On 20/10/17 23:22, nospam wrote:
In article , Tim Watts
wrote:


what's even better is that with scene recognition and machine learning,
the need to manually keyword everything is going away.


At least in as much as "I keyword photo A" and the computer suggests all
similar photos in my collection and I chose which ones to replicate
keywords to. Ditto faces.

no. scene recognition can figure out what is in the scene for you
without needing to do anything.

for instance, you can query for paris in winter and it shows you just
that, without needing to have previously keyworded it. this capability
continues to improve and is better than humans in many cases.


Although I might prefer to be in control of the keywords chosen, and
merely have it recognise similar scenes.

the user is always in control.


Oh - there seemed to be a suggestion that AI would "just do it" - maybe
I misunderstood.