Comic Discussion > QUESTIONABLE CONTENT
WCDT Strips 3541 to 3545 (7th August to 11th August 2017)
gopher:
Liking Winslow's look, feels quite Fallout Boy to me.
Gyrre:
--- Quote from: Thrudd on 08 Aug 2017, 10:45 ---I would not go so far as saying that being able to walk about without issue is a higher level of learned skill set.
He already has the basics of locomotion and navigation from being in his ipod chassis.
Do you think about how you place your limbs and flex each muscle as you walk?
I like to hearken mack to Macross/Robotech with what was referred to as Reflex technology [muscle memory for the organic types] - that is that each subsystem had built-in dedicated control systems to look after basic repeatable operations with a high level of skill. A master control signal comes in saying "keep upright balance" and so you stand there as the individual subsystems interact, flexing and relaxing to compensate for environmental irregularities and personal shifts in weight due to upper body actions and yet maintain an upright stance. This also explains the old "ow - I just walked into a lamp post" scenario where the reflex system was doing its job just fine propelling you forward but the higher level functions did not do theirs, in this case collision avoidance.
--- End quote ---
Humans learn fine motor skills over the course of 3 to 5 years. Some humans take even longer to grasp finer motion control.
BenRG:
Of course, AIs have the advantage of having bodies with their own separate 'brains'. You could pre-load the chassis with basic motor control software (walking, grasping and the like) and all the AI algorithm has to do is trigger it. The AI can and will learn to do other things that it wants to do (Momo gave dancing or playing music as good examples) but it doesn't need to do it all.
The closest biological parallel I can think of is that the AI algorithm is only the Cerebrum; the brainstem is part of the chassis and that handles most if not all of the routine bodily operations.
--- Quote from: Kugai on 08 Aug 2017, 23:38 ---Pintsize is going to want one now
--- End quote ---
Calling it now: gender-ambiguous hermaphrodite aesthetic. Actually having the parts will radically change his outlook though and maybe he might be less aggressive... Or possibly just as aggressive but isn't prepared for the occasional positive response.
flondrix:
--- Quote from: BenRG on 09 Aug 2017, 01:47 ---The AI can and will learn to do other things that it wants to do (Momo gave dancing or playing music as good examples) but it doesn't need to do it all.
--- End quote ---
I don't recall that discussion. Any idea when it was? I would like to read it.
Case:
--- Quote from: flondrix on 08 Aug 2017, 10:26 ---
--- Quote from: ChipNoir on 08 Aug 2017, 09:24 ---So basically, I think an AI still has to have experiential education, and process it like a human does. They might be better able to organize than a human student, but they'd still have to put in the time and dedication that a human would. So I don't think Winslow can just jump into any field he wants.
--- End quote ---
This overlaps with something I was wondering about...after downloading into the new body, Winslow can stand and walk and safely hug people without staggering about for a while first. Shouldn't he have a learning curve?
--- End quote ---
We humans might be biased in that regard, since it takes us so very long to learn locomotion after birth, but I believe that we're rather the outliers here - in fact, a lot of mammal infants acquire basic locomotion within minutes (cute vid of baby elephant, cute vid of baby Vikuna), whereas the first thing human babies learn in terms of locomotion is reaching.
Apparently, being 'a bit of a slowpoke' wrt. motoric learning is common for many species of primates:
--- Quote ---There are obvious differences in adult locomotion for humans compared to the great apes, but there are additional differences related to the development of locomotion. Many mammals are able to locomote shortly after birth, and it is likely that most mammals are born with an innate locomotion circuit, but the time to reach mature, adult-like locomotion varies significantly.
Human infants display coordinated leg movements that seem to mimic walking movements at a very young age, but they do not reach typical, adult walking behavior until 6-7 years of age. Newborn infants supported stepping patterns do not match the plantigrade locomotion of adults.
[...]
Chimpanzee infants begin locomoting primarily with their upper limbs, with torso-orthograde suspensory locomotion as the most common mode. At about 3 years of age, they begin to transition to more quadrupedal locomotion. By 5 years of age, they begin to locomote completely independently. During this stage, they decrease clinging and torso-orthograde suspensory locomotion and increase quadrupedal locomotion. Locomotion finally reached the adult form by adolescence (10-13 years), when the majority of movement became quadrupedal walking.
[...]
Gorillas begin crawling on mother locomotion by about 3-4 months of age.
[...]
By 4 years of age, gorillas are locomoting much like an adult, with knuckle-walking quadrupedalism as the most common form.
--- End quote ---
True, human babies' bodies are simply not built for adult locomotion - think e.g. of that (relatively) huge head that the body cannot even support initially, or the fact that bipedal locomotion is mechanically less stable and therefore also harder to learn than quadrupedal locomotion. Primates in general appear to be built for "painfully slow learning", even those primates that primarily walk on four limbs. Maybe that's a function of the time that our bodies need to reach their adult form? Wouldn't make much sense to have hard-wired locomotion circuits for a body that cannot execute its commands for six to seven years?
Not that I have the first idea about neurology, but I vaguely recall that human natal learning is less developing connections than pruning them.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version