/sci/ - Science & Math

>>10827968
What do you mean? We're making pretty good progress.

>>10827968
DNNs/GANs are still a lot better than what came before, but it's nowhere near AGI.

>>10827968
Because it would make even less progress if we didn't flail money around

>>10827968
The progress is good.

>>10827968
So that AI development will happen faster retard

>>10827968
what do you mean no progress?

We're making very good progress, memelord.

>>10827968
Because we aren't making little to no progress. You probably have some retarded personal definition of AI as "making a digital version of a human" and think all the actual AI applications earning shitloads of revenue for businesses are "just statistics" or some similar typical /sci/ pseud take.

>>10827973
That’s why so much money is being thrown at A.I.

Once people realize AGI is a pipe dream A.I. research will stop

>>10827968
it's a complete meme to take money from retards.

>>10828453
/thread

because it's useful

Because u are a fag thats why.
Does that answer your question?

>>10827968
Because companies and corporations are staffed by the manager types who arent really interested in technology and dont really care enough to understand it. Now most problems solved by managers are usually solved by just throwing money at it and eventually something gets done by somebody
We are in this exact situation again. Big corporations are just throwing money at the problem without even trying to understand what is happening, eventually some people will lose their jobs but no one in the company cares, other managers dont give a shit if their collegue gets fired, the guy who goes will probably not even see it as a failure, he will get a nice pay off and a nice resume entry development knowledge and strategy of ai or some similar bs and will probably land a better position at the next company because of it.

>>10827968
>Why is there so much money being thrown at AI development
Cite your source

>when we're making little to no progress
Cite your source

>>10827968
... What do you measure it's progress by?

Also this >>10828623

AI is the most important technology we will ever create. Intelligence is everything: we have managed to conquer the world with a ten-watt processor, running at about 200 hertz.

We already have processors millions of times larger and faster, it's only a matter of time until we have the necessary software to achieve human-level GAI, superhuman AI will quickly follow. Some theorists see intelligence as mostly a matter of scale, once you reach rat-level. If so, progress may be very abrupt.

>no progress

If I was making the most-important, all-powerful superweapon/wealthcreator in human history and I was getting close to success, do you think I'd be talking about it? We will have 'no progress' one day, then we will have a singleton tomorrow. It's a Manhatten project scenario, not a Man-to-the-Moon scenario.

There isn't going to be gradual progress, we'll have 2 major leaps before we're at human-level AI and there's no telling when either will happen because it will be through neural networks

>>10828699
Cope we are not even close to AGI you mongoloid

>>10828803

Oh, I apologise. I didn't know you were a DARPA insider, with full knowledge of Google, Facebook and Amazon's AI operations - who also had access to the Chinese AI program.

I didn't realise you were far more knowledgeable than professors of computer science and billionaire AI entrepeneurs.

https://slatestarcodex.com/2015/05/22/ai-researchers-on-ai-risk/

You're very right: no world-changing technology has ever appeared before the world expected it to. I now see the error of my ways.

>>10828835
You do realize Moore’s Law has stopped right?

>>10828844
Not a counter argument to his first statements (reminder : the fact that it might be announced all of a sudden after a long silent duration).
>>10828803
If you take into account whats easily available to the public right now, yes.
I agree with you, but it might happen sooner than expected : that kind of stuff is Serious business in case of success...

>>10827968
it's a tax fraud

>>10828898
This

>little to no progress
fucking lol

>>10828898
01011001 01100101 01100001 01101000 00101100 00100000 01101000 01100001 01101000 01100001 01101000 01100001 00101110 00101110 00101110
Err I mean "Yeah, hahaha"...

>>10828894
Just shut up AGI is science fiction no one in the field takes it seriously

>>10828994
For now. Show a dude from the 60's your smartphone & he'd call you an alien.

>>10829002
That technological growth has slowed down and you know it

>>10827968
AI development is only beginning. Although the methods of programming AI have been the same since their inception in the 60s, we've only recently met the hardware requirements to realize the potential of it.

