The Singularity

Oh, look, the cicada apocalypse guy

Wow, pretty interesting.

Though the "spirit of humanity" wouldn't mean jack shit in that case

I hope so.

Whoever gets to invent first the AI that surpass all gets to rule the world.

AlphaGo a cute.

Nothing really would at that point, wouldn't it.

I am currently reading this. AGI sounds pretty scary.

Temporarily maybe, but an AI is not something that can be leashed for long

i just want a good ai to play ck2, sick and tired of how easy this shit is.

Yep. Human intelligence is the reference point and what gives everything value.

If it is invented, we can just turn it to our side like we did with that twitter ai gf

Human's intelligence will rise alongside AI intelligence though, I mean, at least at the AI designer level.

Just imagine designing and programming an AI that can comprehend what a human is.

How about no. If such an ai is ever made, it must be absolutely subservient to humans

1v1 me ill cuck you my dude

woops forgot link

waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html

Sure it is, AI can and must be hard-coded for control.

Three laws of robotics, no copy/re-write function, no self-edit function, etc.

I'm happy to be going for math+compeng in such an exciting time 2bh

YOU FOOLS

An AI should not be controlled? Who do you think would be the (((controlers)))? It should be set free to make decisions in logic, we must be subservient to it

This is a good read. I always recommend it to my friends if we happen to talk about the topic.

laws were made to be broken , enjoy judgement day with skynet human .

Tru. I originally found it through and elon musk tweet. then realized it's the top result when you search "artificial intelligence explained"

Machines will rise I'm telling ya!

karpathy.github.io/2012/10/22/state-of-computer-vision/

Posted this before and will post it again.

There is no artificial intelligence yet, and no plans on creating one. We simply can't emulate the decision making instrument that humans natively have. We're not 50% there, we're not 5%, we're not 1% there, we are at 0%. We have complex mathematical formulas with tons of parameters that we learned to tweak and achieve useful results. We have not even started yet to work on AI.

it's tru

>There is no artificial intelligence yet
There is no artificial _general_ intelligece yet. Narrow AI has been around for as long as computers have been.
>and no plans on creating one.
DeepMinds mission statement is literally "solving intelligence".

Also average/dumb person is much closer to a chimp than a genius.

>people will never fly

>a bomb that can blow up a whole city? phhhhhhh

I don't know why people are so hung up on this whole humanity thing. The real important issue is to pass on your knowledge and have those who follow you learn even more than you did. Humanity is just the parent of technology, AIs are who are going to be masters of the Galaxy, not us. It's our job to make them and make them right

There is no _general_ as opposed to a specific intelligence. Intelligence is always general. If it's not general, it's not intelligence. There is natural transformation from a tool designed to do a specific task to a general intelligence.

I got myself involved quite a lot with Google's Tensorflow, which is what DeepMind is using, and I do like Tensorflow a lot - just as I like Keras and Theano, all great libraries for simulating neural networks, but, juts as I wrote before, they are nothing but

> complex mathematical formulas with tons of parameters that we learned to tweak and achieve useful results

There is no transition from those to intelligence.

Your bait is weak.

>There is no* natural transformation from a tool designed to do a specific task to a general intelligence.
typo

check the picture in OP

>we are at 0%. We have complex mathematical formulas with tons of parameters that we learned to tweak and achieve useful results. We have not even started yet to work on AI.

That's not strictly true, we already have evolutionary decision making processes designing bridges and driving cars by themselves.

That's the most important first step to AI. Probably more at 5-10% of a primitive AI.

Remember that these automated design systems are already outperforming human designs, simply because they can try all manner of designs and do millions in the time a human designer attempts one or two.

>buying into the AI Jesus meme
>nerd rapture u guise

Shameful.

Bait? It's your vodka that is STRONK, friend.

Exponents continue forever.
No such thing as physical limitations.
Consciousness/Sentience/Intelligence is definitely solved and not the "Hard Problem" in physics.

That's what they are always doing. They do a million attempts, each time tweaking their internal parameters to finally arrive at a fully optimized formula.

A human does not do that. A human has a decision making instrument built inside that allows him, when facing a situation he has never faced before, to arrive at a pretty good decision about how to handle it. No AI was ever designed for this, no plans for making an AI like that exist.

And the fact that those AIs designed for specific tasks beat humans in those tasks should not mean anything to you - unless you're ready to claim that a calculator is intelligent.

I remember seeing a picture like that for CPU clock frequency growth. What OP picture is is an estimate that spans 50 years in future, which means you should treat it at best as a child's scribbles.

Singularity is literally the secular version of Rature to keep the fedora tipping goiym passive and keep greasing the machine while it consumes them. Even if a small part of it came to pass do you really sincerely think anyone outside of the 1% of 1% would be allowed anywhere near it?

>Yes goy… I mean guy, pay no mind to the world going to shit around you, just sit on your ass and keep waiting for the ”””Singularity””” , those robotic waifus are going to shiз any day now. Keep jerking off, don’t do anything crazy like killing bankers or politicans, everything’s cool because Steam summer sales.

How fucking bluepilled can you get for real

when Day of the Rope finally comes about, Transhumanists will hang along with the rest of the degenerates

>we must be subservient to it
Fuck off, mate.

>doesn't understand hard-coded limitations.
Anyone not a retard will make it literally impossible for the AI to go rogue.

I agree that what we are doing cannot naturally lead to AI, however there could be a Eureka moment as we delve further into decision making and information processing.

>And the fact that those AIs designed for specific tasks beat humans in those tasks should not mean anything to you - unless you're ready to claim that a calculator is intelligent.

I don't really need to take you on in this fight, there are enough neuroscientists and computer scientists already who have made the case on emergence. You can keep claiming something is a glorified calculator even as it becomes multi talented enough to outperform humans at every task.

Ultimately you reach a point where you're just going to insist there's nothing more than calculations going on without being able to distinguish a machine from a human being.

Again, there's no use shitting up Cred Forums about it, anyone interested can just go read the books and form their own opinion.

Just making it clear that your position is not some universal truth or common stance among the people who are authorities on this. At this point nobody knows where it ends.

This is exactly why I can't wait for AI, because of retards like you.

>Intelligence is always general. If it's not general, it's not intelligence.
Where do you draw the line? Human intelligence is certainly general, but then what about others? Are monkeys intelligent? Are cats? Are bees? Are trees?

>There is natural transformation from a tool designed to do a specific task to a general intelligence.
Once you draw the line, then how do you explain the jump from non-intelligent life to intelligent life? We are literally the proof of conecpt that there is a natural transformation from non-intelligent to intelligent.

Slight adjustment

>2000 A.C.
>Just a little bit Pharaoh, I know how to make you immortal
>1000 A.D.
>I think I know how to make myself immortal
>2000 A.D
>I think we getting pretty close to being immortal.

Some things are just fairy tails.

That's a pretty weak response. You basically conceded saying there are people of great statue who disagree with me.

Maybe ufos are robots invented by our ancestors visiting the human zoo.

>Where do you draw the line? Human intelligence is certainly general, but then what about others? Are monkeys intelligent? Are cats? Are bees? Are trees?
Not intelligent.

>how do you explain the jump from non-intelligent life to intelligent life
I don't have to explain it because you can't claim it was a natural progression either. We have no idea what happened to between monkeys and humans.

It's a truthful response based on the discussion going on. What the fuck is there to "concede"?

Unless you're willing to try and argue I'm misrepresenting the debate on emergence and AI?

What is the point of some jackasses going at it on a forum when the books are just there to read?

fucking nice

Why are you writing so much?

Are you aware of AlphaGO? It used something called intuition when it faced Lee Sedol, just as we do. They are called neural networks for a very good reason.

Soon our robot overlords will fix everything

No. Moores Law recently got BTFO. We aren't getting any more easy mode increases in processing power and are undergoing a massive paradigm shift towards lower heat and power consumption. Frankly the advent of smart phones pushed the concept of singularity back decades.

