Huh? Kepler was the guy who theorized that the planets lived on spheres whose sizes were determined by nesting of the Platonic solids. Later, finding that the orbits didn't fit the spheres, he finally struck on the idea that the orbits were ellipses. He believed that the motion of the planets was caused by a force radiating from the sun, which grew weaker with distance, giving rise to his law of equal times for equal areas. But he thought the force pushed the planet along its path -- he never invented gravity (though he set the stage for Newton to do so, as Copernicus and Tycho had set it for him.
Astronomy is perhaps the closest to compression of the sciences, since it's difficult to do experiments with planets. However, even 15th-century astronomy used some non-trivial instrumentation, notably astrolabes and the earliest telescopes (and look at this: http://nfo.edu/italy/meridiana.html). So there's a significant amount of work going on there, not just chewing over a given set of data. But to repeat: Kepler's major advance was to trade one theory for another because the first one did not fit the observational data. That's about as empirical as it gets. As far as elegantly compressing one's history, yes, we do that -- but my point is that we actively intervene in the world based on the theories we produce thereby, in such a way as to optimize our expected history thereafter -- both for the further production of theories, and other purposes. If Zen has any useful advice for AGI, I'd claim it was in its rejection of endless philosophizing :-) josh On Thursday 29 May 2008 04:08:14 pm, Tudor Boloni wrote: > Josh, IF we remove the empiricism bias in "future sensory histories" (think > Kepler and the approach he took to discover gravity, such as: what can > explain observed patterns, knowing our mind is always BLINDED by the senses > (similar to zen teachings), he had to negate the tendency to build a > hypothesis that relied on the "shadows of reality" (i.e. observable > phenomena limited by our arbitrary senses), and then arrive at the most > compact hypothesis that could not be disproved at the time) THEN this seems > to be a very powerful roadmap: > > Recipe for Intelligence: elegantly compress one's history, then hypothesize > all reasonable permutations of past events (include permutations of time > scales also), elegantly compress hypotheses until not disprovable > > The more intelligent a system the smaller the information footprint... > pretty zen > > Tudor > > > > On Thu, May 29, 2008 at 8:41 PM, J Storrs Hall, PhD <[EMAIL PROTECTED]> > wrote: > > > I would demur. There is a huge overlap in the techniques used in > > compression > > and those in intelligence. However, the significant difference is that > > intelligence, in interacting with the real world, has a motor component > > which > > allows it to select among possible future sensory histories in a way that > > is > > not a part of the standard formulation of compression. This leads to > > techniques such as experimental science -- the source of 99% of current > > human > > knowledge. > > > > Josh > > > > > > On Thursday 29 May 2008 02:27:43 pm, Matt Mahoney wrote: > > > --- Tudor Boloni <[EMAIL PROTECTED]> wrote: > > > > > > > as a side note, does anyone else feel that intelligence and compression > > > > (or less formally the ability to summarize) are identical? > > > > > > Yes, http://cs.fit.edu/~mmahoney/compression/rationale.html<http://cs.fit.edu/%7Emmahoney/compression/rationale.html> > > > > > > See also Hutter's work on AIXI which proves the equivalence. > > > > > > -- Matt Mahoney, [EMAIL PROTECTED] > > > > > > > > > ------------------------------------------- > > agi > > Archives: http://www.listbox.com/member/archive/303/=now > > RSS Feed: http://www.listbox.com/member/archive/rss/303/ > > Modify Your Subscription: > > http://www.listbox.com/member/?&; > > Powered by Listbox: http://www.listbox.com > > > > > > ------------------------------------------- > agi > Archives: http://www.listbox.com/member/archive/303/=now > RSS Feed: http://www.listbox.com/member/archive/rss/303/ > Modify Your Subscription: http://www.listbox.com/member/?&; > Powered by Listbox: http://www.listbox.com > ------------------------------------------- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244&id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
