Will AI brains need a teacher or a guide to living among humans or to even exist successfully in the universe? Can they figure out how to behave without a long period of learning like humans enjoy. I believe that they might actually require a longer period of learning. Simply because they have the capacity to learn quickly, does not mean they will use it right away. It takes humans longer than any other animal to fully mature. If you add walking and talking and socializing into the fold, it could be a while before a young AI could be let out in public.
In my short stories about androids who have been given awareness upgrades, I have them learning about their new condition through a teacher. Sometimes the teachers are humans other times they are other AI infused androids. What I don’t show is the long, tedious efforts to bring the AI up to speed on basic things like walking and interacting with humans on a basic level. That would be boring. Short stories are not a format for boring. So my stories move at a fast clip and usually expound upon a single idea.
I’m also interested in human and android relationships. If you had an android, would you be its friend or its owner? They say a dog is man’s best friend but what they mean is that the loyalty really only goes one way. Except for Payton in Silverado. This is not to say that humans don’t love their pets, but that is different than loving your best friend. Or is it? Would androids and humans ever be buddies? Now that would be an interesting character study. Note to self: possible story about human/android buddies.
One thought on “Plastic Pals”
I can accept that when the AI starts out, it will take a long time for it to develop the same as a human would. However, unlike humans, the AI has the ability to be transferred from one body to another, intact. (No M$ vs *Nix vs Mac vs…jokes, please.)
What I can foresee is that the first several AI androids will have to develop like a human, and learn whatever the heck common sense is. Emotions? I believe at least a rudimentary understanding of what emotions are will be a minimum requirement, just how much they need to evolve emotions I’m not convinced of yet.
But now, the time has come and the AI android has been interfacing with humans for a few years. Other androids can now be built and the AI copied into the other android bodies. There will, obviously, be some changes such as the need to re-adjust and fine tune to operate limbs or wheels or whatever with different lengths and frictions than the original, adjust to new/better/different eyes, and so on. However, the actual AI that enables interaction with humans and the world around it should be ready to go, cutting down the time needed to develop and be released “into the wild”. Even that can be reduced over time since as the AI is transferred, it eventually has memories of a multitude of bodies and needs less and less time to tweak to any new, similar bodies. The limitation then becomes storage space.
Indeed, with the AI and androids, I see a potential issue that I’ll call the Methuselah factor: The original AI could well be continually evolving, not just for a human equivalent lifetime, but potentially until the end of time. Destroy one, deliberately or accidentally? There’s plenty others. With the resulting huge amount of knowledge will come the need to prune, but such pruning could also be selective: certain strains of androids would continue developing selective areas, becoming specialized while still maintaining the basic AI that allowed the growth in the first place by the first successful AI.
So, the way I see it, once a successful AI android such as we’re talking about is developed, it will take increasingly less time for each successive generation of that AI-based android to insinuate itself into the real world.