Hacker News new | past | comments | ask | show | jobs | submit login

You don't have to design the machine. Once you have enough computer power, you can use genetic algorithms (or another method).

Con: It can be slow.

Pro: It worked at least once.




Ugh, no. You would just create a super-parasite. Evolution is parsimonious. You would need to know how to build a mind the simple way in order to direct evolution down just that one path that leads to minds - and meanwhile you'd have cruelly killed millions of by-then thinking feeling persons who didn't quite measure up. And who knows what other tendencies you'd have created aside from intelligence? The whole thing would be a nightmare.


That only works if you have a fitness function. No one does.


No one has a fitness function for a search interface. But Google does a lot of A/B tests and the interface evolve.

I think that they are not using random mutations in the code. They design the "mutations". But the important point is that they don't have a theoretical fitness function.


Of course they have a fitness function! What do you think an A/B test IS? It's a fitness function!




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: