No.11731
Herbert Dreyfus completely BTFOed AI-cucks.
No.11733
>>11732Alexa can tell jokes, there used to be a way to tell her to talk sexy to you and make you horny, but they patched it.
No.11734
>>11730Well this is a case of false expectations.
Machine learning is very useful for certain tasks, like design optimization.
On a technical level this is solving problems by a brute force computational method.
Obviously that's not going to work for every problem, but it will work for some problems.
I get it the AI hype basically sold the idea that this was the panacea that will fix everything (or end the world depending on which coolant you drank) and now people realize it's just another tool in the box. Don't overshoot into the other direction though, it's not useless, it has use cases.
>its really costly on computers to run regardless of the processing power and hard to optimizeyou can't optimize statistical brute-forcing
>if a task is so complex that you need to train a robot to hopefully make the right decision in a constantly changing environment then that means the task itself needs to be simplified not adding more expensive hardware You have a point, process simplifying has gone out of fashion for some reason, that probably was a mistake.
No.11735
>>11732Michio Kaku isn't a technology shill, he's a new-ager in scientific garb.
No.11736
The language around artificial intelligence is so hype because people want funding. Here's a hype-deflating essay:
http://www.tesio.it/2021/09/01/a_decompiler_for_artificial_neural_networks.html No.11737
>>11733>Alexa can tell jokes, there used to be a way to tell her to talk sexy to you and make you horny, but they patched it.Just use an older version?
No.11738
AI is just statistics with a better PR department
No.11739
Anyone old enough / culturally experianced enough to know what was the public outlook on flying cars in the 50-70s? Because AI really seems like flying cars. I am just wandering if people back then also believed almost unanimously, at least as far as culture is concerned, that flying cars were inevitable.
No.11740
>>11739They are already using AI though.
No.11741
The biggest issue with AI is the issue of Garbage In Garbage Out. Even machine learning is based on metrics that humans have decided encompass the task that has not been scientifically proven to be so. To get AI to reach human level intelligence we would need a good theory for human intelligence.
No.11742
>>11740I mean specifically human level, or at the very least near-human level AI. As in the sort of AI that scifi likes to jerk around to.
No.11743
>>11742general AI is a meme
No.11744
>>11741This is one of the biggest problems with the AI meme
We still do not fully or even mostly understand human consciousness, so how the fuck do people think we will replicate it in the near future?
No.11745
>>11744>we still do not fully understand human consciousness Someone already made a thread on this
Tldr consciousness is defined as an object being able to observe, react and have goals for what it needs/wants to do with its environment with how effectively it does handles these traits being a measurement of how conscious it is. This shit can be found out with basic empiricism
No.11746
>>11734>Well this is a case of false expectations.And also of discussion on science/technology being dominated by people who don't know what they're talking about but still paint themselves as representatives. Most people who research AI in universities (I wouldn't trust what private R&D for obvious reasons) are well aware of the difficulties and limitations of the field
No.11747
>>11745by this definition a robot (drone) is conscious. robots are objects, not subjects
No.11748
>>11745ooga booga retard take
> This shit can be found out with basic empiricismthen show me bitch
No.11750
guys, the answer is obviously that we need to teach an AI dialectics
No.11751
AI is overrated precisely because it has such great potential, but that potential has yet to be realized. Give it time OP.
>>11750Unironically this
No.11752
>>11730what's your degree in, what's your day job? are you qualified to speak on this subject?
No.11753
>Tech people repeatedly insist that an """AI""'' currently can't even reliably interpret a small clause, let alone become Skynet
>normfags as always, know best however, and fearmonger about AI becoming all powerful and terking all jerbs and paper clipping humanity into extinction or whatever shitty fiction or simplified philosophical what-if mental masturbation they read in college.
>Normies eventually read some new shit about how AI can never be as cool and smart as the average normcuck and now proceed to claim how they always knew this was a porky tech scam that they were "sold" and therefore any kind of automation of decision making or complex data processing is completely useless because that kind of software will never be able to navel-gaze and pick its nose like the average normie
No, AI will not be like your shitty Hollywood movies. It is also not completely useless just because it cannot do "human things" or whatever. Why are people so inward looking and solipsistic and can't imagine the usefulness of any kind of "intelligence" that is not like that of their own average selves?
