[The Sabbaticus]

The quality isn't great, but it's an interesting perspective that Chomsky provides on amongst other things artificial intelligence, Google Translation, chess-bots and the notion of an imminent singularity (which he dismisses outright), suggesting instead that we should focus on global warming.

[youtube]0kICLG4Zg8s[/youtube]

Dr. Noam Chomsky is a famed linguist, political activist, prolific author and recognized public speaker, who has spent the last 60 years living a double life -- one as a political activist and another as a linguist. His activism allegedly made him the US government's public enemy number one. As a linguist he is often credited for dethroning behaviorism and becoming the "father of modern linguistics" (and/or cognitive science). Put together his accomplishments are the reasons why he is often listed as one of the most important intellectuals of the 20th century. And so I was very much looking forward to interviewing him on Singularity 1 on 1.

Unfortunately our time together was delayed, then rushed and a bit shorter than anticipated. So I was pretty nervous throughout and messed up some of my questions and timing. Never-the-less, I believe that we still had a worthy conversation with Dr. Chomsky and I appreciate the generous though limited time that he was able to grant me.

During our 28 minute conversation with Noam Chomsky we cover a variety of interesting topics such as: the balance between his academic and his political life; artificial intelligence and reverse engineering the human brain; why in his view both Deep Blue and Watson are little more than PR; the slow but substantial progress of our civilization; the technological singularity...

[quetzalcoatl]

Any sensible person realizes that the Singularity is nonsense. It is simply eschatology deprived of its spiritual underpinnings. While it is true that technology has a profound impact on human life, the things that matter most now are the things that mattered most in past centuries. The singularity won't feed your children. It won't put a roof over their heads. It won't pay your mortgage. And it can't deal with the profound sense of disquiet that comes from being at the mercy of people and forces that do not respect your humanity.

[Saeko]

It don't think Chomsky knows what he's talking about when it comes to AI.

At 8:00 he seems to be saying that Deep Blue's chess playing ability is derived solely from fifty grand masters thinking really hard about chess beforehand and then "programming" their strategies into Deep Blue. In other words, he seems to be saying that Deep Blue is a mere reflex-agent that looks at the board and executes a pre-programmed sequence of actions.

But this is simply not true. Deep Blue, like all of its successors, uses heuristic functions and search algorithms. Given enough memory and speed, a chess AI could solve the whole game and at least draw every time. Making a program that does this would require no chess-playing ability whatsoever on the part of the programmers. The only part of Deep Blue's programming that was influenced at all by actual chess players are the heuristic functions it uses to judge whether one position is better than another. However, even naive heuristics (such as simply looking at the material each player has) are sufficient for very high level play.

That's why today, even a complete Chess novice can easily make a chess program that can play at the master level. If what Chomsky believed was true, this wouldn't be possible.

[quetzalcoatl]

But this is simply not true. Deep Blue, like all of its successors, uses heuristic functions and search algorithms. Given enough memory and speed, a chess AI could solve the whole game and at least draw every time. Making a program that does this would require no chess-playing ability whatsoever on the part of the programmers. The only part of Deep Blue's programming that was influenced at all by actual chess players are the heuristic functions it uses to judge whether one position is better than another. However, even naive heuristics (such as simply looking at the material each player has) are sufficient for very high level play.

What you say may be true, but HI (Human Intelligence) is not adequately described by heuristics and search algorithms. I would go further and challenge the Turing idea that any physical process is necessarily computable on any platform. It is quite possible that some physical processes are unique to a particular physical matrix and cannot be emulated.

...even a complete Chess novice can easily make a chess program that can play at the master level. If what Chomsky believed was true, this wouldn't be possible.

You don't go nearly far enough. Even a sufficiently large group of sailors stationed on rowboats and communicating by semaphore could emulate a high level chess player.

[Brother of Karl]

It don't think Chomsky knows what he's talking about when it comes to AI.

