4 ms·
So if we can’t trust an AI to write routine code because it is unsafe, how shall we ever trust it to drive a car where physical security of human beings comes i
by goalieca 5y ago
So if we can’t trust an AI to write routine code because it is unsafe, how shall we ever trust it to drive a car where physical security of human beings comes into play?
- sk0g 5y agoAI is not a hive mind, and each will have varying levels of research, effort, and validation put into them. This is like asking, IMO, "how can humans be trusted to pilot planes, when my barista can't ride a bike?"
- NavinF 5y agoIt only has to be better than the average human. Self driving cars are already there, but GitHub copilot is just an alternative to googling for code snippets rn
- jimsimmons 5y agoSelf driving cars aren’t there. Which planet are you living in