Would using AI for experimentation be considered ethical /sci/?

>>7719921
we're doing it
it's called 'software testing'
look it up b i a t c h
rekt

>>7719921
What kind of experimentation? The sort that would be unethical if performed on people?

If it was demonstrably sapient, then no.

>>7719933
>The sort that would be unethical if performed on people?

I'm thinking social sciences. Thousands of AIs in a double blind randomized controlled experiment. On one side, you have a control society, and on the other side you have an experimental society, where you change some policy like taxes or a regulation to look at its effects. Time would be accelerated and we will be able to see its effects in 5, 10, or 20 years.

Then they would be reset.

>>7719921
>Would using AI for experimentation be considered ethical /sci/?

Yes.

>>7719921
Yes no maybe, with ethics you essentially try to dumb down a pretty complex situation into a yes no question (would you harm X if Y). You then proceed to self-insert yourself into the portrayed victim position to answer this question. By this point you realize that this actually depends highly on how much you value yourself and on how much you value your environment. Combine that with different generalised approaches drafted by different schools for morals and philosophy and you end up with the certainty that you will always do it wrong, regardless of your choice. This is ofc too highly generalised, but this is what you get if you try to determine which person(s)/entities you're supposed to kill (first).

>>7719941
>is it ethical to kill or mindwipe people after doing experiments on them

>>7720077
AIs aren't people though.

People can't be endlessly copied.

>>7720102

>People can't be endlessly copied

Which is probably why our genes are sequenced in such a way that we can only be cloned so many times.....

>>7720043
You seem to not really have an idea about philosophy,ethics or morals so your generalisation is pretty wrong.

The different approaches and school of thoughts don't have to be combined.

"Self inserting into victim position" where does this happen?

And the dumbing part is also never true, ethics purpose is to try to answer such question to a certain degree, and this answer can then be used to make a decision on a certain subject, such as in law for example.
And there everyone(most people) will agree that there is a difference between for example killing someone out of fun and killing someone out of self-defense.

Also ethics and morals are not interchangeable terms.
Anyway to OP:
As stupid as it sounds I can imagine it beeing looked as totally fine at first, but as some point either people will get too attached to Ais, or they will play too much of a big role in our society, so that it will be started to be looked unethical upon.

But I can imagine that it wouldn't be too hard to still experiment with them, unless we will live in somekind of 1984esque world with no privacy.

Depends on if the AI is conscious.

>>7720149
>cloned so many times.....
>cloned
u wot m8?

Some of this is explored in the book Altered Carbon including simulation and termination of conscious intelligences. And there is little ethics going on there - instead it is all soaked in an epic blood bath, so be warned before you buy the books (now a series).

>>7719921
>plays Fallout 4 once
An AI will never actually feel so it will never be unethical to experiment with it you underage faggot

>>7720102
>_____ aren't people.

Not everyone was had the privilege to be born from human tissue, asshole.

>>7719921
>>7720043

this brings up an interesting question

if such simulations are possible but unethical, won't enacting policies on a live population be the same or worse?