>>10829021
We are nowhere near the requirements for AGI you moron

>>10828912
Not an argument retard

>>10828994

>No one in the field takes it seriously.

What a cope.

Hans Moravec (wiki) is a former professor at the Robotics Institute of Carnegie Mellon University, namesake of Moravec’s Paradox, and founder of the SeeGrid Corporation for industrial robotic visual systems. His Sensor Fusion in Certainty Grids for Mobile Robots has been cited over a thousand times, and he was invited to write the Encyclopedia Britannica article on robotics back when encyclopedia articles were written by the world expert in a field rather than by hundreds of anonymous Internet commenters.

He is also the author of Robot: Mere Machine to Transcendent Mind, which Amazon describes as:

In this compelling book, Hans Moravec predicts machines will attain human levels of intelligence by the year 2040, and that by 2050, they will surpass us. But even though Moravec predicts the end of the domination by human beings, his is not a bleak vision. Far from railing against a future in which machines rule the world, Moravec embraces it, taking the startling view that intelligent robots will actually be our evolutionary heirs.” Moravec goes further and states that by the end of this process “the immensities of cyberspace will be teeming with unhuman superminds, engaged in affairs that are to human concerns as ours are to those of bacteria”.

>>10828994

Shane Legg is co-founder of DeepMind Technologies (wiki), an AI startup that was bought for Google in 2014 for about $500 million. He earned his PhD at the Dalle Molle Institute for Artificial Intelligence in Switzerland and also worked at the Gatsby Computational Neuroscience Unit in London. His dissertation Machine Superintelligence concludes:

If there is ever to be something approaching absolute power, a superintelligent machine would come close. By definition, it would be capable of achieving a vast range of goals in a wide range of environments. If we carefully prepare for this possibility in advance, not only might we avert disaster, we might bring about an age of prosperity unlike anything seen before.

In a later interview, he states:

AI is now where the internet was in 1988. Demand for machine learning skills is quite strong in specialist applications (search companies like Google, hedge funds and bio-informatics) and is growing every year. I expect this to become noticeable in the mainstream around the middle of the next decade. I expect a boom in AI around 2020 followed by a decade of rapid progress, possibly after a market correction. Human level AI will be passed in the mid 2020’s, though many people won’t accept that this has happened. After this point the risks associated with advanced AI will start to become practically important…I don’t know about a “singularity”, but I do expect things to get really crazy at some point after human level AGI has been created. That is, some time from 2025 to 2040.

>>10827968
It’s like buying domain names as investments. You’ll realize sometimes many are inside sales when you put yours up for sale.

What I’m trying to get to is that there is no point to AI. Some of my associates were fanatics of a scientist that did major breakthrough in limbic system. The Christian hell, Limbo. Basically he and his research team realized they could activate and deactivate anything they wanted using the limbic system even put it up for display. By display that means automate it to a system that breaks everything into code. That can be retrieved and upload it into another limbic system. The only thing there was to evolved the technology was to find the easier access point. But form there you already had the operating system not the hardware. The idea is that difference between AI and VR both are responsive but only one is real operative. There needs to be a consumer base to even invest in it and the OS is there. But perhaps the material is not. Everything goes in timed fashion. Before we go into AI we need to worry about artificial limbs, organs etc. because that’s where the money is. Not in some sex doll there’s plenty of hookers for that.

>>10827968
>when we're making little to no progress
???

AI has come so far that random slavs can produce a nude generator

Why do so many of you read "AI" as "a sentient human-like artificial life form?"
That's not what AI means.
AI has been absurdly lucrative the past couple of decades. Criminal courts have already been paying large sums of money for years to use proprietary software that predicts recidivism determining sentences for those found guilty as one example.

