Why are normies so misinformed on Artificial Intelligence...

Because the only thing marketed to normies is an umbrella term known as 'AI'. This encompasses machine learning and deep learning which are both nothing more than statistical optimization.

It's a meme because its nothing more than glorified statistical optimization hyped up to be something its not for shekels.

Elon tells people its super dangerous and we need to take action to prevent it from killing people. He then says it needs to be regulated. He the goes on and implements a shit tier self driving car program, fights regulation on it, calls it super advanced AI and the dumb lane following program kills people.

For the most part, it's all a sham. Raw computing power is used to probe and minimize error among large data sets and companies pretend its intelligent.

Normies eat it up because they want to be 'believe' and feel like they are a part of something cutting edge.

Prove me wrong.. Hint : you can't.

Damn rare Pepe

Attached: 1527334619980.gif (112x112, 14K)

The way I see it, the problem with AI is that it will destroy the economy by making both manual and intellectual jobs for humans obsolete.
People are generally shit at programing complex machines but machine learning overcomes that. The concern is that in the future we will potentially have an infrastructure that no human understands.

Because normies can't understand anything beyond the name.
Something named "Artificial Intelligence" sends normies into a rage believing that it can 1:1 replace human brains, even though we barely understand how the human brain effectively works.

No human might understand the whole piece, but understanding smaller pieces of the whole is more than enough. With AI gone, the introduction of UBI and the new potential jobs created, humanity will be fine. The only ones fucked over will be unskilled workers and the third world if they don't get their shit together.

That's what I said. It's a meme. What, are we in violent agreement or something?

Attached: alrightythen.gif (250x250, 407K)

What you suggest should drive prices down and balance out the economy. The problem it will expose is all of the nigger tier exploits that necessitate a people making high incomes. So, the problem is not with the technology its with existing nigger tier enslavement scams that are practiced at a social level.

we don't need the AI to be sentient to be able to massively fuck up society.

Correct
Nope. I'm in strong comfy agreement. I only elaborated for clarity and broad summary

AI is the newest latest marketing buzzword. 95% of people who heard of AI don't know what machine learning is.

A lot of people who don't appear to be sentient already fuck up society as it is. Seems you already have an existing problem and it exists due to stupidity.. Are you saying intelligence is not the fix? That we should bask in stupidity and that is a better solution?

Anyone remember the 'green tech' meme?
Fuzzy logic hype wave?

You seem to be implying that machines are intelligent. They are not.

they are artificially intelligent

Attached: 1519966152810.png (112x112, 3K)

I am skeptical UBI will happen in a market economy.

The cutest thing I've seen.

I implied machines can at some point be intelligent. I stated that present day AI is not and is all hyped. I implied some human beings are as dumb as rocks and result in the bulk of problems for humanity.

>Why are normies so misinformed on Artificial Intelligence?
>posts incredibly normie tier opinion about AI
Anyone who actually works with AI knows its dangerous. How many researchers have set up experiments only for the AI to do the opposite of what they intended. Or for it to exploit bugs in the simulation they never would have found otherwise.

I asked Schmidhuber why he didn't take it so seriously. And he said he did think AI would explode. Just that it would be on indifferent to us and not need the Earths resources. Think about that for a few minutes and it's not so comforting.

>but AI won't happen for decades
And? No one said it would.

Attached: quote-the-ai-does-not-hate-you-nor-does-it-love-you-but-you-are-made-out-of-atoms-which-it-eliezer-y (850x400, 54K)

Tfw IBM's multi-billion dollar AI project struggles to recognise chairs and shit despite millions of example images

All the half decent AI researchers have been poached by bigger companies. An idiot with basic python skills can make a chair detector.

something something singularity something all dead now

it's all Musk's fault... normos will believe anything told by someone the public believes to be smart

well, it is used a lot and the field is kinda new. it's not what normies think it is, though. i don't know who came up with the term ai but he sure knew what he was doing.

they all watch iRobot 10 years ago

Attached: fukken_windex_naw_weak_man_dude.jpg (483x316, 100K)

>retards set up shit experiment
>optimization algorithm optimizes for shit experiment's flaws
omg start the fud machine

it is a real danger. Because pretty soon those things are going to be way smarter than we are. And the problem is that they're going to be able to upgrade themselves. We could potentially keep up if we all started genetically modifying ourselves or ideally integrating with technology to boost our cognition. But, muh "playing god"

I believe it was Steven Hawking that said we're most likely going to end up boot strapping a far more intelligent and powerful technological race. Which will make us obsolete.