At 8:00 he seems to be saying that Deep Blue's chess playing ability is derived solely from fifty grand masters thinking really hard about chess beforehand and then "programming" their strategies into Deep Blue. In other words, he seems to be saying that Deep Blue is a mere reflex-agent that looks at the board and executes a pre-programmed sequence of actions.

But this is simply not true. Deep Blue, like all of its successors, uses heuristic functions and search algorithms. Given enough memory and speed, a chess AI could solve the whole game and at least draw every time. Making a program that does this would require no chess-playing ability whatsoever on the part of the programmers. The only part of Deep Blue's programming that was influenced at all by actual chess players are the heuristic functions it uses to judge whether one position is better than another. However, even naive heuristics (such as simply looking at the material each player has) are sufficient for very high level play.

That's why today, even a complete Chess novice can easily make a chess program that can play at the master level. If what Chomsky believed was true, this wouldn't be possible.

I think you mean brute force algorithms, not search algorithms. A chess program looks at all the possible moves, and then branches out all the possibilities for the opponent's move, and then back and forth a few more times, as far as the processing power of the computer can handle it. Then based on the best outcome of possibilities for each particular move, the computer will give each move a score. The move with the highest score is then executed. You're right that this is different than the way that chess grandmasters play, but you got the mechanics of it completely wrong.

Good chess players on the other hand use experience and heuristics, and memorize algorithms for certain circumstances. They also use psychological techniques to intimidate the opponent and/or predict the opponent's thought process. However, the computer's method will win every time. A human novice could, in theory, use the same method as a computer and beat a grandmaster, but it would many years to play out a single chess game at the speed that a human can repeat the same simple action.

Listening to that lecture, it's obvious that Noam Chomsky knows something about translation algorithms (he mentions that they are brute force algorithms, but nothing in science or philosophy has discovered a more fundamental thing about language that can be translated into a much more efficient algorithm for computers). However, he probably doesn't know about chess algorithms, he was just using it as a metaphor for what the interviewer was asking and shouldn't be taken as anything more.

[Saeko]

I think you mean brute force algorithms, not search algorithms. A chess program looks at all the possible moves, and then branches out all the possibilities for the opponent's move, and then back and forth a few more times, as far as the processing power of the computer can handle it. Then based on the best outcome of possibilities for each particular move, the computer will give each move a score. The move with the highest score is then executed. You're right that this is different than the way that chess grandmasters play, but you got the mechanics of it completely wrong.

Good chess players on the other hand use experience and heuristics, and memorize algorithms for certain circumstances. They also use psychological techniques to intimidate the opponent and/or predict the opponent's thought process. However, the computer's method will win every time. A human novice could, in theory, use the same method as a computer and beat a grandmaster, but it would many years to play out a single chess game at the speed that a human can repeat the same simple action.

Listening to that lecture, it's obvious that Noam Chomsky knows something about translation algorithms (he mentions that they are brute force algorithms, but nothing in science or philosophy has discovered a more fundamental thing about language that can be translated into a much more efficient algorithm for computers). However, he probably doesn't know about chess algorithms, he was just using it as a metaphor for what the interviewer was asking and shouldn't be taken as anything more.

Brute force and search algorithms are not mutually exclusive. What you are describing is how a chess program actually works, but that is an informed search algorithm and not a brute force one so I don't even see what it is that I got "completely wrong" when you've pretty much repeated exactly what I said.

[One Degree]

The only purpose behind winning or losing a chess game is an emotional reaction. A human always wins because the AI gets no emotional feedback. Superiority in true/false reasoning serves no purpose for humans (unless you are foolish enough to believe it does), so why would we ever be in competition with AI?
Edit:
Computers were made possible by true/false reasoning. We have perverted that into believing we must live by true/false reasoning and compete with future AI's in this realm. It is very illogical. Why should we live and compete by a type of reasoning that was necessary to create computers?

[Il Doge]

This thing to understand about chess is that it's a man-made game with a limited number of possible moves, while in real life there is an essentially unlimited number of possible moves.