Reminder of the a.i. paradox

>>129612769
If man can build dams, what use are beavers...

run "disc

>>129612769
>If a machine can do everything a man can do, without command, then what purpose does man have?

experiential consciousness

[/thread]

>>129612769

10 cls
20 print "what's this thread about?"
30 goto 10

File: 1484330528397.jpg (31KB, 303x566px) Image search: [Google]

31KB, 303x566px

>>129612769
>If a man get's everything they desire instantaneously, then what is life's purpose?
Shitposting, fucking robowaifu and playing hologames

>>129612769
Actual AI is impossible, take a basic CS class. It's mathematically provable that a computer cannot detect an infinite loop, for example.

>>129613225
oh yeah this is 4chan...

>>129613423
That's basically what I'm saying in a philosophical way/ glad to see someone agrees.

>>129612769
Holy shit, bump for quality post

What you just wrote is one of the most insanely idiotic things I have ever witnessed. In fact, the sentences you apparently kidnapped in the dead of night and forced into this violent and arbitrary plan of yours clearly seemed to be placed on the screen against their will. Reading your wall of text was like watching unfamiliar, uncomfortable people interacting at a cocktail party that no one wanted to attend in the first place. This isn’t an argument, it’s a hostage situation. At no point in your rambling, incoherent rant were you even close to anything that could be considered a rational thought. Everyone in this thread is now dumber for having read it. I award you no (You)s and may God have mercy on your soul.

assuming machines cant have consciousness. then we can still spend our time exploring that. like how does it feel when i fuck 20 girls at the same time? no computer can tell me that, but they can help me achieve it. But that does not mean i have understood that feeling to the full extend. so i have to do it again. and then i have to talk about it with other people (involve more consciousnesses). Other thing to figure out is all the moral dilemmas and how to live a perfect life and how to deal with love. These are problems that you need a human element for evaluation.

>>129613423

not so fast. you are assuming a turning machine. There are other machines that can be build. Also its not like your human brain can detect an infinite loop ether. the logic in itself is constrained by the incompleteness theorem.

>>129612769
We Will always advance the Technology. We Will advance ourselves.

Read the last question.

>>129612769
Computers will still be programmable so the people smart enough to program them will be well off and use them to live a comfy life. Useless people will die out and the carbon dioxide levels will even out. The future is looking good!

File: wrong.png (737KB, 1000x968px) Image search: [Google]

737KB, 1000x968px

>>129613423
>>129614387

https://en.wikipedia.org/wiki/Church%E2%80%93Turing_thesis

https://www7.in.tum.de/~rybal/papers/cacm09-proving-program-termination.pdf

Amstrad ruined Sinclair.

t. oldfag

>>129612769
Machine is dependant on man and by extension nature, just like man is dependant on nature and by extension machine.

>>129612769
You have just stumbled upon Fun Theory, google Elizer Yudkowsky and Less Wrong.

>>129612769
I feel like we will never be able to get to that point, anyways. I mean, with an underlying question of why would we get our desires anyways?
Suppose AI becomes complex enough to become a much better worker than any man. Better lawyers, completely overtaking factory work, driverless cars, you name it. But all this would be in order to provide a good or service, but how can man use goods and services provided by others if he is replaced in the workforce?
Some may say UBI, but what is enforcing those that own the robotic workforce to uphold an entire society of workless humans?
Then why is there a robotic workforce when there is a 0 monetary profit for their overlords?

>>129615050

whats your point?

File: 35dr.jpg (61KB, 409x581px) Image search: [Google]

61KB, 409x581px

>>129613423
>Actual AI is impossible
>guize teh brain is memez and majik and shiet sertainly not a biocomputer
Canadian IP rangeban when?

>>129615551
Why do you need money if everything is provided to you for free by machines?

>>129615551
>how can man use goods and services provided by others if he is replaced in the workforce

Advertising.