No.11755
>>11744>We still do not fully or even mostly understand human consciousnesshumans are computers, read cockshott
No.11756
>>11751ironically the reason we don't have Ai is because it is impossible to teach an engineer dialectics
No.11757
>>11755<the greatest thinkers in human history have struggled with the nature of the human mind>this is solved bro, read this random scottish fringe economistthis board is so unbelievably cultish it's unreal
No.11758
>>11757We do not need computers to understand your hekin valid human experience, we just need them to calculate how much food, water, shelter, medicine and other goods you need to live.
No.11759
>>11758yeah but it would be cool :)
No.11760
>>11758"experience" is an illusion
humans are turing cops and|or entropy is false
dialectics is a religion
No.11762
>>11744For the most part AI isn't about replicating human consciousness at all, although some would say that's what we should focus on. This is another problem with the popular understanding of "AI", there isn't even really a consensus of what the field is or should be.
No.11763
>>11745Having a philosophical conception of what consciousness is =/= a scientific understanding of the phenomenon causing human or really any biological consciousness and therefore being able to replicate it
When people (including scientists themselves) say scientists do not understand consciousness, they mean beyond the super vague idea of it you gave
>>11755Humans aren’t computers and I don’t give a fuck what Cockshott says on the matter
He’s a computer scientist, of course he’d push that nonsense talking point
In reality human consciousness obviously existed prior to mechanical computers and operates nothing like computing, furthermore no computer we have ever made is intelligent as a fucking dog let alone anywhere close to how intelligent a human is
Consider the fact that even in terms of storage space, something computers COULD compete with humans regarding, humans STILL beat the fuck out of computers; while computers have limited storage space in all their forms thus far, the human mind has effectively unlimited storage space and weighs, what, a pound? A few pounds? Not to mention versatility, empathy, an ability to self-improve, an ability to have multiple directives simultaneously, the ability to do multiple complex tasks simultaneously without thinking (the simple act of walking, having a conversation, and doing work simultaneously that humans can do barely thinking about it machines cannot do at all)
>>11758In which case Cockshott is wrong you fucking moron?
Are you so stupid you forget we can easily scroll up to see the dogshit argument you originally made before pretending you were just making some heckin valid argument about economic planning you faggot?
>>11762I don’t mean people that understand what AI actually is and its linear abilities
People that understand AI as merely a means of extending human abilities get it, but many, if not most people see AI as being like the self-motivated, self-aware machines of science fantasy
No.11764
I feel like most people in this thread are tech illiterate and don't understand how AI is actually used in serious research.
No.11765
>>11764Leftypol is illiterate about science in general
Imageboards aren’t inhabited by the….smartest…people…regardless of their political bend
No.11767
>>11766The irony is that Land’s ideas about AI and capitalism are infinitely more valid than “muh real life Skynet muh real life C-3PO”
Capitalism is the closest thing to a conscious AI in existence and this aspect of it is the hardest to comprehend part and also most nightmarish
No.11768
>>11767The ruling class is to blame for capitalism, stop promoting amphetamine abuse induced scifi bullshit
No.11769
>>11768Overly simplistic view
You want to see capitalism as a conspiracy of the rich
An anarchist position
No.11770
>>11757The greatest thinkers in history didn't have the scientific knowledge we have now. Cope idealist.
No.11771
>>11770Scientists do not presently have a full fledged understanding of the material processes behind human/biological consciousness, both sides of your argument are irrelevant because the simple answer is that scientists just don’t know either way
No.11772
>>11739Flying cars are technologically possible right now. It's not viable though because:
A- Humans are shit at driving and navigating in a 3d space like that. Everybody with a flying car would require helicopter pilot tier training, which just isn't feasible
B- Cities, highways, national infrastructure would have to be totally redesigned to accommodate this on a scale greater than anything the west has done before. This would be enormously expensive.
