The Robot Birds And The Virtual Bees

Have you been looking through my character sheets again?

Thomas Boys

Niels, you're assuming a human child; the point was an AI 'child'. It would be tricky as hell to digitize DNA into some set of 'personality characteristics' with any sort of actual coherence, though. Might be better simply to use a random number generator and a D&D encounter table with the original settings overmarked and personality traits scribbled in, in pen, with added margin notes in pencil and colored markers.

All *male*

Andy Sleeper

Female Komodo dragons, if alone without any males, will lay fertile eggs. It's called parthenogenesis. Offspring will be all make, who can eventually mate with the mother.

Andy Sleeper

In 3071 I can't get enough of the little drunk rectangles over May's head! Lol

Andy Sleeper

That's a nice thought, but obviously the devil is in the details. Although development of the external womb has made a lot of progress, I'm not sure if they got the right mix of nutrients down you'd have to provide, nevermind the fact that although we've made huge leaps understanding DNA, it's still complicated enough to be unable to predict how to generate a fully viable genome.

Niels Roosen

Okay, but if you think about it that way, who is experiencing the joy? How is the experiencer of the joy different than the joy? And while it's true that a sensory impression gives rise to blue, and to red, are blue and red sensory experiences? How is experiencing joy, or blue, different than any other activity of the mind? When you see yourself doing something, is that different than the experience of joy? Is my "doing typing" any different than my "doing joy?" If so, how?

Abhayakara

But that's just it. Joy and other emotions are things that humans (and almost certainly other animals) *feel* , experiences that are inflicted upon them in a mostly passive manner. A thing you perceive. You can easily think of it as a sensory mechanism for parts of their own chemistry and brain activity. As for the subjective aspects and frames of reference, that comes down to exactly how the mind is architectured and how it responds to stimuli. Is red meaningfully distinguishable from blue, What if you can and do entirely forget the concept of "blue", etc

Now now, this is some important stuff here, namely that science (as it is in the post-singularity QC-verse) seems to support the idea of Faybles smooches actually being feasible. :P

Will Weaver

I don’t know, Evie, I’m pretty sure Pintsize would just repeat the process you described over and over again for all eternity if he could. Especially if it involved a tugboat.

Will Weaver

I was chewing this over a little bit and it occurs to me a female robot and a human of any gender could have children, in a way. If we assume that forming an AI takes two semi randomised strings from parent AIs, like how human children take two sets of DNA from each parent, then you could read a human's DNA sequence, convert it into a numerical code, match it to their AI partner's code and the result would be a child AI with traits from both parents. Just a thought.

Thomas Boys

The thing is, joy isn't a sense experience. It's entirely in the mind. You can feel joy while experiencing pain (e.g., a new mother meeting her baby for the first time feeling joy despite feeling really shitty from the birth experience). Also, _even if_ you get used to the feeling, is it the case that you would then benefit from feeling shitty just so that you could feel joy in contrast to feeling shitty? Or would you be better off feeling joy all the time, even if it became normal for you?

Abhayakara

But isn't adaptation a chemical process? Unless you programmed an AI to have a similar gated response to input, it could simply experience ecstasy at the same levels forever. (Now I'm thinking that very gating is a safety mechanism, so it might well be useful in the development of the AI sensorium.)

David Pipes

Whoa, cutlery on the wall?!? . . . Oh, hi Mark!

Willinwoods

In sensory biology there is a process called adaptation. Senses slow and/or stop being transmitted to the brain if the stimulus is constant. If the intensity of the stimulus changes from a base stimulus then the brain becomes aware of the change as a sensation. Senses that fade out faster are called phasic and ones that persist for longer are tonic. The sense of smell is a phasic one which is why we all get used to the funk of our own bodies and don't notice it (to the detriment of our fellow humans). I'd hypothesize that constant joy/ecstasy would also become less noticeable with time and that even more intense jolts would be required to achieve the former level of intensity. Then it would be too much effort/energy spent to make being joyful worthwhile.

Douglas E. Smith

Meanwhile Roko is still learning to master the skill: <a href="http://jephjacques.com/post/153991088855/roko-unwinds" rel="nofollow noopener" target="_blank">http://jephjacques.com/post/153991088855/roko-unwinds</a>

Andrew

I feel like we have seen parts of this before <a href="http://www.questionablecontent.net/view.php?comic=2133" rel="nofollow noopener" target="_blank">http://www.questionablecontent.net/view.php?comic=2133</a> <a href="http://www.questionablecontent.net/view.php?comic=3071" rel="nofollow noopener" target="_blank">http://www.questionablecontent.net/view.php?comic=3071</a>

Bagge

Right this minute? Hmm, hands above the table please Mandy.

Andrew

as long as he isn't robo jizzing everywhere.

Stoodmuffin

And exactly how much are you getting done now, Amanda?

Bagge

I'll admit that this is largely a personal theory confabulated out of half-remembered Hofstadter and Hawkins, based on a very...functionalist? Lense. Qualia as a product of certain types of complex sense-process-react agents.

ROBOGASMS! Winslow?

BobC

No, he just gets bored.

BobC

I haven't seen any research that actually shows anything particularly suggestive about this. People love to write about it because it's a fascinating topic, and being an iconoclast on this topic (e.g., the "the hard problem of consciousness isn't that hard" people) draws a lot of discussion. Have you seen anything concrete? Most of the interesting stuff I've seen has to do with introspection; the fmri stuff is fascinating and useful for modeling brain function, but doesn't seem to point in a definite direction as far as consciousness itself goes.

Abhayakara

So good.

Devin Forbes

I would believe that if we didn't have all the cake incidents as counterarguments.

Alex Mullenix

"A system watching and tasting itself"? This is Pintsize we're talking about, so, eeeewwwww...

Andrew Diseker

is that an A&amp;W? something about those burgers...

At least for biological intelligence, that experience appears to grown out of sensory perception in a physical body. Chemistry, vibration, patterns of light. You could think of reasoning and introspection and conscious memory as a kind of proprioception, a brain processing it's own states as another sense. If AIs are sentient in a recognizeable way, it's probably similar. A system watching and tasting itself

This is actually a much deeper question than it seems. Sure, ecstasy can get old, but what if joy is your baseline? Do you need to experience misery to appreciate the joy, or is it better to just experience joy all the time. For an AI to be meaningfully sentient, it has to be the case that there's a way that it's like to _be_ an AI. If there is such a way, then it's not just bool: ecstasy. Ecstasy is experiential; how experience arises for an AI is a very interesting question.

Abhayakara

This is a really fascinating take on how a human-like ai would approximate and experience "endocrine" phenomena. Awesome as always, Jeph!

I think I see it more as being classically trainable. An action resulting in an undesirable outcome will naturally be avoided.

Marlo Delfin Gonzales

Wait, so you're saying Pintsize has...self control?

Alex Mullenix

I wanna see Bubbles stop dragging this out!!!!! 😭