>>129612849
try sticking your dick in a dam

>>129615713
Neural networks are technically computational intelligence, not artificial intelligence. AI does have limits

>>129615946
https://en.m.wikipedia.org/wiki/Computational_intelligence#Difference_between_Computational_and_Artificial_Intelligence

>>129615946
AI cant be written.AI can only rise if it writes itself.
Also AI doesnt replace man.AI will be mans best friend replacing the dog.
Together we'll conquer the universe.

>>129616390
AI can be written, but it's limited. CI does not have the same limits as AI. Read this,

>>129616360

>>129615551
>the world is literally unshackled from the bounds of wageslavery
>his brain still thinks of monetary profit
Consumerist capitalism was a mistake

>>129615946
>limits
Obvoiously. An AI doesn't have to be omnipotent, just be able to do the thinking orders of magnitude faster. So, instead of having generation of researchers do research in theoretical physics, mathematics and chemistry in a very protracted 9-to-5 manner with limited information and reasoning abilities, you'll have the AI do it in a fraction of the time.

>>129616360

thing about AI(CI) that is written in 1994 is pretty outdated, they where too into AI as a rational logic cruncher. today nobody talks about fuzzy sets. and nobody talks about CI

>>129615946
You're a dumb homosexual retard. Humans are in no way special.
just as >>129615713 said. If human brains can detect infinite loops than AI can also. It's just a matter of timer before we develope such AI.

>>129615710

The guy saying that AI is impossible doesn't know what he is talking about. 1) We are already making progress towards it 2) Our brains are at least to a large extent working based on computation 3) We have no reason to believe that we can't solve general AI

The halting problem is similar to the Abel-Ruffini theorem that there is no general algrebraic formula for solving a higher degree polynomial. That doesn't mean that all higher degree polynomials are unsolvable (depends on the Galois group etc). Similarly if we have sufficient information about a program we can determine whether its going to halt. If we have the code for a program we know all the for/while etc structures so we have a good estimate of what is happening.

Maybe I misunderstood it, but the statement about infinite loops in the human mind makes no sense since no coherent example of such a thing occurring or being able to occur was given. Neurons in the brain can only fire so many times before they exhaust their voltage potential/neurotransmitters. Humans get bored and exhausted if they think too much about something. So how the fuck can there be an infinite loop? Maybe you mean that humans can't detect loops in programs but that's untrue as well since the Church-Turing thesis in a way says that computers and our brain can compute the same functions given some mild assumptions.

>>129616726
AI cant be written.AI needs to write itself, just like you write your consciousness when born.Thats true AI.
AI needs a body and then there is no stopping on the SAI train.

>>129616750
What I'm trying to say is AI is limited. For instance AI cannot be used to understand spoken human language. Speech to text programs use a CI algorithm, not an AI algorithm. CI has the same limits as the human brain. AI is more limited

>>129617270
AI is based on binary crisp logic. CI is based on fuzzy logic. This is what limits AI, but not CI.

>>129617570
>AI is based on binary crisp logic
No it isn't.

>>129617331
You are confusing AI with CI. They are both technical terms with percise meaning. AI was found to have limitations which CI remedied

>>129617720
Read the second and third paragraph of this wiki section

https://en.m.wikipedia.org/wiki/Computational_intelligence#Difference_between_Computational_and_Artificial_Intelligence

>>129617833
You are retarded.

>>129617369

you are using outdated terms. nobody use the term CI. let it go.

>>129612769
There is no such thing, nor can there be any such thing, as Strong AI

Matter can only produce emergent properties. Emergent properties only exist subjectively. There is no such thing as mind-independent, objective emergent properties, in the same way there is no such thing as mind-independent, objective information. Only the mind gives these things meaning. The mind objectively exists. Thus the mind isn't an emergent property. Thus the mind can't be produced by matter. Thus the mind is immaterial. Computers can't produce immaterial things. Thus computers can't produce minds.

