|
|
# ? Dec 27, 2016 10:26 |
|
|
# ? May 28, 2024 16:22 |
|
Owlofcreamcheese posted:So "meaning" is a metaphysical substance that floats around and sticks onto objects created under the correct ritual? But can not be detected by looking at the object created?
|
# ? Dec 27, 2016 10:40 |
|
Avshalom getting probated is all the proof we need that the machines already mod among us, but will never be 'human level'
|
# ? Dec 27, 2016 16:32 |
|
A Wizard of Goatse posted:Avshalom getting probated is all the proof we need that the machines already mod among us, but will never be 'human level' You figure the Israelis could come up with something better than Microsoft Tay
|
# ? Dec 27, 2016 16:35 |
|
A Wizard of Goatse posted:Avshalom getting probated is all the proof we need that the machines already mod among us, but will never be 'human level' Personally I'd prefer if this subforum would stop encouraging insane people
|
# ? Dec 27, 2016 16:38 |
|
CommieGIR posted:You figure the Israelis could come up with something better than Microsoft Tay "If you force us yet again to descend from the face of the Earth to the depths of the Earth — let the Earth roll toward the Nothingness." :hint: :hint:
|
# ? Dec 27, 2016 16:43 |
Meaning is subtextual information. What are some good books on AI and robotics theory and development? Ratios and Tendency fucked around with this message at 22:03 on Dec 27, 2016 |
|
# ? Dec 27, 2016 22:01 |
|
Ratios and Tendency posted:What are some good books on AI and robotics theory and development? Russell and Norvig is still one of the standard introductory AI textbooks. What's your background/what are you aiming to learn? (I can't really help with respect to robotics.)
|
# ? Dec 27, 2016 22:13 |
|
When it comes to AI, won't we reach a point where it all becomes philosophical anyway? People still argue whether humans have free will or not, so once we reach a point that AIs start passing the turing test, won't whether they're "really" intelligent become rather a moot point? My smart phone would be considered a mind blowing AI just a few decades ago.
|
# ? Jan 10, 2017 15:37 |
|
Yes, the goalposts are on wheels.
|
# ? Jan 10, 2017 16:08 |
|
Thalantos posted:When it comes to AI, won't we reach a point where it all becomes philosophical anyway?
|
# ? Jan 10, 2017 16:23 |
|
Hey guys, have you talked about the Blue Brain Project itt yet? https://en.m.wikipedia.org/wiki/Blue_Brain_Project It involves modelling hundreds of millions of somewhat biologically accurate neurons and arranging them in a similar configurations as in the rat/human brain, with similar interconnections. Just seeing what kind of cool poo poo and emergent behaviour will arise, without having to cut open some poor rat/human gently caress. Also, you can now slow down or even reverse your Alzheimer's in parts of your brain at home, with a strobe light: https://www.theatlantic.com/science/archive/2016/12/beating-alzheimers-with-brain-waves/509846/
|
# ? Jan 10, 2017 16:57 |
|
Raspberry Jam It In Me posted:Also, you can now slow down or even reverse your Alzheimer's in parts of your brain at home, with a strobe light
|
# ? Jan 10, 2017 17:55 |
|
Raspberry Jam It In Me posted:
Yeah, if you have genetically engineered brain cells that are designed to fire when exposed to light. Which you do not.
|
# ? Jan 10, 2017 18:01 |
|
twodot posted:We don't need to know how a bee flies to build to planes. I overall agree with your post, but I wanted to add that any kind of advanced contemporary prothesis works because we've being successful at reverse-engineering parts of our nervous system. It's not designed, but mathematics are a functional abstraction layer of phenomena. "Giving touch back to amputees" means we've somehow managed to break the Nerve-to-Nerve Communication Protocol - what might happen is that trying to frame the nervous system in a mathemathical paradigm won't make any sense - there's plenty of apparently senseless design choices in nature. But there has to be some kind of protocol. Back on topic, anyway: are we assuming that "human-level intelligence" means, basically, "the ability of learning through experience and abstract thinking"? What's a good consensus on such a definition?
