Friday, March 02, 2007

AI and Google

Here is a little video snippet of Larry Page talking about AI at the recent AAAS meeting. He points out that our genetic information is about 600MB compressed, so smaller than any modern operating system. The connection between compression and AI has been made by many people -- what is intelligence, after all, if not an algorithm capable of taking in data (observations about the world) and turning it into predictions about what will happen next? Prediction (or, equivalently, modeling) is nothing more than compression! Newton's laws plus some initial data compress all the information about the trajectory of a spaceship -- trading bits of stored information (the spacetime coordinates of the trajectory) for CPU flops (necessary to uncompress the trajectory from the initial position and velocity -- i.e., evolve it forward in time).

Page guesses that AI will result more from "lots of computation" than from "whiteboard stuff" (i.e., we won't really "understand" how it happens from a theoretical or analytical perspective) and that "we aren't as far off as many people think"!

A lot of people like to speculate that Google is working like mad on AI, and indeed certain related problems like machine translation, pattern recognition and, of course, search, are things they devote a lot of resources to. However, the vast majority of their 10,000 employees (yes, the number really has been doubling every year for a while) are working on just keeping the existing services up and running. There isn't yet a blue-sky, Bell Labs-like research arm at Google.

See here for previous related posts. I have a bet with a former PhD student about machines passing a strong version of the Turing test in the next 50 years.

7 comments:

Anonymous said...

I hope that Google will focus more on improving their core business (= search).
e.g. image search is pretty bad compared to web search (e.g. in terms of stale pages),
similar critic applies to their blog specific search etc.
lot of room for improvement, before
we need to talk about the fancy stuff like AI, imho.

Dave Bacon said...

Is your bet for or against?

What will happen to physicists if you do build a strong AI system...give it a few "Moore" doublings and it will be able to beat the best physists? :)

Steve Hsu said...

I bet that no machine passes the strong version of the Turing test in the next 50 years. Strong means it has to fool me over a very long period of time -- long enough for me to try to teach it quantum mechanics, for example. (And it has to simulate a *smart* human.)

The "strong singularity" believers think we'll all be useless and irrelevant once machines pass the point described above -- they'll go on exploring things we can't comprehend, and, if we're lucky, keep us around as pets :-)

Steve Hsu said...

wolfgang: yes, there are a lot of things for them to improve. That's what most of the 10k employees are doing! They are sitting on $11B in cash now. Very hard for anyone to catch them now...

Dave Bacon said...

"I bet that no machine passes the strong version of the Turing test in the next 50 years."

Cool. I'm no fan of the singularity (it's a far too simplistic model of how future technological change will impact the world, IMHO), but I'd be willing to bet on strong A.I. in 50 years.

JaY said...

Word or character recognition software uses relationship logic, which is baby steps of AI. Breaking past the math logic is a pretty steep incline.

On or off...

wish it was that simple for me. Instead my thinking goes a little further than a instant result.

The damn what if's.

Then you get to pick the one that will lead you to what you think will give you an outcome close to your goal.

Now even if you are able to get a computer to do this. The problem now becomes... without self awareness and self preservation... the computer will never have true AI. Its just a suicidal machine... without a care in the world.

Just my thought on the issue...

losingthegame said...

I think that the general idea at google is to make a search engine that functions like/as quickly as a human brain. Ultimately, the internet can return information directly to your brain at the rate of firing neurons.
The Atlantic Monthly just had an article on how Google was making us stupid. (I argue smarter, though much more shallow.) I believe that it quoted a Goole exec rhapsodizing the possibilities of a brain on computers. So maybe the Google intention is not to make a smarter computer, but, through using computers, make a smarter brain?
Peace and love.

Blog Archive

Labels