ok hear me out, let's say in a fictional future, AI was so good it could act like a person. I take some generic dude named Todd. Todd has no mental complications or disorders, he's just an average dude. You upload all of Todd's memories to an AI bot, but only the memories of his life, not his thoughts. You then have both Todd and the AI bot answer a series of questions, from something simple as cats or dogs or to something like what classifies as evil.
Will they make the same choices?
posted
I believe yes!
I've at least noticed that past events and memories do very much affect our conscious and unconscious decision making.
Say someone has many bad experiences from talking with new people.
Will they go and talk with strangers or will they fear because of those past experiences?
posted
Well, I don't really think AI ever could get to such a point, and most experts on the matter don't, from what I understand.
But taking the hypothetical at face-value. Im not really sure they would answer the same, no. Probably similarly, maybe even mostly the same, but I think there are a mix of pre-determined and developed characteristics. Scientifically this is mostly concluded, and philosophically, this is about what checks out for me.
There is also an important question, which is; is the AI now conscious? Which is again, mostly deemed unlikely and impossible by many experts currently, or is it simply following through a bunch of algorithms to act like the person?
If its conscious, then perhaps it would try to deviate from the singularity in order to assert its own Self-Hood. If its not conscious and is merely following algorithms, then imaginably it may predict with startling accuracy what Todd would respond, but I think it would mostly do this as a result of it having a bunch of data already on Todd and how he responds to things.
posted
Forum:
Report Topic
Permalink