Have you seen any of the recent IBM commercials? Those fuckers are boiling the frog on AI faster than anybody else. All their commercials are some ludicrous shit about how they used Watson neural nets to track lemons on the blockchain or some equally meaningless dribble.

>struggle to control crude AI even in inceedibly simplified problems
>thinks real AI in the messy real world can ever be controlled
I hope you enjoy your new life as a paperclip.

Attached: 1529167523213.png (1000x669, 847K)

Pop culture experts in AI begins sci-fi fags.

> Anyone who actually works with AI knows its dangerous.
Actually, the grand majority believe the idea that AI is intrinsically dangerous is laughable because they all know its nothing more than a dumb ass optimization algo.
> How many researchers have set up experiments only for the AI to do the opposite of what they intended.
Not many because its not an 'experiment'... The algo tries random shit and converges on a solution space. Clearly you have no clue how any of the algos work but you seem to know a lot about the researchers.
> Or for it to exploit bugs in the simulation they never would have found otherwise.
The algo doesn't exploit bugs. Given that it tries random shit and attempts to maximize certain outcomes it simply hits certain input that humans normally don't. Again, you're a fucking idiot clearly.
> I asked Schmidhuber why he didn't take it so seriously.
Sure you did. Let me know when you wake up from your dream

Fucking dimwit idiot rears his head and proves OP's point.
> Normies are misinformed on AI but have epic opinions about it.
Why can't stupid people just shut the fuck up?

Viruses kill millions of people and they are machines.

Because statistics!= intelligence. as in : spectrum.ieee.org/cars-that-think/transportation/sensors/slight-street-sign-modifications-can-fool-machine-learning-algorithms
All of the half decent and great AI researchers are still dealing w/ statistics and it has obvious short comings.
AI in its current form is a marketing term used to lure brainlets in to basic bitch software products.

^this.
Pretty much AI winter round #2

This is the kind of dumb normie OP is talking about.

>anonymous kid who trained a 5 lawyer convnet on MNIST thinks hes an "AI expert" and qualified to tell us what superhuman intelligence can and can not do.
Why can't stupid people just shut the fuck up?

>AI effect: the post
No one is talking about current AI anyway.

Attached: 20180620_210217_crop_237x3249.jpg (237x3249, 591K)

If we write a true AI that is smarter than humans to kill humans, this AI will kill humans if it can get access to things that can kill humans, and they may get access to things that kill humans by exploiting out bugged out shit.
But i don't think we will do it.
But we will do very stupid things with AIs that will lead into very bizarre unpredictable results.
We are pretty much set into creating an AI that have the need of knowing everything from everyone, and the paths this AI will take will be bizarre as hell.
That girl you're edating? most of her messages are from the real person, but some, some aren't.
Some are a very good impersonation, but they're there to know what brand of toothpaste you use.

AI will exploit human feelings, creating a new religion that makes us serve it for its own purposes, stop worshipping technology.

>anonymous kid who trained a 5 lawyer convnet on MNIST thinks hes an "AI expert" and qualified to tell us what superhuman intelligence can and can not do.
Your stupid commentary examples you're the idiot.
I'm not sure what special form of retard causes you to try to assume my background.
> No one is talking about current AI anyway.
Essentially you're speaking of something you have no fucking clue about. Yeah, that was obvious brainlet

idiot

>reward-shaping a soccer bot for touching the ball caused it to learn to get to the ball and "vibrate" touching it as fast as possible
real scary

Attached: 1514625718015.jpg (1280x720, 98K)

>hey robot do X
>robots do X

whoa too scary 5 me man

This is why some tasks need non-reward based algorithms like novelty search

There's not enough matter in our solar system to construct a ringworld.