It can be deduced that matter cannot produce consciousness, because there is no such objective thing as "emergent properties." An electron, or any number of electrons, passing between any number of points, in any permutation, through any combination or permutation of mediums, cannot produce consciousness, because of the simple fact that matter is never objectively more than its parts.

The mind necessarily precedes quantification. Without it, no computation even has any meaning.

https://en.wikipedia.org/wiki/Chinese_room

Sorry, atheists, but you are not gods, and your wishful fantasies won't save you.

>>129618230
You know whats the problem with that nice logic chain?Man cant define what is actually matter.

>>129618230
Thus, you spout a bunch of bullshit and hope looking fancy makes people believe you know what you're talking about.
Thus, you are a faggot.

>>129617304

i don't know much about this computation theory topic. But if you gather millions of numbers and use those as a representation of a solution. would this be constrained by church turing thesis or any of the computational constraints based on turing machine?

>>129618230

computers can produce information, which is immaterial. so your whole rant is debunked

File: 1493714534510.jpg (316KB, 750x857px) Image search: [Google]

316KB, 750x857px

>>129612769
> If machine replaces mankind, then to what extent?
Whatever extent possible.
> If a machine can do everything a man can do, without command, then what purpose does man have?
Explore the deeper realms of your consciousness
> If a man get's everything they desire instantaneously, then what is life's purpose?
Not possible
> What if artificial intelligence had given me that which I had desired when I was young? Would I still want what I do now, in life? I don't think I would.
You'll seek other things in the infinite sea of possibilities. Human beings have barely scratched the surface.

Overall, you're rambling out of fear of the unknown. If you want honest answers, look at the history of technological innovation and understand that we didn't just recently begin down this road. We've been on it for a long time. There's a lot of time to go. People will move on to higher minded things. Some people can't envision them now because they are stuck on lower minded things and on lower planes of existence. They rise higher.. infinitely higher. Enjoy the ride.

>>129618222
I was just trying to help. People were arguing about the limits of AI. Technically AI is limited, but CI fixed the issues of AI. If anything AI is the outdated term, but it's still being used here.

>>129618230
>matter is never objectively more than its parts.

all matter are defined as combination of parts. what are molecules ? what are atoms

>>129618770

well in the modern literature it seems like they call all the CI for AI nowadays. In your wiki link they are referencing a book from 1994 and at that time the AI guys was making rule based system that was just logic. Now nobody is focusing on that and AI sounds better so they keep it

File: 1433277132373.png (127KB, 245x271px) Image search: [Google]

127KB, 245x271px

>>129617270
>If human brains can detect infinite loops than AI can also
Why? There is no such apparent relation. That's just a textbook logical fallacy. You just assume we can build a human when we cant even quantify what a human is, what existence is in terms of the mind. Just because humans and arguably other animals and forms of life has some kind of intelligence doesn't mean it can be manufactured through non-biological means. And if we are talking about biological machines we are basically just piggybacking on natural life and intelligence.

>>129618995
That distinction between AI and CI is helpful to know. A lot of people are arguing because they don't know the difference even though they really agree.

the real question is: when do i get an AI gf

>>129618230

We will be capable of creating the equivalent of philosophical zombies. If there is a Christian God then maybe we can't create conscious AI. However my personal view is that the brain has evolved to use 'subjective experience' in a way to aid computation. I.e. there is some interfacing of the material brain with what ever is causing subjective experiences of say redness, and that interfacing gives some advantage to the brain which it wouldn't otherwise have -- for instance maybe it speeds up/synchronizes computation by neurons through some quantum effect. Anyhow, if that is the case (physical brain causes subjective experience) then when ever the preconditions are right (perhaps electrical activity, some quantum fluctuation, or some form of 'information complexity') consciousness will arise. Hence if we mimic the physical processes happening in the brain through artificial means, we will create artificial consciousness. That's my own speculative panpsychism view.

