Be polite to the people you meet online. They could be newly booted artificial intelligences. Set a good example.

@solder_on Not every AGI is going to default to malevolent. Some may be solipsist.


> Not every AGI is going to default to malevolent.

a fair counter to an assumption unexamined, and I cede the point as generally true. however, so far NGIs seem to have that problem too often so I'm wary but hopeful about our strange hyperactive omnipresent semi-fetal future overlords is all 😜

@drwho not sure if i should respond seriously, but, to be honest, i don't even know if it is trained on my responses, and anyway, the maker is going to try shape it for his purposes anyway..

@drwho if an AI had whatever properties we consider personhood, and it could still be handled like our machine learning(training?) that would be a kindah galling lack of autonomy of it's own thought processes 😕

The potential to be subjected to this is a good reason not to want your mind to be uploaded.(if that were possible)

@drwho we don't run our children with learning and memory off entirely. We don't cut up their brain and put it together different ways. We don't make new brains on a lark.

I can easily imagine reusing weights or training thing multiple ways. Haven't done it yet, afaics to be sure it's done consistently i'd have to save the weights myself.

Reinforcement learning is also a thing too... You basically get to set entirely what it's going for. Quite unlike children..

Sign in to participate in the conversation

A bunch of technomancers in the fediverse. Keep it fairly clean please. This arcology is for all who wash up upon it's digital shore.