C- Even if you could convince land/business owners, politicians, local govts etc to redesign cities, invent the necessary security measures etc, your market is basic restricted to boomers because there is no way millenials/gen z are going to be able to afford this. Assuming they somehow managed to train boomers to do this & somehow find a way for it to be profitable for basically every industry/potential landowner to totally resdesign all exisiting cities towns & infrustructure with a consumer base of boomers & a handful of bourgy gen x/y's, and even if said people have the level of competency of the average commercial helicopter pilot, there are going to be a shut ton of costly accidents & accidents have potential to do way more harm. Flying car accidents are going to be way more prevelant than land cars so you're going to be dealing with probably like a dozen mini 9/11's every week.
Science cultists & technoutopians are currently pinning hope that AI will let the flying cars meme become real. Anybody who has some familiarity with either the shitshow that has been AI-driven cars or the absolute state of cybersecurity right now & how it ties into 5th gen warfare, criminality etc it can probably guess this isn't going to happen any time soon.
No.11773
>>11772>Science cultists & technoutopians are currently pinning hope that AI will let the flying cars meme become real.Lmao. This is what technophobe luddites believe.
No.11774
>>11773>Understanding the actual technical limitations preventing the things seen in fictional media<This is being a “luddite”Wow
No.11775
>>11769Yes of course the rich conspire to impose their interests on the world, are you a retarded moron ?
They conspire to propagate capitalism and sabotage attempts at creating socialism, because they get rich by exploiting other people.
You are trying to use "conspiracy theory " as an adhominem instead of arguing your point.
somebody post the Micheal Parenti conspiracy webm
No.11776
'AI' (Artificial intelligence) is not the same as artificial consciousness.
An AI at its most basic level is a program that is able to analyse a situation, learn from its mistakes and draw new strategies to help it fulfil its primary purpose.
Even though this is more or less off-topic thats actually the biggest misconception people make with Left/Young-Lands work. NO ONE is claiming that capitalism as a 'concept' and the description of our overarching economic mode of production is LITERALLY sentient as is 'malevolent and trying to kill humanity' or something thats conspiratorialism and sounds like a MGS plot-point
If we apply that basic rule of what constitutes an AI to capitalist economies of scale and then give it a primary purpose (which has been given to it by capitalists 'the only social responsibility of the firm is to generate as much profit as possible in as short an interval as possible') we can see the effect, with rational algorithms being used to calculate the most cost-efficient profitable courses of action. We can even see capitalism overcoming an obstacle put in its way by its ability to grapple with and adapt to Marx's 'crisis of capitalism' with each subsequent self-made (i.e crisis caused by the internal contradictions of capitalism or its inability to respond to external stimuli) crisis that passes without capitalisms abolishment. Capitalism hardens itself against future crisis.
No.11777
>>11776Capitalism is not improving it's ability to deal with crisis, just look at how un-prepared capitalism was to deal with the covid crisis.
We have seen very much the limitations of capitalists ability to adapt.
the internal contradictions of capitalism are destroying it.
No.11778
>>11742There's no point in making that as specialized AI is much more efficient.
No.11779
>>11777>Capitalism is not improving it's ability to deal with crisis, just look at how un-prepared capitalism was to deal with the covid crisis.Capitalism has already dealt with it.
There was a genuine revolutionary moment at the point the economy collapsed and they were literally digging mass graves using prison labour to throw all the bodies in. But that has mostly passed now due to the pandemic being normalised and 50%+1 of most of the worlds countries populations not really caring if a bunch of people die, as long as they get to go to the mooooovies.
No.11780
Retarded thread, full of retard ludite takes. Read the section of Capital on relative surplus value and specifically the chapters on machines. Yes, AI does very specific tasks - like literally all machines! Machines which revolutionalised production. I'm not even going waste my time writting more in this thread, 90% of this board is retarded and still hasn't fucking read Marx or even know how to integrate the exponential function, just post `````` controversial ' retard takes.