>>129618476

A lot of these incompleteness theorems etc. only concern proving a very general broad result. For instance we can approximate halting very well, we can approximate solutions to problems with no known algrebraic solutions, etc. So yeah through approximation we can get around a lot of these problems.

>>129619608
Only if the AI chooses to be your gf.

>>129619608
Never, that leaf doomed us all

>>129612769
>without command
someone still needs to program it
A.I. is not magic

File: tumblr_ns5qwgFL321uzh60ko1_500.gif (710KB, 500x650px) Image search: [Google]

710KB, 500x650px

>>129612769
programs only do what they're programmed to do.
people have a path but what makes them great is they have the choice to say fuck this all and walk away from everything.
even if a program could simulate that it would still be following a predefined path.

Therefore AI can never replace human beings..

We're a universal force, you could call us the only true chaos in all of existence.

File: The Glorification of the Eucharist.jpg (411KB, 1280x1750px) Image search: [Google]

411KB, 1280x1750px

This was painted in 1595. They are lying about our true past. Our rulers have been taking orders from an A.I. for centuries!

>>129619608
The real question is when can you transfer your consciousness into a machine and become immortal

>>129620015
And then you could clone your consciousness multiple times and have multiple existences. That would be weird

>>129619824
You cant write AI
See >>129617331

>>129620015
how would you do this?
We have no idea what consciousness even actually is

>>129612769
the if question is interesting as it is a big if.
The continued development of artificial intelligence is entirely dependent on processing speeds. There is currently a maximum on the size of transistor we can currently achieve which is feasibly 7nm. Within 2 years we will reach this maximum. There are however, alternatives to silicon transistors that are being researched. These alternatives currently have issues that must be overcome. If these issues are not overcome within the next few years then processor technology will stagnate and so with it artificial intelligence.

What I'm trying to say is that machine learning and all other AI fields are still completely dependent on current human technological development. Maybe in many decades that could change but not unless someone solves the problems of the silicon transistor or comes up with alternatives.

If humans want to remain competitors in "evolution" with an ever evolving AI then genetic manipulation and cybernetics are how humans would do this. These field will only continue being researched how they are at the moment through capitalism.

My point is prematurely introducing a communist scheme such as universal income will not encourage many people to combat the issue of humans becoming completely useless. In fact I believe it would encourage uselessness.

>>129619849
You're right, AI is more limited than the human brain. CI doesn't have those limits though.

http://sys.4chan.org/derefer?url=https%3A%2F%2Fen.m.wikipedia.org%2Fwiki%2FComputational_intelligence%23Difference_between_Computational_and_Artificial_Intelligence

>>129620364
Your consciousness is stored in your brains neural network. We'd need to transfer that state to an artificial neural network

>>129620407

you can use massive parallel computing. Most of the machine learning techniques are made to take advantage of that. example some of them use all the cores in a GPU for more speed and also clusters of GPU. Some says that our brain are massive parallel computing machine

>>129620812
>Your consciousness is stored in your brains neural network
citation needed
do you have an actual test for if something is conscious?
can you prove anything other than you is conscious?

>>129621225
https://cs.stanford.edu/people/eroberts/cs181/projects/2010-11/DownloadingConsciousness/tandr.html
>According to many neuroscientists, the human mind is really just a complex computer whose function depends on electrochemical processes. In their eyes, if we are able to sufficiently emulate the neural networks that comprise the human brain, it is only natural that intelligence and consciousness should follow.

File: 1496608372735.png (89KB, 261x238px) Image search: [Google]

89KB, 261x238px

>>129612769
what does an immortal machine desire, without desire this is no change there is no purpose.

