Forgot your password?
typodupeerror
Programming Technology

Can Curiosity Be Programmed? 269

Posted by samzenpus
from the killing-the-computer-cat dept.
destinyland writes "AI researcher Jurgen Schmidhuber says his main scientific ambition 'is to build an optimal scientist, then retire.' The Cognitive Robotics professor has worked on problems including artificial ants and even robots that are taught how to tie shoelaces using reinforcement learning, but he believes algorithms can be written that allow the programming of curiosity itself. 'Curiosity is the desire to create or discover more non-random, non-arbitrary, regular data that is novel and surprising...' He's already created art using algorithmic information theory, and can describe the simple algorithmic principle that underlies subjective beauty, creativity, and curiosity itself. And he ultimately addresses the possibility that the entire Universe, including everyone in it, is in principle computable by a completely deterministic computer program."
This discussion has been archived. No new comments can be posted.

Can Curiosity Be Programmed?

Comments Filter:
  • by Urza9814 (883915) on Thursday January 28, 2010 @12:47AM (#30929694)

    Basically - there's no way to store more information in a given area than what it already contains. In order to fully simulate the universe, at full (or greater) speed, you would have to know absolutely everything about absolutely every particle and subatomic particle, etc. And that includes the particles that make up the processor itself.

    It's like this: Say you have a 300 DPI printer. You print out a full page of text. Now, you want to fit all the information about that page into some sub-region of the page, printed on the same printer. Ok, so you say you can just shrink the text or encode it in binary or something, which is fine - except somehow also fit the information about the shrunken/encoded text in there. As you can see, you enter a recursive nightmare. And as your printer is a fixed resolution, you would quickly reach a point where any attempts to fit more information results in a blurred pixelated mess.

  • by Trepidity (597) <delirium-slashdotNO@SPAMhackish.org> on Thursday January 28, 2010 @12:52AM (#30929726)

    A minority of AI researchers have tackled the problem on and off, and even built some small-scale models of curious agents. One of the classic precursors is Doug Lenat's 1977 system Automated Mathematician [wikipedia.org], which shifted from the idea of using AI to prove theorems, to instead looking for theorems that would be interesting if they were true (it didn't actually prove them; it was an interesting-conjecture generator). Essentially a model of mathematical curiosity.

    Some interesting more recent work is a 2001 thesis [usyd.edu.au] that modeled curiosity as a social phenomenon in societies of agents, where agents try to find things that are: 1) new enough to interest its fellow agents; yet not 2) so new that they were incomprehensible in its cultural context.

    (I'm an AI researcher, though not precisely in this area.)

  • by BitZtream (692029) on Thursday January 28, 2010 @01:12AM (#30929840)

    Yes, I agree, I should have specified.

    We will not be able to simulate in real time or faster.

    However, glossing over bits means you are also, wrong, however so slightly.

  • Re:Show me the runny (Score:2, Informative)

    by Internalist (928097) <fred.mailhotNO@SPAMgmail.com> on Thursday January 28, 2010 @02:50AM (#30930282) Homepage

    No, he knows and has explicitly stated in a few places that it's uncomputable, in much the same way that Kolmogorov Complexity is uncomputable, but an interesting and potentially useful theoretical construct, nonetheless.

    This vein of Schmidhüber's work is more or less descended from Solomonoff's work on induction and Chaitin's Algorithmic Information Theory stuff (the line of descent is less explicit with the latter), and a bunch of Schmidhüber's descendents, most prominently his student Marcus Hutter [hutter1.net] and *his* student Shane Legg [vetta.org] have taken this ball and run with it in interesting ways.

  • by HateBreeder (656491) on Thursday January 28, 2010 @06:49AM (#30931482)

    You made so much sense in your previous post.. too bad you had to make this one as well.

    Simulating the universe from within the universe is impossible - regardless of the rate, as your simulated universe should contain the simulation itself.... which is a positive feedback loop.

  • by sourcerror (1718066) on Thursday January 28, 2010 @08:54AM (#30932010)
    I guess not. At the atomic level there's a lot of randomness.
    Einstein wasn't quite statisfied with these consequences, that's why he said: God doesn't play dice. [hypography.com]
  • Re:Not informative (Score:3, Informative)

    by HateBreeder (656491) on Thursday January 28, 2010 @10:21AM (#30932812)

    err, no...

    your analogy is so bad.

    i don't have the will to get into this.

    just have a long think... and figure out if a program printing itself is the same as a program simulating itself.

  • by Anonymous Coward on Thursday January 28, 2010 @12:17PM (#30934660)

    Here is a short version of his incredible talk, only 10 minutes:
    youtube.com/watch?v=Ipomu0MLFaI

  • by MobyDisk (75490) on Thursday January 28, 2010 @12:30PM (#30934914) Homepage

    but if you could account for all the variables with enough precision; angle of the coin, angle of the thumb, force of the flip, distance to the floor, etc, you could likely predict each and every toss.

    Unfortunately, you can't. :-( That's called the hidden variable theory [wikipedia.org]. It has been proven that there can be no set of information that could be used to compute quantum randomness.

    Einstein refused to believe that, and proposed the EPR thought experiment as a way to disprove it. Unfortunately for him, he died before John Bell resolved the EPR paradox, finally disproving hidden variables.

Top Ten Things Overheard At The ANSI C Draft Committee Meetings: (9) Dammit, little-endian systems *are* more consistent!

Working...