Logic problem

February 19, 2009

D was telling me about the game called Go and how computers are not very good at playing it. Unlike chess, at which computers can beat a human even with half a processor

Anyway, D was reading something on the AAAS website and had this page opened in his browser.

And just as the Go-related conversation ended, I asked D jokingly:

“So who’s the sexy fox in the picture?”

“He’s a PhD” said D. “See, this is how you’re going to look like when you have your PhD.”

This entire piece of conversation reminded me of the Logic class in highschool when we were learning about syllogisms. Stuff like:

Nobody’s perfect.
I am nobody.
———————–
Therefore I am perfect.

So by applying the same rules of logic D came to the conclusion that when the letter P, H, D will precede my name and everyone [himself included] calls me doctor I will have a semi-bold head and a bushy beard. I see great days ahead!

Posted in Life, universe and everything

6 comments

Alexandra - March 4, 2009 - 7:28 pm

Phil, I re-read your post on Turing and Wall-e. Actually saying “it’s a matter of time” has somehow been imprinted in my brain somewhere. Maybe it’s what I want to believe, maybe it’s something I fear. (i guess it’s a little of both.) But I always tend to say that.
Then again, maybe the philosopher was right. Maybe we science will progress, greatly, but never enough. Maybe we’ll always be asymptotically close to AI but never quite be able to touch it!

Phil - March 4, 2009 - 5:35 am

> It’s only a matter of time

Then again, every other one of Dreyfus’ warnings have sort of become a reality. You’ll notice that research into A.I. these days is almost nonexistent, especially in the UK.

The one big problem is how to program common sense. Why is it that it’s so difficult for a computer to understand a children’s story when a 5-year old can get it?

I had a small article about this on my blog (Did a philosopher kill WALL-E?) last year.

Really interesting stuff. And one of the few times that philosophy has preempted modern science.

Alexandra - February 26, 2009 - 9:00 pm

Laura, my dear! I am so thrilled you reminded me. That was so funny! Oh, those were the days.

Phil, I swear, it’s exactly what my boyfriend told me. Except for the Dreyfus part. I really think in a few year’s time computers will beat us at Go. It’s only a matter of time. As you’ve said, they thought a chess-playing computer was a wild goose chase, but here we are…

Phil - February 25, 2009 - 5:06 pm

Sorry Alex, but since this is an interesting topic you’ve raised, I’ll go ahead and bore you to death.

Dreyfus, the great philosopher who critiqued the AI programs in his 1972 book “What computers can’t do,” basically said that chess computers would not become a reality because of the sheer number of possible games (greater than the number of atoms in the universe).

Given this, it’s interesting that the chess problem has been effectively solved (today, the top 5 Grandmasters in the world would probably struggle to beat the best program), but they’ve done so by making a program that combines brute force with a little bit of human intuition.

Problem with Go is that the number of possible games is enlarged by several orders. Will we do the same thing with Go as we did with chess? Maybe. Maybe not.

That said, it’s interesting that in a purely mathematical sense, Chess and Go are perfectly solvable games (this can be proven!) — meaning that it’s theoretically possible to generate an unbeatable strategy every single time, so in that sense, it’s a trivial pursuit.

laura - February 24, 2009 - 7:11 am

this is so funny, dear
i remember when i forgot my logic at home