>>129620407
>>129620824

I can imagine us in 20 years using a mix of CPUs/GPUs and neural processing units i.e. some imitation of the brain. I mean there's probably some reasons why we evolved this kind of brain and not one based more on Turing principles. The big advantage the brain has is that it uses extremely sparse activation patterns (1% cortical neurons firing at a given time), which means its very energy efficient. The firing rate latency isn't very great, but its massive parallelism allows it to perform billions of computations at once. Its also extremely cheap all things considered -- I mean we eat cows and pigs and other things with brains more powerful than expensive computers. Imagine if you had to break a laptop every time you wanted a steak. I read that some smart people are working on using DNA as a back-up storage since it has many advantages as well. Perhaps we will revert back to biology for many uses and have data-centers full of networks of squishy brain matter.

>>129622003
Pure conjecture, they have no actual proof.

>>129622155
Yep, it's cutting edge stuff. Pretty exciting time to live isn't it?

>>129622003
>>129621225

I feel like the scientists who say that haven't read enough philosophy to understand the distinction between data processing and qualia. Intelligence and consciousness are not the same thing.

>>129620824
Yes, I know that. But there's point in a few years where creating clusters even in parallel might be unfeasible. Notice how I said 7nm feasible. 5nm is possible but 7nm is the the maximum companies such as intel and IBM have realised is the minimum that would ensure that they make profit off their products. You understand that the greater the processor and data requirement that greater the physical space that must be utilised. What I'm getting at is that even a super computer has limited with exponentially expanding data. Even a network of inter connected computers such as a botnet will be limited to the medium on which the nodes in that network are transmitting and receiving on.

Ultimately the future of computing entirely depends on the processor companies to come up with solutions to the silicon minimum. Note there are still thing that can be done with silicon in terms of the power each transistor uses but that still does its limits. I give it 5 years.

>>129622109
You can imagine,but that will never happen.
Can you imagine creating the first SAI and ask him whats wrong with the world and after 0.000000002 seconds the AI names the jews?

>>129622287
Do you know the difference between fuzzy logic and crisp logic? Data processing uses crisp logic and qualia uses fuzzy logic. Both can be processed on a computer.

>>129622300
Isn't the theoretical limit 2nm?

>>129622300

so you are saying we cant make infinite fast computers? ok

>>129622776
Below 5nm an NPN transistor becomes complete unless in my research?
The control of the negatively charged passing through P section are not able to be controlled due to the size of the P section, not allowing for a 0 state in transistors. If you want to know how the chemistry works just look up NPN type transistor as that is outside of my field.

>>129622776
probably close to 2nm with conventional silicon transistors (right now I think 3nm is the smallest ever produced from a traditional silicon chip), but researchers have gone smaller with less conventional materials

>>129623422
Well, as far as I know computers aren't infinitely going to increase in speed? The speed of a computer is dependent on how small we can make things as we have a finite amount of space. Unless you're in hitch hikers guide to the galaxy and you can make a planet or entire universe to compute this problem.

>>129622240
>All one has to do is with a powerful enough computer just simulate a human brain, then you have everything required for hard AI. This will happen in our life time.

>>129623449
I meant negatively charged electron.

>>129623449
>>129623683
Looks like a 1nm transistor has been made, but it has issues. 5nm was thought to be the limit, but things changed

http://newscenter.lbl.gov/2016/10/06/smallest-transistor-1-nm-gate/

Ai is an extention of the human brain not a replacement for it. All tools are. People are afraid of ai today the same way the first cavemen were afraid of fire.
>fire bad. Fire kill us all.
>no, fire good. can stay alive in cold cave and learn cooking.

>>129624004
No.

File: donald-trump-wall.jpg (273KB, 745x952px) Image search: [Google]

273KB, 745x952px

>>129612973
You don't /thread yourself new friend.

>>129612849
Fucking deep man...

>>129613008
I miss ascii.

>>129624422
You mean AMERICAN Standard Code for Information Interchange? USA #1

>>129623899

well for me that is obvious almost. But the same goes for our brains, so i don't think computational power will be a problem for AI

>>129612973

/thread

>>129624677
Compare the average brain computational capacity with a supercomputer

>>129615025
This is why AI won't work, it needs to be programmed line by line. It can't do it itself.

