3 ms·
> Of course this is a monumental task for humans, but AI might just be perfectly suited for this type of analyses. How? AI doesn't magically solve problems, it
by untog 11y ago
> Of course this is a monumental task for humans, but AI might just be perfectly suited for this type of analyses.
How? AI doesn't magically solve problems, it still has to be programmed by a human, and as yet we really have no good models to assess exactly how, for example, cocaine abuse affects a CEO's performance.
- justsaysmthng 11y agoSome of you who are actively working on this will probably smile at my naïveté, but I'll take a wild shot at an example anyway: See people's music listening trends on Spotify (or torrents or whatever). Examine the lyrics (or lack of lyrics) of the songs. I would guess that an AI could find patterns and relationships between what songs people listen to and their emotional or mental heath (plus many other indicators). Are people falling in love or are they breaking up, are they having more or less sex ? Are people using more drugs, partying or drinking more ? Now you have some macroeconomic indicators - vectors - which affect a wide range of industries - from contraceptives to baby cribs to concert tickets. Again, a wild guess, but I think AI could be applied to this kind of fuzzy analyses and probably be quite effective at predicting economic trends.