|
|
This is basically how image search works. It breaks everything down into lines, edge tracing, shape tracing, etc, and tries to fit them into what it's been trained on. Like Waton's image recognition has seen thousands of images, can look at pic of a dude playing soccer and guess its a dude playing soccer. It's an insanely difficult computing problem and we're getting closer all the time, but yeah, this is the meticulous action of something studying clouds with a perfect memory, and matching every little detail to something in their memory. It's quite impressive. Our big data processing technologies are getting better all the time, making more and more kinds of data accessible in new ways. (Like Watson's expert system for medicine: There's so much new (and old!) medical research information out there right now that no doctor can possibly keep up with everything they might need. But Watson can read every journal, and every instance can learn from everything every other instance does, correlating and matching the relevant details and weighing them up or down based on what it experiences. Every year it will get better and better. Great video about the future of automation: https://www.youtube.com/watch?v=7Pq-S557XQU |
#
¿
Jun 22, 2015 07:53
|
|
|
|
| # ¿ May 9, 2024 09:18 |
|
|
|
In the far future when only computers write Wikipedia Dogs The thing most things that aren't eyeballs are. Most human's arm's end in dogs. |
|
#
¿
Jul 7, 2015 01:10
|
|
|
|
dogcrash truther posted:uh? actually? it was the worst day in american history? rear end in a top hat? when those giant dogs burst out of the WTC, no one really knew what was going on, but everyone remembers where they were that day. |
|
#
¿
Jul 7, 2015 01:34
|
|
|
|
Hick Magnet posted:holy shiiit wow |
|
#
¿
Jul 8, 2015 04:05
|
|