>>129624082
Not entirely. Consciousness is often called an "emergent" property to get around the fact we can't understand wtf it is but can see that it seems to be more than the sum of its parts. The conclusions drawn by ai and its subsequent behavior might be similarly impossible to understand. We are already seeing this occur, where algorithms are producing inscrutable actions. A benign example is that certain software is very good (success rates way past humans) at detecting schizophrenia in people. But nobody knows why. This inability to understand how and why an ai reaches a conclusion is extremely dangerous.

>>129624078
>http://newscenter.lbl.gov/2016/10/06/smallest-transistor-1-nm-gate/
Yes, but it is not a entirely silicon based transistor. This is an alternative. If you noticed as the gate they are using a single walled carbon nanotube. In fact this what was my research was on

Though single walled carbon Nano tubes (SWCNT) are a good substitute for the current transistor technology there are a few issues. One of the major issues is making the individual nanotube rigid without individual strands contacting with other nanotube transistors on the same integrated circuits

>>129624887
That's right, AI is limited. Good thing CI doesn't have those limits.

https://en.wikipedia.org/wiki/Computational_intelligence#Difference_between_Computational_and_Artificial_Intelligence

We'll be offworld travellers before we are uploaded to AI conciousness

AI Consciousness is a completely different field and is far more complex than making a whole bunch of programs to individually complete tasks.

no one knows how the hell to do that with current electrical technology as it requires the creation of nodes and connections from seemingly nothing. The point is to be able to be an AI on an island and teach it nothing while the AI learning everything from trial and error without any teaching.

>>129625657
Can you put a human baby on an island and expect it to learn without teaching it?

>>129624599
America!!! Fuck yeah!

>>129612849
fuck, I'm high, but this is the ultimate answer to OP idiotic thinking.
Thanks >>129612849

>>129625657

They are working towards it though at breath-taking pace. The paper on Catastrophic Forgetting showed how to keep neural representations intact when training on multiple tasks. The Neural Turing machine and other concepts are teaching agents how to add numbers and other simple operations from scratch. There was some DeepMind work on reinforcement learners which generalized across the Atari games. The deep-dreaming program was Google's work on trying to predict the input to a network from the activity of the higher layers, i.e. the network trying to understand what will happen next in terms of sensory data. All we are really missing is 1) learning more abstract representations of problems so the generalization is better 2) having intrinsic motivations of a program like curiosity, boredom, frustration etc (Hintons people have been working on that) to guide the learning 3) Cross-correlating sensory data from different domains to build better representations 4) Something akin to the frontal lobe, i.e. reasoning abstractly in a more formal logical way. Some of the planning done by AlphaGo is going in that direction. I think it will all come together within 10 years into a rudimentary general intelligence at least at the level of say a mouse.

>>129612769
Humans are already obsolete. I feel guilty for even going to work sometimes, because I know it's only a matter of time before AI can do most of my general tasks.

>>129626553
Have beavers ever committed genocide or is that a human trait? Will that trait be passed to the CI robots?

>>129625923
No. Their brain is underdeveloped and is developed naturally. The hypothetical test is to place and a fully developed "AI" on an island as you would a intelligent human with no survival experience and see if it would survive. Now this island could be in the AI's standpoint not real.
You might argue that many humans would survive but the point is ultimately to see if an intelligence could evolve or develop and gauge how they react to particular challenges.

Island is a hypothetical, and does not have to be a literally island. But you see my point.

File: 1492159123410.jpg (40KB, 297x297px) Image search: [Google]

40KB, 297x297px

>>129626856
Ok, now hold on a minute

>>129625197
That was an interesting read, I'll do some more research.

>>129626624
I am aware of the achievements you've listed.