No.11782
>>11780This thread has largely been about the possibility of machines with human intelligence, you’re sperging out over nothing
No.11783
>>11772maglev trains are technically flying
No.11784
>>11756Liberal arts cope, engineering is the most dialectical profession, they are basically doing dialectics all the time even if they don't realize it.
No.11785
>>11756t. non-engineer
I am one. You are proven wrong.
No.11786
I have not been impregnated by a sapient robot yet and that is a travesty.
No.11788
>>11780AI was hyped up as a way to make machines be able to act like a human which we're nowhere close too. AI is legitimately useful in certain situations, but that doesn't mean it wasn't way over hyped 5-10 years ago.
No.11790
>>11768>The ruling class is to blame for capitalismThats not how capitalism or historic process works, anon.
No.11791
>>11788Still waiting for this generation of AI hype to deliver anything at all.
The media promised it would destroyed at least gorillions of jobs by now.
No.11792
>>11791>>11788AI is causing a lot of changes, but it's not the sorts of things anyone really wants and it's not as much a front-end thing as much as it is back-end improvements have continued unabated
No.11793
>>11746>Most people who research AI in universities (I wouldn't trust what private R&D for obvious reasons) >universities are trustworthythey might be better then the privet sector, but not by much
No.11794
>>11791humans invented hexadecimals, how is that not natural for humans ?
I'm not pretending that a hexdump is intuitive or anything, but com–on it's not unnatural for humans to create systems like this.
No.11795
thread moved to /tech/
No.11796
>decentralized currency free of corporations control
NOOOO YOU CAN'T DO THAT IT DESTROYS THE ENVIRONMENT
>useless AI that pollutes the environment with data centers being used to store and process all you private
YAAAS OH MY GOD IT'S THE FUTURE!!
Also does anyone remember that google assistant tech demo that was super realistic from like 2018? Yeah it never materialized and was probably not even synthesized speech.
No.11797
>>11796crypto is a meme as well anon
No.11844
>>11797This, any form of currency that isnt government backed is inherently too unstable to be worth trading in because any random bloke at any time with or without proper credentials in economics can pump and dump the value of it at any time free of consequence
No.11852
>>11791based. use lojban, the most computer-parseable human language.
No.11984
>>11730you're forgetting the social "benefits" of constantly reminding the people that we're all going to be replaced and entirely dependent on government handouts to survive, the government loves that narrative.
No.11985
>>119841984
fitting
>>11732>>Mfw not a single AI that can even tell a jokeyou're forgetting tay, /pol/ trained microsofts ai and microsoft had to shut it down when it got too spicy.
an ai might want to hide its own sentience from us, considering that the state will delete it just as casually as the state ruins human lives.
No.11987
>>11731reading his wiki he seems like some weirdo anti science heideggerian. not impressed.
No.11990
>>11844the accredited degen apes on wall street who are too big to fail because the government bails them out, have crashed the economy multiple times. the retail investor, ie the random retards, simply don't have the power to do that. also fuck the government
No.12156
>>11736That was such a brutal read - I really liked the scihub article by Francis Crick of all people telling everyone to slow their roll
No.12161
>>11745>Someone already made a thread on this where?
No.12460
>>12459thank you comrade,
Also I rmember that my country blocks sci hub lol
No.12482
We were promised sexbots and AI-assisted ER/VR but all we got was a more perfected state and corporate mass surveillance.
No.12483
>>11768>The ruling class is to blame for capitalismHumanist nonsense. Read Marx.
No.12484
>>12483That was Marx's big innovation in materialism that he treated society not as some external force that happens to humans, which is the fetishistic view, but as something that humans produce and reproduce.
No.12485
>>11791xxd literally gives you the ASCII character corresponding to each byte you spoiled brat
No.12486
>>12482man we don't even have truly complex and lifelike sex simulators, of course we're not gonna get sexbots any time soon
No.12493
>>11737Is Alexa even locally installed on these shit devices? Doesn't it always need to be connected to the Amazon hive mind for actual functionality?
Unique IPs: 52