When a researcher invokes Kolmogorov complexity in this way, it's a sign either of a basic misunderstanding, or simple fraud. KC *is not computable*—it should be an embarrassment if someone uses it in an explanation.
-
-
Replying to @SimonDeDeo @dwilliams999 and
"weights reward summable measures by the Kolmogorov complexity of the environment" is as sensical a statement as "weights reward summable measures using the first three even primes". No such task is possible, or even conceivable.
2 replies 0 retweets 1 like -
Those are strongly judgmental words, a sign of a sore topic. However, your assertion is incorrect. It’s true that Kolmolgorov complexity is not explicitly computable, but it is well defined. There is a narrative specification for our world, though it may not be compressible.
2 replies 0 retweets 1 like -
Since Solomonoff published his seminal papers the question became even more subtle. It's not about incomputbility in principle but weak vs strong effectivity.
1 reply 0 retweets 2 likes -
Happy to Google search more. Do you have a link to a canonical paper you recommend for getting started with the Solomonoff formulation of strong effectivity?
1 reply 0 retweets 0 likes -
A Formal Theory of Inductive Inference, Part I and Part II. You can find them here: http://raysolomonoff.com/publications/pubs.html …
2 replies 0 retweets 2 likes -
Replying to @NegarestaniReza @dwilliams999 and
But the Solomonoff method is show to be on the weak range of effectivity so as Hutter's thesis.
1 reply 0 retweets 0 likes -
It seems like everything I read about defining intelligence is weak. Even the measure of intelligence by F. Chollet boils down to moving around colored boxes with an implicit understanding that his measure is about function finding for small samples of visual effects.
1 reply 0 retweets 1 like -
Fundamental issue: how can you distinguish between quantitive and qualitative compression? But then there is also the technical problem of defining effective computation in the proper way as someone like Tom Sterkenburg has done it. I think Simon is right but we are still in dark
3 replies 1 retweet 2 likes -
How about this thought experiment: if you were the admiral of a Star Trek fleet, and you were tasked with visiting a new world and documenting all of the “intelligent” beings, what sensors and criteria would you use?
1 reply 1 retweet 2 likes
Good early Kantian question (ex. anthropology from a pragmatic point of view). I must think about the scope of this question.
-
-
Do they throw good dinner parties? http://tuvalu.santafe.edu/~simon/rules_social_minds_NEW.pdf …
1 reply 0 retweets 2 likes -
There is also this one which one of the early examples of Kant's fascination with extraterrestrials and the outside view of agency: https://www.amazon.com/Universal-Natural-History-Theory-Heavens/dp/1935238914 …
1 reply 0 retweets 2 likes - 1 more reply
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.