6 ms·
MENACE
- kabdib 9y agoFred Saberhagen wrote a short story (Without a Thought) in which this little "computer" played a role. Published in 1963; I always wondered where the idea came from.
- dsr_ 9y agoAlso appears in the classic AI novel The Adolescence of P1 by Thomas Ryan.
- kabdib 9y agoThat's right, I'd forgotten. Still a good read . . . though I have to keep multiplying the numbers the author throws around for capacity (so many megabytes here and there) by about a billion, to keep from giggling. When the book was written, a megabyte was a really big deal.
- teddyh 9y agoWhat I usually think is that maybe they change the byte sizes in the future, so that a “byte” is no longer 8 bits. (Historical note: A “word” is nowadays usually four bytes, i.e. 32 bits, but older mainframe computer had varying word sizes; 36 bits for a word was not uncommon.)
- dsr_ 9y agoSadly, P1 is set in the 1970s and has figures appropriate for IBM systems of that era. Near the end, an experimental cryogenic memory system is attached: * Testing indicated that the results of a live experiment would be a data storage unit with 87 percent addressability and sufficient structural integrity to allow a map of the "dead" areas to be constructed. The interface being designed to link the cryostat to the 360/50 was geared to address just over 16 million bytes. It was understood that this capability would grossly understate the estimated final capacity of the cryogen. Modifications were under way to add eight more addressing bits to the computer. This would allow it to address over 4xl0^9 separate locations, or in excess of 4 billion addresses. This was expected to be more in keeping with the ultimate capacity of the cryogen at 150 millimeters diameter by 400 millimeters high. A six-inch cylinder, sixteen inches tall. * The System/360 series started with 24 bits of addressing and moved to 32 bits later on, so we're looking at a whopping 4GB. Even if the storage system uses 32 bit words, that's just 16GB.
- pavlov 9y agoThis suggests a thought experiment for those worried about superhuman AI. If you add more matchboxes to this setup while teaching it to play something else, is it going to become intelligent eventually? Maybe at a billion matchboxes? If that doesn’t seem believable, why be worried about machine learning algorithms running on digital computers instead of matchbox arrays?
- comex 9y ago> If you add more matchboxes to this setup while teaching it to play something else, is it going to become intelligent eventually? No, because the design uses a different matchbox for every possible state of the game. If you taught it to play chess, a billion matchboxes wouldn't be enough to get four moves in.[1] Yet digital computers can play chess. On the other hand, if you switched to a completely different design that uses matchboxes to simulate an artificial neural network… a design that would require opening each matchbox on each time step… …and pumped the number of boxes up to 100 billion (number of neurons in the human brain) or higher… …and ran the simulation for, perhaps, trillions of time steps (need to simulate evolution, not just the operation of the brain during a single lifetime)… …and changed the goal from playing a game to something that would require more general logical reasoning... …then who knows, it might well become intelligent (though it's hard to be sure). But at that point I'd say the lack of believability has more to do with the difficulty of performing 100,000,000,000,000,000,000,000 matchbox operations, than the likelihood of achieving useful results if you did. [1] http://www.bernmedical.com/blog/how-many-possible-move-combinations-are-there-in-chess http://www.bernmedical.com/blog/how-many-possible-move-combi...
- TeMPOraL 9y agoRelated: https://xkcd.com/505/ https://xkcd.com/505/.
- MereInterest 9y agoSounds like the Chinese room thought experiment [1]. [1] https://en.wikipedia.org/wiki/Chinese_room#Chinese_room_thought_experiment https://en.wikipedia.org/wiki/Chinese_room#Chinese_room_thou...
- pronoiac 9y agoUgh, reset_menace() is hairy - https://github.com/mscroggs/MENACE/blob/master/menace.js#L185 https://github.com/mscroggs/MENACE/blob/master/menace.js#L18...
- nerdponx 9y agoCan someone help me understand why it has to look like this? for(var i=0;i<9;i++){for(var k=0;k<i;k++){for(var m=0;m<k;m++){ for(var j=0;j<9;j++){if(i!=j && k!=j && m!=j){for(var l=0;l<j;l++){if(i!=l && k!=l && m!=l){for(var n=0;n<l;n++){if(i!=n && k!=n && m!=n){ pos="" dummy_moves=Array() for(var a=0;a<9;a++){ if(a==i || a==k || a==m){ pos+="1" dummy_moves.push(0) } else if(a==j || a==l || a==n){ pos+="2" dummy_moves.push(0) } else { pos+="0" dummy_moves.push(s[3]) } } add_box(pos,dummy_moves) }}} }}}}}}
- tyingq 9y agocontinue; could reduce the depth a bit, but it's still gnarly. https://gist.github.com/anonymous/f3607a28a4c8ad1322f13a2d10f28b5e https://gist.github.com/anonymous/f3607a28a4c8ad1322f13a2d10...
- pronoiac 9y agoI have a ridiculous urge to refactor this. It's enumerating all of the boards, legal or not. Judging by the letters, it's doing one player's moves at first, in a way that's trying to avoid repeats. If I were to do this, I'd have, well, menace_x and menace_o, and build move n from move n-1. Depth-first would also work, really, as we're talking about only 300-ish boards.