People think AI will model cognition and are really disappointed when math-done-really-fast simply gets job done. https://twitter.com/sriramk/status/707955202687688704 …
-
This Tweet is unavailable.
-
Replying to @patio11
This is the oldest story in AI: "How would you go about evaluating an individual's likelihood of being dishonest?"
2 replies 6 retweets 11 likes -
Replying to @patio11
Academics and amateurs alike: "We begin by imaging a machine which can define 'truth'..." Industry: "We use five gigs of transaction data."
2 replies 9 retweets 20 likes -
Replying to @patio11
"Wait but where is truth and beauty in the transaction data?" "Eff if I know, but I've got 42 three-line heuristics for features."
2 replies 0 retweets 9 likes -
Replying to @patio11
"Well you have to realize that's philosophically bankrupt." "Sorry, didn't hear you, too busy running simulated annealing."
2 replies 1 retweet 14 likes -
Replying to @patio11
"Are you even listening to yourself? That's, that's cheating." "We successfully identify 80% of fraudulent transactions; 2% false positives"
3 replies 1 retweet 8 likes -
Replying to @patio11
And if you think this imagined dialogue is a joke just read up on FICO.
1 reply 0 retweets 7 likes
@CiPHPerCoder Success rate is hard to measure across all the applications; suffice it to say that FICO is *scandalously* predictive.
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.