An AGI by definition is capable of self improvement. Given enough time (maybe not even that much time) it would be orders of magnitude smarter than us, just like we're orders of magnitude smarter than ants.
Like an ant farm, it might keep us as pets for a time but just like you no longer have the ant farm you did when you were a child, it will outgrow us.
(Warning this is a weird read, George Hotz shared it on his Twitter awhile back)
https://www.fimfiction.net/story/62074/friendship-is-optimal
Just because you can imagine something and define that something has magic powers doesn't mean that the magic powers can actually exist in real life. Are you capable of "self improvement"?
We’re talking about a potential intelligence with none of our hardware limitations or baggage.
Self-improve? My brother in Christ, have you heard of this little thing called stochastic gradient descent?
No, you're capable of learning things. You can't do brain surgery on yourself and add in some more neurons or fix Alzheimer's.
What you can do is have children, which aren't you. Similarly if an AI made another bigger AI, that might be a "child" and not "them".
> We’re talking about a potential intelligence with none of our hardware limitations or baggage.
In this case the reason it doesn't have any limitations is because it's imaginary. All real things have limitations.
> Self-improve? My brother in Christ, have you heard of this little thing called stochastic gradient descent?
Do you think that automatically makes models better?
> No, you're capable of learning things. You can't do brain surgery on yourself
What principle do you have for defining self-improvement the way that you do? Do you regard all software updates as "not real improvement"?
>All real things have limitations.
Uh, yep, that doesn't mean it will be as limited as us. To spell it out: yes, real things have limitations, but limitations vary between real things. There's no "imaginary flawless" versus "everything real has exactly the same amount of flawed-ness".
Software updates can't cause your computer to "exponentially self-improve" which is the AGI scenario. And giving the AI new software tools doesn't seem like an advantage because that's something humans could also use rather than an improvement to the AI "itself".
That leaves whatever the AGI equivalent of brain surgery or new bodies is, but then, how does it know the replacement is "improvement" or would even still be "them"?
Basically: https://twitter.com/softminus/status/1639464430093344769
> To spell it out: yes, real things have limitations, but limitations vary between real things.
I think we can assume AGI can have the same properties as currently existing real things (like humans, LLMs, or software programs), but I object to assuming it can have any arbitrary combination of those things' properties, and there aren't any real things with the property of "exponential self-improvement".
That's part of my reasoning. That's why we should make sure that we have built a non-hostile relationship with AI before that point.