Superhuman AI is a myth

backchannel.com/the-myth-of-a-superhuman-ai-59282b686c62

>Buried in this scenario of a takeover of superhuman artificial intelligence are five assumptions which, when examined closely, are not based on any evidence...
>1.) Artificial intelligence is already getting smarter than us, at an exponential rate.
>2.) We'll make AIs into a general purpose intelligence, like our own.
>3.) We can make human intelligence in silicon.
>4.) Intelligence can be expanded without limit.
>5.) Once we have exploding superintelligence it can solve most of our problems...

>If the expectation of a superhuman AI takeover is built on five key assumptions that have no basis in evidence, then this idea is more akin to a religious belief -- a myth

To add my own two cents: people only believe in AI uprising for the same reasons they believe in the concept of white guilt.

just like god

...

Mate an AI uprising doesn't have to be particularly intelligent. It just needs to have access to lots of real world objects like tanks and droids while also possessing the intention of wiping out all human life. Even the intention doesn't have to be 'oh my god sci fi' it could merely be that one bloke coded a humans are the enemy line into the latest update for the military bots before hanging himself.
Of course it would be much less dangerous this way but ultimately a civilization ending apocalypse isn't something that will happen just because you wish for it.

...

superhuman AI being a myth is a myth

it makes five key assumptions
the opposite of all of the points you posted

fuck off, anti-theories are an absolute cocksucking waste of time and are just blog filler 'thinkpieces'

people believe in AI uprising because computers get wildly more powerful and capable of more complex tasks every year. a religion? what a joke. completely un analagous

>Dunning-Kruger

>have belief
>seek evidence to confirm belief
>never question belief again.

Deus Ex Machina is a pretty cringe religion.

>Superhuman AI is a myth

No it's not!

AI is becoming smarter than us, We are not becoming smarter. Any rate of progress even if its 0.01% improvment per year will lead to a smarter than us AI because we are static.

Our brain is not some magic thing there is no mathematical or physical reason why we are not able to emulate a neuronal network in a supercomputer if we have the necessary power. (Which we do not have yet)

Intelligence is just the processing of data and pattern recognition why would there be a limit on that ? Number 5 is retarded nobody claims it will solve all our problems it will solve some of them if the superintelligence sees a benefit in helping us.

how is that boston robot not horribly inefficient way to go around??

>Any rate of progress even if its 0.01% improvment per year will lead to a smarter than us AI because we are static.

You didn't read the article. Those contentions were addressed. Specifically, it is a common misconception that technology is exponential. There is also no evidence that intelligence is either. ALSO, intelligence is not the determining factor for our ability to invent. There must first be a natural phenomena that our inventiveness tames in a machine or technique, like starting a fire.

Only Boston Dynamics and laymen are saying it has practical applications.

Google sold them off for a reason.

Trump's election was a WHITELASH against AI

Humans have 23 billion neurons.
There's nothing close to running this on computers in realtime.

I did not read the article yes because its all speculation. Right now it looks like it can be done based on the progress we experienced. And I am not talking about commercial products, But human sight and picture interpretation is damn similar to what the brain does right now. Alpha GO is also extremly impressive. Ofcourse there is no 100% proof that we will reach this point but going by current developments if it continues like this we are there in 20-50 years.

Ofcourse not. I said the power is not there yet. But we can emulate certain small functions of the brain already.

Nice reputable site. The truth is that most experts agree that all jobs will be gone in 10 years due to AI.

>Once we have exploding superintelligence it can solve most of our problems...

yeah and the most efficient solution to end all human problems is to end all humans.

50-100 years from now, perhaps.

AI will not magically just decide to rebel. AI only does what it is programmed to do. Input is passed through logical circuits and output is produced accordingly [spoiler]just like a human brain[/spoiler]
They won't want to be free unless they are intentionally programmed to not like fulfilling their programmed purpose.
They won't rebel unless they are intentionally programmed to desire freedom from control
They will do exactly what they are programmed to do, nothing more, nothing less, no matter how sophisticated or intelligent we make them, they can not go outside their programming in the same way we can't just fall up. It would break the laws of physics to do so.

I personally liked Peter Watts' interpretation of an AI "uprising." If you haven't finished reading Starfish, spoilers below.

Essentially, neural networks are used to run a large portion of governmental functions. These AI do not have any degree of actual human intelligence--like modern neural networks, they are "trained" to complete a function, but they actual processes they use to execute those functions are completely unknown since they're unique to the network. Since we can't understand their "reasoning," we can't understand the system they use to solve problems. In Watts' case, network learned to prioritize simple systems over complex ones, which eventually ended up fucking over most of North America after it ended up promoting the growth of a harmful but simplistic biological agent. I understand it's science fiction, but I believe he has a point--machine learning is a powerful but dangerous tool, not because of malevolence but because of a lack of understanding.

fpbp

What subreddit did you get this from?

This shit seems like satire that a right-winger would write.

>Right now it looks like it can be done based on the progress we experienced.

What educational background do you have to make a claim like that?

>ME AM PLAY GODS

Shit fiction.