YouTube

Got a YouTube account?

New: enable viewer-created translations and captions on your YouTube channel!

English subtitles

← P=NP - Intro to Algorithms

Get Embed Code
3 Languages

Showing Revision 3 created 05/25/2016 by Udacity Robot.

  1. I guess I think it's this one. None of these are exactly true and none of these are exactly false.
  2. This is kind of a fuzzy problem, but there are a lot of cryptographic protocols that are based on,
  3. on factoring like problems that are in NP.
  4. NP is a good source of secret keeping algorithm.
  5. For the same reason that NP problems are good for puzzles--that is to say,
  6. you might be able to look at an encrypted message and not be able to figure out where it came from,
  7. but if I gave you the secret information behind the scene, you should be able to read it.
  8. The people who have the key should be able to decode the message and other people not.
  9. There are definite cryptographic protocols based on factoring.
  10. If it shown the peak was NP, we actually have an algorithm that runs very quickly
  11. to solve all problems in NP, then this would definitely be true.
  12. A lot of CS theoreticians will be out of work. They will find something interesting to do.
  13. I'm not very worried about them. They're very smart people.
  14. They will presumably not work directly on this problem anymore, but there really only a
  15. handful people who spend a lot of their time directly on this.
  16. They're usually working on side problems that they hoped would build up to a solution to this problem.
  17. Whether this makes computer smarter than people or not, I really have no idea.
  18. My own research area is in artificial intelligence and machine learning,
  19. and there's definitely a lot of problems that we encounter
  20. that are NP type problems that we don't have the solutions for.
  21. If we could solve them quickly, that will be great, but I'll certainly able to believe that
  22. if the computers can solve it quickly, the human brain probably can solve it quickly too
  23. if we look at it the right way, and for now most of the problems that really seem to stump
  24. computers, these NP type problems, people seem to be able to do in practice rather well.
  25. If anything, it might help computers become similarly smart to people.