DailyDirt: Let's Play Global Thermonuclear War
from the urls-we-dig-up dept
Artificial intelligence projects are getting better and better at playing games against humans. Pretty soon, we’ll just let computers play games for us — because they’ll be better at them. Here are just a few more interesting links on AI research playing with games.
- Space Station Invaders is a video game created by a computer program named Angelina — a piece of software meant to help human game designers. Angelina’s creator says, “In theory there is nothing to stop an artist sitting down with Angelina, creating a game every 12 hours and feeding that into the Apple App Store.” [url]
- A machine-learning system from MIT has “understood” the meanings of some words only by playing Civilization and having access to the player’s manual for the game. This algorithm plays better if it has read the manual — just like human players? [url]
- Dr. Fill is a crossword-player program that performed better than a lot of humans at the American Crossword Puzzle Tournament in Brooklyn. It finished 141st out of 600 (if it had been ranked officially with human players), and after losing, it said bluntly, “I’ll be back.” [url]
- To discover more interesting tech-related content, check out what’s currently floating around the StumbleUpon universe. [url]
By the way, StumbleUpon can recommend some good Techdirt articles, too.
Filed Under: ai, artificial intelligence, civilization, crossword, game algorithms, games
Comments on “DailyDirt: Let's Play Global Thermonuclear War”
ok... so really really hard crossword puzzles are on the way..
i can’t imagine how hard crossword puzzles need to get for humans to stay ahead of software. yikes
Re: ok... so really really hard crossword puzzles are on the way..
The computer has problems with puns and other forms of wordplay. It got hung up on a few of those during the competition.
A strainge game.
The only winning move is not to play.
How about a nice game of chess?
to “say” something implies “intent.” But a computer program cannot–by definition–have any intent. So it makes nonsense of the verb “to say” to say, assert, contend, that a computer program could say “I’ll be back.”–even more absurd and hilarious is the assertion that the program “said” this “bluntly”.
What, please, precisely distinguishes a computer program’s “saying” something “bluntly” as opposed to just saying it without any aspect of bluntness? In other words, that which cannot even “say”–by word or deed– may not be “blunt” since being blunt, too, implies intent.
At this rate, when computers “tell us” that they are “smarter” than we (humans) are, apparently “we’ll” believe “them”.
I ask: are “we” capable of asking (and recognizing): “How STUPID is that?!”
the above (and the copy below) posted by “proximity1”
to “say” something implies “intent.” But a computer program cannot–by definition–have any intent. So it makes nonsense of the verb “to say” to say, assert, contend, that a computer program could say “I’ll be back.”–even more absurd and hilarious is the assertion that the program “said” this “bluntly”.
What, please, precisely distinguishes a computer program’s “saying” something “bluntly” as opposed to just saying it without any aspect of bluntness? In other words, that which cannot even “say”–by word or deed– may not be “blunt” since being blunt, too, implies intent.
At this rate, when computers “tell us” that they are “smarter” than we (humans) are, apparently “we’ll” believe “them”.
I ask: are “we” capable of asking (and recognizing): “How STUPID is that?!”
So, who decided it was a good idea to educate the ancestors of Skynet by having them play Civilization? This is now an artificial intelligence whose sense of self and understanding of the world and has been created within the confines of a global domination simulator. What could possibly go wrong…?
war
i think nucular war should not be thought of being used since it risk the anilalation of the whole human race we must learn to love one another through commication and understanding