![pcray1231](/data/avatars/m/1/1353.jpg?1649698015)
pcray1231
Well-known member
Md, that sounds like normal software. I.e humans put in equations. The computer merely does the calculations much faster than a human can. Even if those equations are differential equations rather than standard spreadsheet fare, it's not really AI.
The central definition of AI is that it perceives it's environment in some way, learns, and adjusts itself based on it.
That robot is not AI, for instance. Sure, it perceives it's environment and reacts to it. But it doesn't learn. For instance, the guy keeps sliding the box away with the hockey stick. It keeps going after the box. AI, after he does that once or twice, would decide it needs a new plan. And would attempt to prevent the guy from sliding it away. Could you imagine how surprised that dude would be if, on the second try, it grabs the stick and rips it out of his hands, breaks it, throws it aside, then proceeds to pick up the box? That would be AI.
It's scary because an equally valid "solution" to it's "human keeps sliding target away" problem is to kill the human. And machines don't have ethics.
The central definition of AI is that it perceives it's environment in some way, learns, and adjusts itself based on it.
That robot is not AI, for instance. Sure, it perceives it's environment and reacts to it. But it doesn't learn. For instance, the guy keeps sliding the box away with the hockey stick. It keeps going after the box. AI, after he does that once or twice, would decide it needs a new plan. And would attempt to prevent the guy from sliding it away. Could you imagine how surprised that dude would be if, on the second try, it grabs the stick and rips it out of his hands, breaks it, throws it aside, then proceeds to pick up the box? That would be AI.
It's scary because an equally valid "solution" to it's "human keeps sliding target away" problem is to kill the human. And machines don't have ethics.