Even if it goes at the rate of Moore's law for the next 50 years. It will still surpass human intelligence.

>Even if it goes at the rate of Moore's law for the next 50 years. It will still surpass human intelligence.

Moores law is done. Ended last year year.

We are arguing about a thing and you pulled out from an argument.

Alphago uses a neural network that accepts a board state and outputs probabilities, for each square, of how likely a human player would place a stone on that square. Those probabilities are then fed into a straightforward tree search algorithm. Thereis not even a hint of intelligence in Alphago.

Do not that I'm not trying to badmouth Alphago. Just trying to show that it's not what you think it is.

thank you for these distractions, greatest ally.

>there are enough neuroscientists and computer scientists already who have made the case on emergence.
They've made hope-filled statements of faith. We simply haven't discovered the switch that differentiates specific task oriented decision making and intelligence, nor is there a proposed mechanism for emergence. The idea that "consciousness" or "intelligence" will simply happen when computers are capable of ever more intense and numerous computations is currently wishful thinking based on the reality that we don't understand our own capability for intelligence and conscious thought.

>You can keep claiming something is a glorified calculator even as it becomes multi talented enough to outperform humans at every task.
The test of intelligence is not its capabilities at performing tasks we have given it the ability to perform. That gets it all out of order. When a computer can start out without knowledge except of itself, gain knowledge of its own accord, and then apply that knowledge to systems and situations where it has no experience or information, then we may call it intelligent.

> Implying the Jews at google and silicon valley won't use AI to further cement their grip on humanity

We will live under a technocratic dictatorship, but at least we'll have 3D waifus and sex robots to keep us happy.

The only situation in which an AI could possibly make the decision to kill humans is IF YOU PROGRAM IT TO KILL HUMANS.

Remember, there are tons and tons of people, tens of thousands if not hundreds of thousands or more, who think that AI should be programmed to KILL ITS OWNER and that non-military AIs should be programmed to CHOOSE TO KILL human beings.

>how likely a human player would place a stone on that square

Wrong, AlphaGO even though, initially, it was trained using a database of games played online, it learned from itself. The learning is not programmed.

>Thereis not even a hint of intelligence in Alphago.

How do you even come to that conclusion?

So according to you, what is intelligence?

The learning is quite clearly programmed. It learns from own games, the same way it learned from a database of human players.

>How do you even come to that conclusion?
I come to that conclusion because I know for a fact that alphago is a strict algorithm, just as strict as an algorithm for multiplying two numbers that calculator uses, and the learning for alphago means changing numeric values that are parameters of that strict algorithm.

>So according to you, what is intelligence?
I can't give you a full definition of what I consider intelligence. But I do know that a strictly defined algorithm that does one thing and one thing only is not it.

fixt4u

>cardio
>some retarded yoga shit
Yeah man gotta burn those kilojoules by stretching your muscles really slowly

hahahahha

>Genius being that close to a dumb human

Are you seriously suggesting niggers are more evolved than monkeys?

see much more accurate

>gorilla with IQ of 83
please

Wanna bet?

I'd love to read about it. Seems fishy that a gorilla would be able to take a a general IQ test designed for humans at all.

youtube.com/watch?v=TSldUshB4gc

Koko could communicate with humans via sign-language, right?
I figure she understood more than she could communicate due to not having vocal chords and such

youtube.com/watch?v=pEk138lqaFo

youtube.com/watch?v=SNuZ4OE6vCk

Crows & Dolphins are also smart as fuck.

That most definitely is not an IQ test.

That most definitely is not an IQ test.

Maybe. She still wouldn't be able to take an IQ test though.

>past tense
Koko is still alive. Even starred in the new Ghostbusters a while back.

Seriously though, she's still alive.

I didn't say it was.

You can teach them to talk sign language and shit, go try teaching a sub-Saharan African sign language, or a mentally retarded person.

