of the things AI is currently being developed for, and may not be that far off. Again, it won't be reasoning as we understand it, more a set of state tables that allow inferences. But people are definitely trying to develop AI that can take a description of an event, and break it down into legal concepts such as "who was at fault here?"
bbctol 10y ago
I agree that that would be impressive, but beating a human being at Go is also very, very impressive, and had been considered a high benchmark for many years. I can't help but think that after a machine can do basic story analysis, there will still be people saying that this is just a trick, that doesn't show any "intelligent" analyses of the story. AI has a long way to go, but I don't think we should dismiss intelligence as an either/or binary, and call things that are not "intelligence" mere "pattern recognition." Strategy games are widely accepted across cultures as a mark of intelligence when humans do them; you'd be very comfortable saying a human chess or Go grandmaster was intelligent. There was a long time when people were dismissing basic AIs as mere toys by saying they'll never beat a human at Go. There's a little goalpost-moving any time someone responds to an AI breakthrough by saying "You know what would be really impressive..."
- dmreedy 10y ago
  The Go victory is definitely impressive, for what it was, but I think this is a pretty unfair comparison; it's an entirely different class of problem, and a strictly less difficult one at that (at least given current knowledge of the mind). It seems likely to me that strategy games are accepted as an intelligence gauge partly because they're so hard for humans to do; thinking so many plys ahead by exploring and pruning across the game tree seems really hard for humans. For some reason, Natural Language and its semantic binding (I think the notion that these two things are often treated separately is part of a lot of the bigger deltas between expectation and reality when it comes to 'AI'. Honestly, they're probably the same thing anyway) doesn't seem really hard for us. That doesn't strictly imply that 'strategy games' are strictly harder than 'natural language understanding'; in fact, as far as research goes, strategy games are easy to tackle, because we know all the rules; we designed them! And often, perfect information is available to all agents involved. It's about as black and white as problems get. Sure you can't brute-force Go before the heat death of the universe, but it still theoretically -can- be brute forced. The rules of the game (known and accepted by all parties) encode the algorithm, and it's just a matter of reverse engineering it step-by-step for a given situation. What a brain is doing when it speaks or listens or thinks is currently a much bigger, much more primitive mystery. As far as we know right now, you can't just brute force sentence permutations until you hit a correct answer to a question like the one posed by the parent. So I don't call this goal-post moving (though acknowledge that the Vanishing Problem is systemic in AI Research and Pop Science); instead, I think this is more akin to the work in say, separating problems into classes of computational complexity. Finding the upper limits of things can be very helpful when we're dealing with such an old, deep, controversial problem.

by lottin 10y ago

dasboth 10y ago

I agree about the story. Text generation is already well within our capabilities, but it would be much more "intelligent" if the algorithm understood concepts of plot and characters to come up with a coherent piece of writing. There are already algorithms for summarising sporting and news events, but that's pretty formulaic, not creative like writing a story.

dmreedy 10y ago
I want to be careful here; yes, semi-coherent text generation is possible under the current state of the art, but there is no evidence right now that it 'on the right track', as it were, to solving the harder problem of binding semantics and language that you allude to; it may be the case that our current techniques for text generation, inverses of our current techniques for text 'understanding' (and I use the term loosely) are just sufficiently powerful to pick up whatever signal is available from the surface-forms of language, but the information is too scarce, the entropy too high for that next big step of 'why'

summerdown2 10y ago

This is actually one of the things AI is currently being developed for, and may not be that far off. Again, it won't be reasoning as we understand it, more a set of state tables that allow inferences. But people are definitely trying to develop AI that can take a description of an event, and break it down into legal concepts such as "who was at fault here?"

bbctol 10y ago

I agree that that would be impressive, but beating a human being at Go is also very, very impressive, and had been considered a high benchmark for many years. I can't help but think that after a machine can do basic story analysis, there will still be people saying that this is just a trick, that doesn't show any "intelligent" analyses of the story. AI has a long way to go, but I don't think we should dismiss intelligence as an either/or binary, and call things that are not "intelligence" mere "pattern recognition." Strategy games are widely accepted across cultures as a mark of intelligence when humans do them; you'd be very comfortable saying a human chess or Go grandmaster was intelligent. There was a long time when people were dismissing basic AIs as mere toys by saying they'll never beat a human at Go. There's a little goalpost-moving any time someone responds to an AI breakthrough by saying "You know what would be really impressive..."

dmreedy 10y ago
The Go victory is definitely impressive, for what it was, but I think this is a pretty unfair comparison; it's an entirely different class of problem, and a strictly less difficult one at that (at least given current knowledge of the mind). It seems likely to me that strategy games are accepted as an intelligence gauge partly because they're so hard for humans to do; thinking so many plys ahead by exploring and pruning across the game tree seems really hard for humans. For some reason, Natural Language and its semantic binding (I think the notion that these two things are often treated separately is part of a lot of the bigger deltas between expectation and reality when it comes to 'AI'. Honestly, they're probably the same thing anyway) doesn't seem really hard for us. That doesn't strictly imply that 'strategy games' are strictly harder than 'natural language understanding'; in fact, as far as research goes, strategy games are easy to tackle, because we know all the rules; we designed them! And often, perfect information is available to all agents involved. It's about as black and white as problems get. Sure you can't brute-force Go before the heat death of the universe, but it still theoretically -can- be brute forced. The rules of the game (known and accepted by all parties) encode the algorithm, and it's just a matter of reverse engineering it step-by-step for a given situation. What a brain is doing when it speaks or listens or thinks is currently a much bigger, much more primitive mystery. As far as we know right now, you can't just brute force sentence permutations until you hit a correct answer to a question like the one posed by the parent. So I don't call this goal-post moving (though acknowledge that the Vanishing Problem is systemic in AI Research and Pop Science); instead, I think this is more akin to the work in say, separating problems into classes of computational complexity. Finding the upper limits of things can be very helpful when we're dealing with such an old, deep, controversial problem.