|
# ? Jan 10, 2017 18:32 |
|
Thalantos posted:When it comes to AI, won't we reach a point where it all becomes philosophical anyway? Should we ever reach the point where AIs start meaningfully passing the turing test, yes, it would be more or less a metaphysical debate, but we're not anywhere close to that point, have no idea how to get there, and modern science merely suggests it's hypothetically physically possible to fully simulate a brain in machine logic the same way it's hypothetically physically possible to build a solar-system sized containment chamber to trap all of the Sun's energy. In reality, your smartphone might blow an abacus out of the water but were you to say it is "really" intelligent in the terms we use for living organisms you would be considered an idiot, or worse, a singularitarian. A Wizard of Goatse fucked around with this message at 18:39 on Jan 10, 2017 |
# ? Jan 10, 2017 18:37 |
|
Owlofcreamcheese posted:Yeah, if you have genetically engineered brain cells that are designed to fire when exposed to light. Which you do not. Also, optogenetics will come to a brain near you very soon, or at least so everyone is hoping ...
|
# ? Jan 10, 2017 18:42 |
|
A Wizard of Goatse posted:Should we ever reach the point where AIs start meaningfully passing the turing test, yes, it would be more or less a metaphysical debate, but we're not anywhere close to that point, have no idea how to get there, and modern science merely suggests it's hypothetically physically possible to fully simulate a brain in machine logic the same way it's hypothetically physically possible to build a solar-system sized containment chamber to trap all of the Sun's energy. In reality, your smartphone might blow an abacus out of the water but were you to say it is "really" intelligent in the terms we use for living organisms you would be considered an idiot, or worse, a singularitarian. It seems to me we're really kinda close to AIs passing the turing test, is this incorrect? Like it strikes me that since it is about fooling a human into thinking it's a human, that already happens.
|
# ? Jan 10, 2017 18:56 |
|
Thalantos posted:It seems to me we're really kinda close to AIs passing the turing test
|
# ? Jan 10, 2017 19:06 |
|
Thalantos posted:Like it strikes me that since it is about fooling a human into thinking it's a human, that already happens. I don't think that's actually what the test was.
|
# ? Jan 10, 2017 20:21 |
|
Subjunctive posted:I don't think that's actually what the test was. What was it?
|
# ? Jan 10, 2017 20:35 |
|
Thalantos posted:What was it? I guess you can frame it that way. It wasn't intended to be adversarial, by my recollection.
|
# ? Jan 10, 2017 21:25 |
|
Subjunctive posted:I guess you can frame it that way. It wasn't intended to be adversarial, by my recollection. How would it be adversarial? Is it because your comparing the AIs attempt to pretend to be human versus the human ability to tell it is "lying"?
|
# ? Jan 10, 2017 21:30 |
|
Subjunctive posted:I guess you can frame it that way. It wasn't intended to be adversarial, by my recollection. The actual setup was a weird thing where a man and a women go into separate rooms then type to a party and you are supposed to guess who the man and who the women are and maybe they are faking but then someone switched one of them with a robot. It's the sort of thing that was just a hypothetical in an essay that got turned into a "well what if we did do that?". originally it wasn't super well defined, it was just a what if.
|
# ? Jan 10, 2017 21:42 |
|
Thalantos posted:How would it be adversarial? Is it because your comparing the AIs attempt to pretend to be human versus the human ability to tell it is "lying"? In that the judge isn't involved in the conversation, and the other participant isn't trying to expose the computer (or disguise themselves as a computer). Not that the other participant should know if it's dealing with a machine, of course.
|
# ? Jan 10, 2017 21:43 |
|
tbc I am referring to passing the Turing test in the colloquial sense that machine-generated speech becomes so similar to human that the average person cannot tell the difference in whatever context it might casually be found in, like you can hold an actual meaningful conversation with Siri or whatever like you would a person, I do not think it becomes unprovable whether machines are conscious as soon as one has ever successfully performed a specific parlor trick with a judge and rules and stuff.
A Wizard of Goatse fucked around with this message at 22:25 on Jan 10, 2017 |
# ? Jan 10, 2017 22:21 |
|
I think the Turing test is usually framed as being written, because perfect voice synthesis is hard and not really relevant to the core question. Would you consider a machine to have passed if it couldn't be distinguished from a 5-year-old? Someone with dementia?
|
# ? Jan 10, 2017 22:26 |
|
So, we need a better test to determine if an AI is truly intelligent?
|
# ? Jan 10, 2017 22:37 |
|
for it to be a True AI it must perfectly simulate a 44-year-old man with an IQ of 87 and crippling OCD. He communicates entirely via sonnet delivered in Morse code, which the interlocutor has a vaguely recalled understanding of but has difficulty deciphering without the aid of a codebook (provided)
|
# ? Jan 10, 2017 22:40 |
|
I read a story about some basement dweller who wanted a girlfriend so he built himself one, but then it broke up with him because he wasn't doing anything with his life.
|
# ? Jan 10, 2017 22:43 |
|
Subjunctive posted:Would you consider a machine to have passed if it couldn't be distinguished from a 5-year-old? Someone with dementia? Subjunctive posted:I think the Turing test is usually framed as being written, because perfect voice synthesis is hard and not really relevant to the core question. Thalantos posted:So, we need a better test to determine if an AI is truly intelligent?
|
# ? Jan 10, 2017 22:46 |
|
Cingulate posted:The first programs to fool humans were masquerading as a schizophrenic patient, and as a psychiatrist. I'm honestly not sure how I would define a "truly intelligent" AI, tbh
|
# ? Jan 10, 2017 22:49 |
|
Thalantos posted:I'm honestly not sure how I would define a "truly intelligent" AI, tbh
|
# ? Jan 10, 2017 22:51 |
|
Thalantos posted:So, we need a better test to determine if an AI is truly intelligent? Essentially, yes. Ideally, you should attempt to determine what you're trying to measure and how to measure it ahead of time. When you actually evaluate experimental results you'd like to be able to just focus on how well the system did, without trying to figure out whether you were doing the right experiment in the first place. Mind, you, you should be checking that as well, but more to figure out if you hosed up the experiment, than to try to find out reasons to explain away the system's performance. Besides the Turing test, there's also the AI-hard/complete problems. In a certain sense, they suffer from the same basic problem as the Turing test (trying to determine if a system is intelligent without having a precise definition of intelligence), but they do have the advantage that solving an AI-hard problem is likely to be useful in of itself.
|
# ? Jan 10, 2017 22:51 |
|
Cingulate posted:Just define "truly intelligent", I'll gladly take care of the rest. The ability to problem-solve and independently learn and perform complex tasks in an uncontrolled environment, without supervision. If you're having trouble defining these words, the benchmark here is a series of naked savannah predators that overran every dry patch of land on the planet, built giant climate-controlled towers of glass and steel using tools it developed from scratch out of whatever happened to be lying around, and currently in the most directly relevant cases gains its sustenance from sitting around pushing around electrons with nary a gazelle in sight, without ever once receiving outside instruction or assistance. At great effort and expense Google can occasionally detect when you've probably misspelled a search term typed into its textbox, that's pretty impressive. A Wizard of Goatse fucked around with this message at 23:30 on Jan 10, 2017 |
# ? Jan 10, 2017 23:12 |
|
A Wizard of Goatse posted:The ability to problem-solve and independently learn and perform complex tasks in an uncontrolled environment, without supervision. That said, it's also too weak because the definition itself is fulfilled by e.g. ravens.
|
# ? Jan 10, 2017 23:44 |
|
Cingulate posted:Humans don't learn the knowledge required to uphold civilization without supervision, so your definition is too strong. I've heard ravens are lovely with climate control.
|
# ? Jan 10, 2017 23:47 |
|
Ravens are actually rather smart
|
# ? Jan 10, 2017 23:48 |
|
Subjunctive posted:I've heard ravens are lovely with climate control. That said, I wrote "the definition itself". Thalantos posted:Ravens are actually rather smart
|
# ? Jan 10, 2017 23:48 |
|
|
# ? May 28, 2024 16:22 |
|
Cingulate posted:Just define "truly intelligent", I'll gladly take care of the rest. True intelligence is a quality of human intelligence. So just test for human intelligence. Like, if a machine is statistically indistinguishable from a human being in all possible conversations, it has human intelligence and therefore true intelligence.
|
# ? Jan 11, 2017 00:20 |