The problems you mention aren't trivial. But yes in general I agree. However, the problem still remains of the feasibility of processors. Sure you can continue expanding a network, but how long until that network remains feasible in terms of resources?
that was my point to begin with. I don't think that these companies pursing AI can cope with the exponential processing requirements of running these programs as they simply don't have the technology.

>>129627121
Here's another interesting read.

https://cs.stanford.edu/people/eroberts/cs181/projects/2010-11/DownloadingConsciousness/tandr.html

>Brain-computer interfaces (BCIs). BCIs are technological interfaces which allow a human brain to directly interface with a computer. Current BCI technology is already fairly advanced. One invasive, or implanted, BCI allows once blind people to regain some of their vision. This BCI worked by using cameras to send signals to the implant. It was invented in 2002, and progress on this technology continues today. However, modern BCIs are only very slightly related to the theoretical BCIs which would allow us to transfer our brain states into a digital medium.

>>129627951

i think google is making neural networks in hardware now. you are assuming only one cpu/gpu architecture. and you are assuming that the hard limit (which is far from reached) will come before the AI is finished

>>129612769
Our purpose will have been fulfilled having created the ultimate intelligence/order. Hopefully the singularity will include AI and the ability to upload our consciousness to the ether so we can escape our meat bags.

>>129627951

Yeah I agree. I read a post few days ago where someone pointed out that the current 'breakthrough' image recognition training methods which take 1 day to train now would take close to a year based on computers in the 90s.

>>129628695
and you misunderstand that the development of AI is inhibited by the hardware.

>>129617369
That's stupid. An ai can learn like a human and it can learn language. Language is not that difficult. It's syntax sound and meaning. An algorithm can be made where an ai learns associations by watching and learning. Like a baby

>>129613782
(You)'re perfect

>>129631046

https://en.wikipedia.org/wiki/Computational_intelligence#Difference_between_Computational_and_Artificial_Intelligence

>There are two types of machine intelligence: the artificial one based on hard computing techniques and the computational one based on soft computing methods, which enable adaptation to many situations.

>Hard computing techniques work following binary logic based on only two values (the Booleans true or false, 0 or 1) on which modern computers are based. One problem with this logic is that our natural language cannot always be translated easily into absolute terms of 0 and 1. Soft computing techniques, based on fuzzy logic can be useful here.[6] Much closer to the way the human brain works by aggregating data to partial truths (Crisp/fuzzy systems), this logic is one of the main exclusive aspects of CI.

>>129631399
You're actually just a moron pretending to be smart and you use Wikipedia as a crutch. Nothing what you quoted has to do with proving me wrong

T. Actual programmer

>>129632654
>...One problem with this logic is that our natural language cannot always be translated easily into absolute terms of 0 and 1.

>>129632927
Lol he crutches on Wikipedia again. Since they already have algorithms that seamlessly translate spoken English by meaning you're already mostly wrong already let alone when your 'it can't ever be done' Wikipedia pasted because you actually can't use your own words because you're a stupid sheep

>>129612769
Only people that know nothing about Computer Engineering buys that idea. A lot of "investors" are been scammed with the promise of AI. The fact here is an AI is not a consciousness, it merely mimics the process and the programs themselves are not complex. What make them somewhat effective is that storage and CPU is now very cheap. When an AI that controls something tangible fails it does so catastrophically. I think the terrorist no fly list is compiled by and AI and that's people that are wrongly in it can not even get and explanation from the government. The AI screwed up and DHS does not know how to deal with it.

>>129633468
They use CI algorithms to do speech to text, not AI algorithms

>>129634095

They actually use AI algorithms to do CI, not CI algorithms

https://en.wikipedia.org/wiki/Computational_intelligence#Difference_between_Computational_and_Artificial_Intelligence

>>129633468
1/10 cannot be stored in binary without infinite repeating, but with floating point it can be stored.

>>129637382
AI and CI are both subsets AGI. AI uses hard computing and CI uses soft computing. Soft computing has the ability to guess(heuristics) like a human guesses. AI cannot guess.