Sort:  

The problems I see applying the Robot Rules is how do you really enforce them?

If you make something like an "unchangeable" part of its OS to be the Robot Rules and really consider every possibility how the robot may want to change this part you still can't be sure the robot won't find any way to bypass your restrictions not to change its source code. I mean... one AI would be able to do huge thinking operations that would take a human like millions of years in the fraction of a second. How do you believe to outsmart something like that beforehand?

And if you do restrict the AI so much that it really can't rewrite much of its own code you really don't have too much of an AI, do you? You could create a somewhat smart, helping robot but not something with an own consciousness and the ability to think on its own imho.

I guess you are right. It is kind of impossible as you said to outsmart them ... That's why I said that the humanity would never be the same as before...

Coin Marketplace

STEEM 0.19
TRX 0.18
JST 0.032
BTC 86998.67
ETH 3233.80
USDT 1.00
SBD 2.92