The journal’s web site is here, from where I’ve just downloaded a copy of the paper. I don’t know if it’s freely available (my university has a subscription), but if anyone wants it and can’t get it from the web site, send me an email address to send it to. (EDIT: Now online, see my later comment.)
The paper describes itself as the third in a series, of which the first appeared in the same journal, volume 19, pp.189-249 (also downloaded). The second is in a volume called “Machine Learning”, which you can find here, but I haven’t checked if the whole book is accessible. (EDIT: sorry, wrong reference, see later comment.)
Personally, I’m deeply sceptical of all work that has ever been done on AI (including the rebranding as AGI), which is why I consider Friendly AI to be a real but remote problem. However, I’ve no interest in raining on everyone else’s parade. If you think you can make it work, go for it!
Isn’t the second paper in the series the one immediately before the Eurisko paper, “Theory formation by heuristic search: The nature of heuristics II: Background and examples” by Lenat, volume 21, page 31? Can you download that one?
Yes, it is, my mistake. I now have all three papers. I’ve temporarily put them up in the directory here, a URL to which you will have to add lenatN.pdf for N = 1, 2, or 3. (I’m avoiding posting the complete URL so that Google won’t find them.)
I’d like to echo cousin_it’s thanks, I downloaded them as well.
I haven’t gotten to read much yet, but I’ve also run into the problem he’s mentioned with theoretical computer science papers being too vague to write code, let alone include it. (Marcus Hutter and Juergen Schmidhuber, I’m looking in your general direction here.)
You’re having trouble figuring out how to implement AIXI? I saw Marcus write it out as one equation. Perfectly clear what the main loop looks like. All you need is an infinitely fast computer and a halting oracle.
+5? Yikes! People, it’s clear Eliezer_Yudkowsky is joking. There are no infinitely fast computers or halting oracles, and an equation is not the same thing as code, let alone pseudocode.
In any case, AIXI isn’t my main complaint in that department. I’m thinking more of
Richard, thanks a lot! Downloaded all three articles. (No, ScienceDirect doesn’t let me download them—the link says “Purchase PDF”.)
First surface impression: none of the three papers are as specific as I’d like. After a skim I still have no idea how EURISKO’s main loop would look in pseudocode. Will try reading closer.
Presumably he meant LW’s direct messaging, which you can use by clicking on the little envelope under your karma score to get to your inbox, and then clicking ‘compose’.
That’s a funny parallel to my own situation. Most hits on the first few pages of Google for “Thom Blake” are me, but the runner-up is a historian from Australia.
The journal’s web site is here, from where I’ve just downloaded a copy of the paper. I don’t know if it’s freely available (my university has a subscription), but if anyone wants it and can’t get it from the web site, send me an email address to send it to. (EDIT: Now online, see my later comment.)
The paper describes itself as the third in a series, of which the first appeared in the same journal, volume 19, pp.189-249 (also downloaded). The second is in a volume called “Machine Learning”, which you can find here, but I haven’t checked if the whole book is accessible. (EDIT: sorry, wrong reference, see later comment.)
Personally, I’m deeply sceptical of all work that has ever been done on AI (including the rebranding as AGI), which is why I consider Friendly AI to be a real but remote problem. However, I’ve no interest in raining on everyone else’s parade. If you think you can make it work, go for it!
Isn’t the second paper in the series the one immediately before the Eurisko paper, “Theory formation by heuristic search: The nature of heuristics II: Background and examples” by Lenat, volume 21, page 31? Can you download that one?
Yes, it is, my mistake. I now have all three papers. I’ve temporarily put them up in the directory here, a URL to which you will have to add lenatN.pdf for N = 1, 2, or 3. (I’m avoiding posting the complete URL so that Google won’t find them.)
I’d like to echo cousin_it’s thanks, I downloaded them as well.
I haven’t gotten to read much yet, but I’ve also run into the problem he’s mentioned with theoretical computer science papers being too vague to write code, let alone include it. (Marcus Hutter and Juergen Schmidhuber, I’m looking in your general direction here.)
You’re having trouble figuring out how to implement AIXI? I saw Marcus write it out as one equation. Perfectly clear what the main loop looks like. All you need is an infinitely fast computer and a halting oracle.
Couldn’t you implement a halting oracle given an infinitely fast computer, though?
So, that’s one requirement down! We’ll have this AIXI thing built any day now.
+5? Yikes! People, it’s clear Eliezer_Yudkowsky is joking. There are no infinitely fast computers or halting oracles, and an equation is not the same thing as code, let alone pseudocode.
In any case, AIXI isn’t my main complaint in that department. I’m thinking more of
Hutter’s fastest shortest algorithm for everything and AIXI-tl; and Schmidhuber’s provably globally optimal Goedel machines, speed prior, and ordered optimal problem solver
Toy implementations anytime, guys?
I think that most upvoters got the joke...
Richard, thanks a lot! Downloaded all three articles. (No, ScienceDirect doesn’t let me download them—the link says “Purchase PDF”.)
First surface impression: none of the three papers are as specific as I’d like. After a skim I still have no idea how EURISKO’s main loop would look in pseudocode. Will try reading closer.
Are you this Richard Kennaway? (You gave no apparent contact info.)
Presumably he meant LW’s direct messaging, which you can use by clicking on the little envelope under your karma score to get to your inbox, and then clicking ‘compose’.
Yes. Almost all Google hits for “Richard Kennaway” are for me. (Number 2 by a long way is a political scientist in New Zealand.)
That’s a funny parallel to my own situation. Most hits on the first few pages of Google for “Thom Blake” are me, but the runner-up is a historian from Australia.