Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Not sure that a business based on taking a photo of something and asking people "what is this?" is destined for success. Isn't that something that Google Now/Siri will probably automatically answer in the near future?

These services require data-entry on the part of their users in order to seed these identification engines. I see this data-entry as a hidden value of the user, and part of why network effects are so valued by VCs/the market.



Doesn't Google have a sort of "index" and "search engine" of billions of photos with captions, link headings, title text, alternate text, and so on?

https://www.google.com/search?tbm=isch&q=lug+nut

You're telling me it only works one way, from text to images, and doesn't work the other way? But Google already lets you upload an image to search...

http://www.google.com/imghp

In any case, Google can simply say "I think it's a lug nut. Do you agree Y/N?" And in effect, play 20 questions with its entire userbase. In very short order it would be capable of identifying anything.


The lug nut question is an important verification step and indeed data-entry on the part of the user.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: