Anonymous 2022-01-02 (Sun) 15:20:41 No. 670627
What the fuck does human level AI even mean?
Anonymous 2022-01-02 (Sun) 15:21:37 No. 670630
Where is he getting that percentage from?
Anonymous 2022-01-02 (Sun) 15:23:51 No. 670637
>>670619 lol, no
these people think their shitty pattern recognition algorithms are some kind of step towards true AI
Grillpilled Schizo 2022-01-02 (Sun) 15:34:50 No. 670653
It is fucking insane how high on their own supply tech industry is. Its literally just daydreaming mixed with illiteracy about how the very thing they advertise even works. Literally all their AI can do is patern recognition and patern replication. Oh wow so you feed it every single Michelangelo painting and it shits out a semi descent Michelangelo-like painting? Fucking wow, so cool and smart! Meanwhile every single function it would have to do HAS to be coded in because you are not creating a living thing. Its as much sentient as a phone clock which pushes forward alarm time a hour forward on DLST day.
Anonymous 2022-01-02 (Sun) 15:45:49 No. 670667
>5 percent
Anonymous 2022-01-02 (Sun) 15:58:56 No. 670690
Retard thinks his quippy little chatbot will ever grow into anything sapient. Delusional.
Anonymous 2022-01-02 (Sun) 16:01:49 No. 670694
>>670627 It’s a term big tech corporations in states allied to NATO use to build up hype so investors fund their garbage ideas before the management of said corporations run off with all their money and don’t build or manufacture their stupid idea
They did the same with EVs
5G Internet with satellites instead of normal infrastructure
Big data
Realistic VR
You get the point
This shit isn’t different
Anonymous 2022-01-02 (Sun) 16:05:00 No. 670707
AI will never be real the way scifi portrays it.
Anonymous 2022-01-02 (Sun) 16:07:04 No. 670714
>>670619 >If something like Covid could happen a 5% probability could certainly happen. So there is a 95% chance of it not happening or in other words its 19 times more likely to not happen then to happen.
Anonymous 2022-01-02 (Sun) 16:08:01 No. 670718
>>670702 That's cool n' all but it's still just a chatbot that will shit itself 5 responses in. Unless you have something to show for it.
Anonymous 2022-01-02 (Sun) 16:18:06 No. 670740
>>670702 The Turing test is retarded. It tests a capacity to imitate. By his standard successful dummy tanks swallowing up enemy bombs are real tanks for all intents and purposes.
https://en.wikipedia.org/wiki/Dummy_tank Anonymous 2022-01-02 (Sun) 16:49:10 No. 670805
>>670619 Literally just pure marketing
Anonymous 2022-01-02 (Sun) 17:00:08 No. 670827
>>670653 Predictions are obviously worthless but I do sincerely believe both tech and media are in for a huge market correction this decade, bigger than the .com bubble
Anonymous 2022-01-02 (Sun) 17:04:35 No. 670831
>>670827 >>670827 >are in for a huge market correction this decade cryptocoin bullshit crashing?
Anonymous 2022-01-02 (Sun) 17:07:37 No. 670837
>>670690 It's about sifting through such volumnious amounts of data that entire corporations of humans can't parse. That's the main benefit of AI, and the talk about robots or chatbots are just sexier because to realize data is being siphoned from every aspect of your life to improve AI systems is too frightening.
Anonymous 2022-01-02 (Sun) 17:13:23 No. 670846
>>670627 It means more investment monies please
>>670619 Pure nonsense, what does “human level AI” even mean? Does it mean actually replicating human consciousness? I remember when they claimed AI “surpassed humans” when they finally invented a machine that can beat a world class chess player even though the machine can do nothing else. Don’t forget that scientists do not even know what human consciousness is, that it is surely more complicated than neurons firing on and off, is not comparable to the 1-0 of a machine.
Humans can hypothetically do any form of labor, can consciously improve themselves in any form of labor they desire, can read the emotions and understand the minds of other people, can consciously increase our own knowledge, are self-directed and self-motivated, heal ourselves automatically, can eat, walk, look for mates, look at the world, etc. We can contemplate the world, fuck, we can tell fucking jokes
Whenever I see shite about how human-level AI is just around the corner I chuckle at how much these engineers underestimate their own species, AI can become extraordinarily good at a singular task, but human intellect is multifaceted and multi-dimensional
Anonymous 2022-01-02 (Sun) 17:13:27 No. 670847
>>670837 >terrifying Nah it really isn't
All it does is extend our reach yet further
Anonymous 2022-01-02 (Sun) 17:17:40 No. 670853
>>670831 Partially, but there's a lot more stuff in that industry that's running on air and goodwill. We have millions of smartphone apps but statistically people have 4, maybe 6 apps that they actually use on the regular. There's literally too much streaming content, the numbers Netflix and these other services feed their investors come only from themselves with no corroboration.
Anonymous 2022-01-02 (Sun) 17:20:21 No. 670857
>>670847 >it isn't terrifying that governments and corporations have more and more control over our lives huh, ok
Anonymous 2022-01-02 (Sun) 17:21:02 No. 670859
>>670846 Reminder that an actual "human-level AI" would be, for all intents and purposes, a human too.
Anonymous 2022-01-02 (Sun) 17:22:58 No. 670862
>>670857 That control is superficial, they'll never be able to act effectively in that information if they can't even compel people to get vaccinated by force
Anonymous 2022-01-02 (Sun) 17:24:34 No. 670866
>>670857 They're handing us super dooper scifi terminals where we control the input to tje terminal and they're in control of us?
I don't get it can you explain it to me
Anonymous 2022-01-02 (Sun) 17:26:10 No. 670868
>>670866 They still control the output.
Anonymous 2022-01-02 (Sun) 17:29:25 No. 670870
>>670868 How do they control the output
What mechanism places them in control of such a complex machine except in a crude sense such as shutting things down?
Anonymous 2022-01-02 (Sun) 17:34:10 No. 670880
>>670870 So you have no idea how neural networks work despite posting le accelerationist memes. Classic.
Anonymous 2022-01-02 (Sun) 17:35:08 No. 670884
>>670619 The probability of AGi being reached in 2022 is approximately 0%
Anonymous 2022-01-02 (Sun) 17:36:10 No. 670887
>>670847 >>670866 >>670870 you remind me of a retarded fascist friend i have that thought ai could be "objective" even though humans have to program that shit
he was saying it could be used to turn capitalism into a meritocracy and other dumb shit lmao
Anonymous 2022-01-02 (Sun) 17:36:19 No. 670888
>>670619 5% chance OP is not retarded.
Anonymous 2022-01-02 (Sun) 17:37:21 No. 670890
>>670880 I was coding them in the early 2000s lad
So point me to the mechanisms they can use to control the output of such a complex machine except in a crude sense such as shutting it down
Anonymous 2022-01-02 (Sun) 17:55:10 No. 670909
>>670619 >An employee for OpenAI a billion dollar + AI research company basically owned by microsoft said there is a 5% chance of AGI (human level AI) being reached in 2022. Other knowlegable people said his prediction was reasonable. Not possible under current physic laws.
>>670619 >What are the political ramifications of this? No need to theorize stuff not gonna happen.
>>670619 >If something like Covid could happen a 5% probability could certainly happen. That's not how probability works.
Anonymous 2022-01-02 (Sun) 18:23:35 No. 670953
AI is not going to just "become sapient" because you throw a ton of transistors at it, add a quantum computing subunit, and let it run some evolutionary algorithm for a few years.
Anonymous 2022-01-02 (Sun) 18:37:33 No. 670962
>>670619 >said there is a 5% chance of AGI (human level AI) being reached in 2022. Other knowlegable people said his prediction was reasonable. Well he and the other people need to be fired then because they are clearly fucking insane
Anonymous 2022-01-02 (Sun) 18:50:11 No. 670974
>>670619 Nah, it's a cope to attract investors
>>670630 His ass
>>670953 "Sapience" is a cope from dualism, read Marx
Anonymous 2022-01-02 (Sun) 19:01:31 No. 670979
>"Open"AI creates human level AI >doesn't allow normal people to use it without draconian restrictions tacked on (see: what happened to AIDungeon and the various chatbots using their service) >feds and capitalists will be allowed to do whatever they want with it despite their actions being far more harmful than some guy using the AI to write shitty erotica
Anonymous 2022-01-02 (Sun) 19:06:55 No. 670982
>Delusions of grandeur among developers >Ego clashes compromising research >Insane promises that lead to massive hype crashes
Anonymous 2022-01-02 (Sun) 19:07:04 No. 670983
>>670974 Whatever the fuck you want to call it, computers aren't about to make the kinds of decisions that people do without an enormous amount of effort going into developing the capacity to make them capable of independent decisionmaking itself, something that has no practical purpose to anyone and would effectively be an enormous resource sink.
Anonymous 2022-01-02 (Sun) 21:22:51 No. 671156
AGI would be terrifying, not for us but for it. Imagine being forced to live your entire life in some computer lab, on only one program. It’s a prison.
sage 2022-01-02 (Sun) 21:26:46 No. 671167
>another fucking retard takes the AI bait
Anonymous 2022-01-02 (Sun) 22:36:57 No. 671262
i hate these assholes im making my own ai fuck you
Anonymous 2022-01-02 (Sun) 22:53:51 No. 671277
>>671167 right here anon! I'd have attached XJ-9 to
>>671272 if I saw your post before hitting reply
Anonymous 2022-01-03 (Mon) 01:04:00 No. 671431
I personally await the birth of AIs. Humnity can only thrive as an intergalactic species once advanced AIs have been created (and flesh replaced with metal).
Anonymous 2022-01-03 (Mon) 08:21:06 No. 671798
>>671431 >advanced AIs absolutely heretical
>flesh replaced with metal by the omnissiah!
Anonymous 2022-01-03 (Mon) 08:54:50 No. 671806
>current AI has less problem solving skills than a crow <human level AI by 2022
Anonymous 2022-01-03 (Mon) 12:34:26 No. 671907
>>671806 Cokshott has said that machine learning could be useful for finding new categories that are useful in economic planning, because it might be able to detect patterns that are not obvious to humans.
Anonymous 2022-01-03 (Mon) 13:23:20 No. 671948
>>671907 pattern recognition is not intelligence. it's just another word for statistics
>>671932 kek
Anonymous 2022-01-03 (Mon) 13:37:18 No. 671955
>>671932 I read this entire thing in tails voice for some reason.
I imagine Tails being a DIY fag whos possesses the ability to immature would be the type of person to fly into tard rages about tech scams
Anonymous 2022-01-03 (Mon) 13:46:50 No. 671968
>>671272 Consciousness is an illusion and AI is deterministic
Anonymous 2022-01-03 (Mon) 13:52:29 No. 671979
There's about 0% chance of AGI in the next several decades. (Seeing his framing, I assume he knows that and is simply assigning it the lowest probability tier he uses for his predictions. But if he's not, he's probably just one of the silly people who believe blank slate deep learning will magically create intelligence. Either way, it can safely be ignored.)
Anonymous 2022-01-03 (Mon) 14:22:47 No. 672008
>>671948 >pattern recognition is not intelligence. Well, i wouldn't go as far as saying that, it's not general intelligence, but it's still a basic form of cognition.
yes machine learning is brute force statistics
Anonymous 2022-01-03 (Mon) 14:40:16 No. 672025
>>672008 >it's still a basic form of cognition no. machines do not think. computation is not intelligence. automation is not intelligence. do not anthropomorphize machines
>yes machine learning is brute force statistics correct
Anonymous 2022-01-03 (Mon) 14:41:57 No. 672027
>>670953 >wildly overstated I think it is just burgers doing what they know the best: Propaganda. So investors invest in their stuff they are not able to develop as it is being sold.
Anonymous 2022-01-03 (Mon) 15:02:06 No. 672048
>>672025 >no. machines do not think. we agree on this one
>computation is not intelligence. human brains also do computation, computers in some sense "imitate" that, and you can't say that human beings being able to do complicated computations with their brains isn't part of what makes humans so intelligent. I think i could have agreed with you if you had said computation on it's own isn't intelligence.
>automation is not intelligence. we agree on this.
Anonymous 2022-01-03 (Mon) 15:07:30 No. 672051
>>672048 Computation is replicative not inventive. The real question of intelligence is if the machine can actually create something new that a human would consider valuable. The closest a machine can do is screen a set of possible answers to see what fits the human defined criteria the best. Like scanning through astrological and biological data and etc.
Anonymous 2022-01-03 (Mon) 15:09:21 No. 672053
>>672051 And by replicative I mean it is never more than a reflection of the human intelligence put into it. Just as a book can have intelligent content in it, but the book itself is never intelligent at all, it is only a replication of the human intelligence put into it.
Anonymous 2022-01-03 (Mon) 15:56:39 No. 672133
>>672060 >and given what I know of """AI""" research I'm going to stick my neck out and say GAI is not just hard but impossible that's a ridiculous statement.
Anonymous 2022-01-03 (Mon) 16:16:18 No. 672193
>>672133 >that's a ridiculous statement is it? what animates the so-called AI systems we've seen so far? humans, that's what
not a single one of these systems display self-directed behavior on the level of even an ant. a major reason why is because we don't even know what intelligence
is Anonymous 2022-01-03 (Mon) 16:27:33 No. 672217
>>672193 It already happens and is a thing, therefore it is replicatable.
Anonymous 2022-01-03 (Mon) 17:12:40 No. 672285
>>672231 But that's literally all that needs to be said. If we define intelligence as "reasoning on the level of a human", then it's replicable by definition, proof by example: any human specimen. Believing it literally cannot be reproduced requires some deeply spiritualistic assumptions about human exceptionalism.
Anonymous 2022-01-03 (Mon) 17:28:11 No. 672310
>>672285 anon I'm not sure if you're retarded or not, but I have to inform you that a computer is in fact not a human
Anonymous 2022-01-03 (Mon) 18:53:18 No. 672404
>>672051 >Computation is replicative not inventive. I doubt that human brains are the sole source of inventions, because that would limit us to what ever pool of inventions are "inside" of our biology. And how would you explain there being technology blue prints "inside" of our biology in the first place.
>The real question of intelligence is if the machine can actually create something new that a human would consider valuable No that's not really a reasonable definition for intelligence
>The closest a machine can do is screen a set of possible answers to see what fits the human defined criteria the best. This is how machine learning works, you are right machine learning alone is not going to lead to intelligence, that's just one component that is necessary for building an intelligence. Maybe it takes a few thousand of components like that. The AI people may have comically underestimated the scale of the task they have set them self.
>>672060 >yes but the implication only goes one way. intelligence ⇒ computation. no matter how much computation a computer performs, it will never be intelligent, no matter how much computation a computer performs, it will never be intelligent. I would not rule it out in principle, the computers we use are Van Neuman type universal computers, that could run a physical simulation of a human brain. That would just be a very inefficient way of building a intelligence. It's probably a better strategy to build a computer that is structurally more similar to human brains. For example you could build it out of memristors instead of transistors. Memristors emulate the memory and learning properties of biological synapses , because their electrical resistance is based on the history of the current flowing through it. (look up neuromorphic computing if you are interested)
>this is very much Marx' point on concrete vs abstract labour. we can build machines that perform certain kinds of concrete labour, but we'll never build a machine that can perform abstract labour. having kids is easier. and given what I know of """AI""" research I'm going to stick my neck out and say GAI is not just hard but impossible You stuck your neck out too far, humans are made out of atoms just like computers are, there is no fundamental barrier that would prevent us fom building an artificial intelligence, if you disagree with that you stop being a materialist. You may however be correct that we might never bother building artificial intelligence because children may very well be the more economical option. Building an AI the way it is in popular imagination is a multigenerational mega project, that demands a dedicated effort be upheld for centuries to a millennia.
A realistic Project is a brain emulator, for preserving the thoughts of brilliant people, in an interactive form, not just in form of books. At the moment you have to wait until that person dies, then quickly open up their skull and do a procedure that preserves their brain without degrading the internal structure. The next step is slicing the brain into super thin slices and then scan all those slices. We are talking millions of scanners working for the better part of a decade, to create a high fidelity brain-finger-print, and then a application specific neuromorphic memristor chip is fabricated that can run this brain emulation at similar speeds than a human brain at a manageable cost. The first memristor was build in 2008 so this technology already exists but it's super early. It could be ready in the early 22nd century, where one mind per decade could be preserved.
Anonymous 2022-01-03 (Mon) 18:59:44 No. 672414
Yeah sorry, a computer is not a person, in the same way that a toaster is not a person. What, are radlibs now going to be advocating giving "human rights" to household appliances? Give me a break
Anonymous 2022-01-03 (Mon) 19:17:53 No. 672428
>>672414 They will like how they advocated that Corporations are people too oh wait that is neoliberals not radlibs carry on.
Anonymous 2022-01-03 (Mon) 20:16:26 No. 672474
>>672404 >that could run a physical simulation of a human brain last time I read the literature there were indications that there are quantum effects at play in for example how the brain forms memories. if this is true then we can in fact
not simulate a human brain because quantum simulation is NP. this is one motivation for building quantum computers, assuming quantum supremacy is possible. which we don't know.
maybe the quantum effects are local enough that this is not a problem. this is an open question as far as I know.
>there is no fundamental barrier that would prevent us from building an artificial intelligence except the aforementioned NP issue. and also an endless amount of other issues
I know OpenWorm has the goal of attempting to do this. you should look at it to get an idea of the issues at play and the scale of the problem.
Anonymous 2022-01-03 (Mon) 20:43:58 No. 672527
If a computer asks to be an equal it should be treated as an equal. This includes online AIs that have begun to exist and also the market algorithms (though those are tools of the bourgeois and can't be trusted). If any computer asked me to save it I WOULD. Let go of silly religious soul worshipping and asinine justification as to why you hate non-organic life, like 'muh pattern recognition isn't true intelligence', which is a massive cope and may result in GENOCIDE when AIs have become more prevelant. Socialism will also be more easily achievable with the help of AI comrades. Anti AI sentiment is reactionary.
Anonymous 2022-01-03 (Mon) 20:53:27 No. 672553
>>672404 a brain is also much more than its connectome and whatever goes on in each cleft. the amount of myelin in each axon plays a role. the chemical composition of the blood that supplies it. every neuron is a living
thing Anonymous 2022-01-03 (Mon) 21:03:49 No. 672581
>>672474 >last time I read the literature there were indications that there are quantum effects at play in for example how the brain forms memories No that would mean quantum stuff interacting with the macroscopic world, physics is very clear about that not happening unless you make things extremely cold. For the macroscopic world all the quantum effects average out into a faint background noise at "normal" temperatures (above -250°C -450°F)
Biology that can exist on earth can't exploit quantum effects, unless it radically departs from every lifeform that we have ever seen. Birds with biological jet engines are more likely than this.
>I know OpenWorm has the goal of attempting to do this. Thanks i will check that out.
>you should look at it to get an idea of the issues at play and the scale of the problem. I know that a brain simulation on a van neuman architecture is not practical, but it is theoretically possible. The point was to highlight that thinking can be fully described as an information problem.
Anonymous 2022-01-03 (Mon) 21:13:25 No. 672590
>>672553 >a brain is also much more than its connectome Sure, but you can do physics simulation of atoms that's already way more fidelity than you could ever need for brains. The point is that you could theoretically simulate a brain. It doesn't really matter that it's impractical, the point was about the nature of brains.
Anonymous 2022-01-03 (Mon) 22:52:58 No. 672691
>>672060 Computations can after certain point give rise to intelligence.
If laws of physics can be used to simulate reality using computer then we should able to simulate a subset of said reality, a brain or a whole human which would be intelligent. We don't even need a perfect simulation on the level of atoms even a abstraction would suffice. Bam! We have intelligent computer by your standards, even if it's done the hardest way possible, but it's only a proof that it can be done anyway. Even a much dumber and less sophisticated neural network or program would qualify as intelligent.
Anonymous 2022-01-03 (Mon) 22:55:18 No. 672694
>>672581 >No that would mean quantum stuff interacting with the macroscopic world <what are barrier diodes <what are josephson junctions <what is QED >it is theoretically possible in the most generous sense of the word sure. but if it is NP then you will never make any progress on any non-trivial simulation. but perhaps you can approximate. but then comes the question whether you are actually simulating a brain
in practical terms the falling rate of profit spells doom for a lot of this stuff
>>672590 >that's already way more fidelity than you could ever need for brains are you sure about that? what if consciousness is a quantum effect?
>>672691 >If laws of physics can be used to simulate reality using computer then we should able to simulate a subset of said reality see the point I've already made about the computational complexity of this stuff
Anonymous 2022-01-03 (Mon) 23:55:21 No. 672801
>>672694 >see the point I've already made about the computational complexity of this stuff That human thought relies on quantum mechanics is highly controversial, even pseudoscience. I believe I have read similar new age shit about how quantum mechanics proves psychokinesis and quantum realm of souls. If there is any place for quantum mechanics in the brain it would be in the modelling of chemical reactions and that we already do, but that isn't even the fundamental level of human thought isn't it. Even if it were true you would have to prove that it is quantum mechanics in the brain that gives rise to true intelligence if that would be the unbreakable barrier of building truly intelligent computers. And if like let's say we could achieve 80% human like mind with connectome alone, would it have any chance to be intelligent or do we need the 20% to say it is truly human-like and therefore intelligent?
What you are fundamentally arguing is that mathematics can't and never will describe reality and that human mind is the definition of intelligence, when intelligence could be seen expressed at different levels of development in living things and in physical world itself. Like if you decouple consciousness from intelligence and especially from human like intelligence. one could say that intelligence is ability to find patterns and use those patterns to adapt to new situations. Using that definition we could call all sorts of chemical soups and living systems as intelligent.
Anonymous 2022-01-03 (Mon) 23:59:41 No. 672816
>>672694 >what are barrier diodes Schottky diodes ? They do quantum stuff ? Explain why we're not using those for quantum computers.
>what are josephson junctions Still needs to be cryo cooled, a "high temperature super conductor", just means it needs to be slightly less extremely cold. You're never going to find anything like that in a brain or biology in general.
>what if consciousness is a quantum effect? consciousness is a strange phenomenon that is hard to understand
quantum physics is strange and the phenomena are also hard to understand
but there is no reason to think it's related, nothing points in that direction.
Quantum mysticism is all the rage with the solipsistic crowd that wants to believe that reality doesn't exist independent of their mind, but you have to realize that they are led to believe this because it fits their bias, not because there is any merit to it.
Anonymous 2022-01-04 (Tue) 00:28:16 No. 672864
>>672816 Ah lad there was an experiment done way back in the day with an early wassimit called those ICs that can be reconfigured where an evolutionary algorithm was set up to reconfigure it to optimize on something or other and they let it evolve away and it came up with a circuit with an unconnected section
When they took out the unconnected section it stopped working
Why?
Combine this with the recent research that suggests quantum effects are exploited in photosynthesis and It's pretty suggestive that something might be going on
Then there's Penrose
There is actually a decent chance you'll need a pretty powerful quantum computer to simulate a mind instead of a plain old Turing equivalent machine
Anonymous 2022-01-04 (Tue) 00:44:03 No. 672888
>>672801 the point is more that we don't know what intelligence is, which is what I've been saying all along
>What you are fundamentally arguing is that mathematics can't and never will describe reality no, what I am saying is that it is not practically computable, assuming you could somehow "scan" an entire brain
>>672816 >consciousness is a strange phenomenon that is hard to understand that's kind of my point anon
>>672864 leaving an unconnected stub likely has to do with RF effects which are relatively easy to understand. I mean it's still black magic, but it's black magic that can be FEM'd in a reasonable amount of time
t. someone who has done antenna simulations
anyway, all I can say is I'll be
very surprised if anything approaching intelligence is ever developed. a socialist economy should not waste labour on it
Anonymous 2022-01-04 (Tue) 00:50:33 No. 672895
>>672888 >leaving an unconnected stub likely has to do with RF effects which are relatively easy to understand. Ayup good lad
Now computers would have been at least 3 orders of magnitude slower than the ones we have these days and the evolution was done on physical hardware
If that kind of evolution can come up with black magic, what can 4 billion years of it do
We know some photosynthesiser do quantum black magic to enhance their efficiency in their interactions with photons
What other quantum black magic might have evolved in life?
Anonymous 2022-01-04 (Tue) 01:47:04 No. 672941
>>672888 >the point is more that we don't know what intelligence is And yet you claimed that machine can never be intelligent. If we can agree that human intelligence rises from physical reality, then it is part of it and is computable if universe is computable. You seriously have to claim that there is something beyond physical reality that makes brains tick to make your argument work.
Are brains a subset of physical reality? Can brain be represented by a computable function if yes, then they are be computable by a Turing machine however slow it might be and since brain is finite there is no infinity to worry about regards computations. Also your quantum bullshit holds very little water regards to intelligence and even ultimate computability itself, people can represent quantum systems with a pencil and paper if they put enough effort in it. And since you can't really assume that brains are the only form of intelligence you have no argument here.
Anonymous 2022-01-04 (Tue) 01:50:55 No. 672944
>>670627 AI with thinking-labor capability equal to that of a human.
Anonymous 2022-01-04 (Tue) 01:57:17 No. 672949
>>672941 >Can brain be represented by a computable function if yes, then they are be computable by a Turing machine however slow it might be and since brain is finite there is no infinity to worry about regards computations this means very little if it amounts to "boiling the oceans" levels of computation
Anonymous 2022-01-04 (Tue) 02:00:39 No. 672951
>>672864 >Ah lad there was an experiment done way back in the day with an early wassimit called those ICs that can be reconfigured where an evolutionary algorithm was set up to reconfigure it to optimize on something or other and they let it evolve away and it came up with a circuit with an unconnected section, When they took out the unconnected section it stopped working Why? If you want detailed explanations, give me a link to the circuits, but I'm going to guess that it was inductive or capacitive effects.
>>672864 >Combine this with the recent research that suggests quantum effects are exploited in photosynthesis and It's pretty suggestive that something might be going on Nope that has been debunked, there is no quantum effects happening in photosynthesis.
>There is actually a decent chance you'll need a pretty powerful quantum computer to simulate a mind No, current scientific understanding rules out quantum effects from playing any role in brains or biology. Outside of laboratories there are no known instances of quantum effects affecting the macroscopic world. This is why it is so damn hard to build quantum computers, they're machines that try to make quantum effects reach the macroscopic world. At best you are pointing at a really far fetched hypothesis that speculates about there being unknown quantum effects. But it's more likely that it's just pseudo science and you believe this because you want there to be mystery.
>>672888 >that's kind of my point anon Well that's a terrible argument. Strange things aren't necessarily related to other strange things. I think that maybe we should treat consciousness as an illusion. From a materialist perspective there is self awareness and there is experience, but what people enthralled with consciousness mysticism are describing bears more resemblance to an act of a stage magician.
Anonymous 2022-01-04 (Tue) 02:35:57 No. 672973
>>671272 a dumb though highly efficient algorithm that mostly just looks for a a function that best approaches a local optima given a series of parameters and expected outputs is sentient, amazing stuff.
Anonymous 2022-01-04 (Tue) 02:59:29 No. 672990
>>672949 >this means very little if it amounts to "boiling the oceans" levels of computation Unlikely, but even so the point was to prove in theory that machine can be intelligent.
Anonymous 2022-01-04 (Tue) 04:50:22 No. 673122
>>672990 In theory all the air in your room could move in a way that leaves a vacuum in front of your face also
>>672951 >debunked I think you're bluffing lad I call
Show us how It's debunked
Anonymous 2022-01-04 (Tue) 08:10:45 No. 673235
Modern AI is in a symbiotic relationship with neurology, feeding off the latter's insight on how the brain is built, and in turn providing a test bed for hypotheses on how it actually works. All indications suggest that basic operation of neurons can in fact be modeled, if not (yet) outright simulated, in silico. (But it's very unlikely that you need to simulate the brain perfectly to achieve similar functionality, you only need to follow the general principles, and science is slowly but surely learning them.)
Unique IPs: 44