6 ms·

So they are right in assuming that people are searching more and more for answers to questions rather than looking up documents containing keywords. They are wr

by owobeid 13y ago

So they are right in assuming that people are searching more and more for answers to questions rather than looking up documents containing keywords. They are wrong in assuming that all of these questions inquire about what some physical object/location/etc is. But I think having something like this is useful in other ways. One application that comes to mind is that they could gather all submitted answers to build an automatic object/scene recognizer. Probably in a few years, Jelly could automatically tag objects in the photo you just took.

MasterScrat 13y ago
> build an automatic object/scene recognizer That's a really hard problem. Google has millions of labelled images and look how well Google Goggles is working...
- owobeid 13y ago
  True, but I would think that since photographs taken by Jelly would a) most like be close-ups and b) might also have some extra annotations (eg. circled objects, hint in the question). This would result in a training set that better suits real world input. I am not really sure to be honest, it's just a thought.
- ahel 13y ago
  >look how well Google Goggles is working Nicely for buildings and arts, but YMMV