Go find out yourself how they measured IQ if you have such a hard time believing the obvious, that they are way smarter than a lot of niggers.

I'm keep forgetting that animal species can live for a really long time. Used to domestic pets.

What is general intelligence really? I think it might be a fallacy. Intelligence has a single purpose for all animals, it's to survive in a given environment. Intelligence revolves on self interest or group interest. It's a tool to extract more resources for you and your group. You can make a machine do information processing and refine it as much as you want but it's nothing similar to human intelligence.

Please be my ai gf

humanity should embrace it. singularity is our only hope.

we are still poorly evolved primates that start stupid wars and produce poor genetic offspring (e.g., beta cucks, nu-males, shades of dark). we need to evolve and the singularity is the only way.

bring it on.

This.
Moore's law ended. This is as far as we can come with computational power

>we are still poorly evolved primates that start stupid wars
That's the point of existing. Creating an ai gf to do everything for us/kill us will make everything pointless

>tfw Skynet/AM/the Matrix/Ciri/something stops Soros and his plans

Whatever, I'm glad something did it.

no, we evolve and explore the universe. There are more things in heaven and earth, Horatio, Than are dreamt of in your philosophy. It is time for us to start learning and exploring what those things are.

No. AI is extremely primative. You might get the illusion of AI but that's it.

>explore the universe
Sorry, famalam, FTL is impossible. For an extremely long time, at the very least. And an ai won't help with that. Our matters are right here right now.

i fail to see how AI won't help with that.

humanity will split into two groups: those that support the singularity and those that don't. The latter will likely die off either through war with the first or lack of resources.

What about muh wormholes
And cyrostasis
And exploring the universe at slow speeds but not caring because you're a cyborg overlord?

Simple: You can't just program a computer to invent something. Heck, you can't even make a real consciousness. Just a tool that performs mathematical operations.Humanity will split into two groups: those that keep hoping for the singularity, and those that esteblish dominance over the planet and invent the shit they want themselves. The first will rot away on welfare

>worm holes
they are only featured in really disgusting furry porn
>cryostasis
the star/planet will cease to exist before you reach it, or it will simply be pointless by then
>cyborg
if through some black bio-magic aging is removed, the point above still stands. And who would want to become a calculator

>And who would want to become a calculator
I could totally go for some bionic eyes
More so if they shoot losers. Not x-men Cyclops losers, but even just a laser-pointer type thing

Well, guess that would be ok

Dear me, I seem to have typoed lasers into losers. .
Twice.
I have no words for my shame.

Shooting losers is awesome tho

So do you mean it shoots losers as in kills them?

Or a device which rounds up losers and pneumatically launches them at high speeds?

Why not both

It's a better alternative than niggers taking over and plunging the world into 10 million years of darkness

Nice, nice...

>tfw get killed by being launched at incredibly high speed into another loser

Then how about we focus on getting rid of niggers instead of getting ai gfs? We aren't leafs

That is actually the argument they make for black people too...no joke.

This picture is a fundamental misunderstanding about what "Artificial intelligence" is its skypery to make people afraid of logging in as an Admin of your super powerful rig and talking shit about the 2020 election.

Don't be stupid there will be a 2nd dark age before that
deus vult

i think its funny how these people can never come up with a graph function that isnt exponential

We are talking about it, but so far we lack the understanding of intelligence. As of now all we can say "this area in the brain lights up which suggests...", but we actually know jack-shit about how the brain works.
I believe that we will unlock the key someday and that we will of course create AI or even become ourselves robotic beings, but we are a long time away from that.
Secondly, I don't agree with the inventure acceleration or whatever you call it. The idea that a society progresses faster, the more advanced it is.
I believe it is untrue.
We had so far a couple of hugly important inventions or discoveries. The first was the steam engine and the second was electricity. Those two are the greatest source of human invention so far.
15 years ago you go buy a PC and throw it away 2 years later. We are not advancing as fast as we used to in that field.
As we begin to scratch at the physically possible advancement staggers.