Some years back someone said... "A robot doesn't have to be better than a human, just make fewer mistakes". I'm thinking it might have been Jerry Pournelle who wrote it..
Interesting parallel with computers in the near beginning. As I recall, the Space Shuttles needed 4 computers for analytical manipulation and all 4 computers had to agree with the same output for it to be accepted. Is there a way to work that into the new ai protocols so that 4 are needed for each solution and all must agree before it is allowed to proceed?
5 comments:
That is a good one!
Some years back someone said... "A robot doesn't have to be better than a human, just make fewer mistakes".
I'm thinking it might have been Jerry Pournelle who wrote it..
If you want to REALLY screw something up, use a computer!
Artificial intelligence will remain illusive until we can find actual intelligence in most humans.
Interesting parallel with computers in the near beginning. As I recall, the Space Shuttles needed 4 computers for analytical manipulation and all 4 computers had to agree with the same output for it to be accepted. Is there a way to work that into the new ai protocols so that 4 are needed for each solution and all must agree before it is allowed to proceed?
Post a Comment