Comic Discussion > QUESTIONABLE CONTENT
So how do AIs learn
JimC:
Forgive me if you've been through this, but I'm curious about how the Ai personalities develop. Clearly there are some pretty damaged personalities there. Excluding Bubbles which feels like PTSD what happened to the others. If Momo is 4 years old and experiences life at similar rate to humans then they must surely get created with an adult personality, so no child abuse (in broadest sense) to stuff up personal development. Presumably too all are created sane, or is there a major QA problem with AIs, but it's considered immoral to abort/euthanase damaged personalities?
Is it cold in here?:
Welcome, new person with interesting questions!
We don't know. Jeph has been vague about how AIs acquire knowledge and skills after they emerge from the creche.
Momo went through a phase where she acted childlike in some ways, and Pintsize is only now growing up, so there's evidence that they have to mature for a while.
We know they can have catastrophically poor impulse control, from May's example. It's a deficit in a necessary capability.
jwhouk:
You think the first AI's ended up like Tay from Microsoft?
Nepiophage:
It might be something like this http://www.damninteresting.com/on-the-origin-of-circuits/
JimC:
--- Quote from: Nepiophage on 31 Mar 2016, 01:02 ---It might be something like this http://www.damninteresting.com/on-the-origin-of-circuit's/
--- End quote ---
Seems feasible. The philosophical/moral/ethical implications are enormously complex. I suppose the number of people who'd be interested in it as a thread in the webcomic series is distinctly limited. I'll applaud if Mr Jacques wants to explore it though at some time in the future.
When homo saps are created we do minimal quality control, and typically take the ethical position that at some point of development (actual point varying according to religious/moral/ethical position) the new being attains human rights, and once it attains human rights society has an ethical duty to support it should it be unable to support itself.
I suppose the same thing could be thought true of QC universe AIs: there's a minimum level of consciousness which grants the AI equivalent rights. If the AI brain is self generating then Maybe it has a point at which it will stop major development, and that's the AI equivalent of adulthood...
Then we get on to the other ethical minefield of who has the right to create new AIS, and what ethical and legal responsibilities they assume with the act of creation, and so on... As well as Robo-Jail do we have Robo-home-for-the-Educationally-Subnormal? And are the educationally subnormal or otherwise inadequately developed supplied with chassis, and if so how are they maintained (shades of May's problems in other threads). Or is euthanasia practised on AIs that fail to reach a state of development that enables them to function in society, and if so how does this relate to their civil rights? Does Pintsize apparently sitting on a shelf until matched up with Martin give us some clues here?
Navigation
[0] Message Index
[#] Next page
Go to full version