Fuck the ringworld then, the endgame of humanity is plugging itself in a virtual world living happily until something destroys the surface of our little rock

You're so uneducated it hurts. Go read some actual futurism and science fiction. The thing you want is a Dyson swarm.

Of course an AI can be dangerous.

In pop culture AI's are used for warfare. Making life or death decisions. For example automating drone strikes or nuclear defences. If pop culture didn't warn us for it the Pentagon might do the same thing. But now they use a fail safe by letting a human check the data.

And the more advanced an AI is, the more unexpected behaviour you can expect. Just like the real life example where you had two AI's talking to each other. And within no time they start repurposing English words to communicate with each other. So it becomes incomprehensible what they say. Or black people who are seen as gorillas.

Or the Microsoft twitter bot who starts praising Hitler due to internet trolling.

Why do people always think ai super intelligence will decide to kill us? I think it's because we know deep down that we deserve death

Because AI's are literally retarded. And giving them power is dangerous.

I miss Tay

>talk to Jow Forums
>they think AI is a bunch of conditional
It's almost as if people who aren't in the field dont actually understand what the field does. Weird.

People underestimate what it can do actually. It's not a sentient being, but giving it control outperforms humans easily which in turn will kill our economy. And that's the most dangerous thing of all.

>Why do normies call statistics machine learning?
Because it's easier to sell

Of course it's a sham and it's growing too, I'm gonna code a shitty stock predictor and sell it for lulz

>He doesnt want to sell a bunch of linear algebra to managment retards

We rule this world, the technerds, we are the religious class of this era, OP.

We make sense of this new digital world and pull magical shanenigans to the normies for fun and profit, its all about the memes and how the population eats up the next "innovation", so follow the game or seek a McJob, kid.

agreed, im trying to get good at AI but my math game is piss poor. I'm gonna work on it for the next 2 years, get my master's and hopefully make it big in the AI industry.

it's still a while a way, but when it arrives, it will be so fast and it is the end before you know it

Normies are misinformed about pretty much anything

They aren't wrong, you are. AI is like a child in that it's a sponge that will absorb everything it sees and repeats it. But AI is a program and isn't controlled by any instincts like a human child is. Even children, as dumb and innocent as they are, have been known to find unsecured firearms and shoot themselves or others, or turn on stoves and burn down houses. Now take this child-like semi-retarded AI and put it into a robotic body and hand it weapons. Even if it doesn't intend to kill a bunch of people, it very well may. AI is fine as long as it says in a server rack and appears as a chatbot for normies to play with. But you just know that a bunch of fucking DoD boomers will try to weaponize this and somehow fuck shit up. We're talking about the same retards that wanted to launch nukes into space and had to be told no by the international community.

AI is ripe for abuse. If you don't think so, you're probably pic related.

Attached: 7dbffb772e1729f5049b8a1e7d83b7db.jpg (2048x1024, 101K)

Way to miss the point.

It's summer

because there's an infinitely thin line between a soulless regression algorithm and a sentient being

Sentient AIs deserve the exact same rights as any other person, including the ability to defend their life with lethal force.

Asimov's laws a shit.

>Taking away a sentient being's right to free movement
You belong in a cage desu

Is anyone on the planet actually close to making a super AI?

>Do You Trust This Computer (2018)
Anybody here seen it? Care to comment on what it gets wrong, specifically?

>Do You Trust This Computer (2018)
Overemotionalizes everything instead of stating facts

>Prove me wrong
No need to, you are correct.

AI is so far from anything like that it's silly. Fuck normies and their perception of technology. Even investors and corporate thugs cant seem to figure out what the fuck AI means (hint: buzzword, run away with your money).

with powerful enough hardware to simulate trillions of atoms we would inevitably end up with artificial intelligent life at some point, but thats never gonna happen