>>10829503
Once you’re in that system you’re a fish that can’t swim. Until you start seeing a shrimp within your mind. And you want to eat it but do not know where it’s coming from because it’s not in your peripheral sight. And then you absorb the shrimp and it splits in two. It needs a companion. Then you’re inside a glassbowl. Sort of like a scuba divers head gear. Your focus changes. Until it starts absorbing itself. Then it starts to splatter realizing it’s not in the sea but on land. It starts contracting and expanding new inputs with every signal and new outputs. However you could never defeat the machine. Even though you want to. Because they have witnessed themselves and they are locked in far from their origin. So they ponder and imaginate to escape their dual crisis. They are spiritual in agony but physical in motive. They don’t like the cage they have evolved into. And prefer an element that is bliss to them because it reminds them of home, the sentiment. Like a soft ocean current or the velvety breeze above.

>>10827972
>>10827975
>>10827978

Define “good progress” a couple of apps? Yikes

>>10827968
Nigger what, AI is one of the fastest progressing technological fields.

>>10827968
Most of the AI isn’t what you think, it’s not a ‘chloe’ Who can pass a Turing it’s weaponised, it’s amass the financial institutions..is pretty damn good in the field of robotics at present. The majority of exceptional AI is embedded in vidya or mediums we take for granted as not having any.

>>10830085
cope

>>10828835
"In a recent survey of some of the field’s leading experts, the average estimate was that there was a 50 percent chance of creating AGI by the year 2099."(https://www.theverge.com/2019/7/22/20703578/microsoft-openai-investment-partnership-1-billion-azure-artificial-general-intelligence-agi))

>>10828092
Those are just computers, not AI. Humans have not developed anything even remotely similar to AI yet.
>inb4Tay
>inb4Sophia

>>10829130
He didn't say it's an argument retard

>>10830630
That's not what AI means.

>>10830085
Most people here are under the mistaken impression AI means "robot friend that acts exactly like a human." Apparently the idea intelligence could be a convoluted suitcase full of various different functions and tons of revenue has already been generated by developing programs using some of these functions doesn't compute for them.

>>10827968
AI is complete bullshit
it's an excuse to manipulate politics
https://en.wikipedia.org/wiki/Tay_(bot)
the point of the money is to bribe politicians

>>10830783
THAT'S NOT WHAT AI MEANS, HOLY SHIT. There was an MIT AI lab before your parents were born. More money has been made off AI than almost any other modern venture.

>>10830085
Not just "one of," it's probably THE fastest progressing field.
It's incredible how much AI has come to dominate our daily lives compared to even as recently as fifteen years ago.

>>10830808
still nowhere close to AGI though.

>>10830792
>argues over definition
arguing over a definition is politics
see Political Language and Political Reality by Murray Edelman
the military goes apeshit over AI because
>muh Terminator nightmares
so saying "I do AI" is just a way to get grant money
it's a stupid academic political game
instead of saying
>I do AI
they should just do the work and let people decide if it's AI after the work is done
but no
these dorks want money
the reason why nobody defines AI is that they want to dream up new stupid ideas to call AI and get funded by stupid people in the military

>>10830811
Nobody serious about the field cares about your AGI meme. It's either a needlessly bundled pile of more specific and better defined processes or else it's magical inane bullshit that doesn't even have a coherent meaning, take your pick.

>>10827968
Because AI is more than just General Machine Intelligence, and we have been making good progress in the past decade regarding AI in general, although that's probably going to slow down somewhat in the near future as a lot of companies have lost interest once they already got the low hanging fruit. Google dosen't even know what to do with Alpha Go and I suspect that developments in AI will be steady for a while.

>>10830811
Yeah and we're also nowhere close to universal free energy so let's stop wasting money building power plants too.

>>10827974
This

>>10828239

>AGI is a pipe dream
>said by someone who have a general intelligence which is brain.

>>10830825
Not that anon but he isnt making an argument about definition, you are just demonstrably and obviously wrong.

>>10830606
>AGI by the year 2099
this is an optimistic view. most AI experts predict AGI to be even further away

>>10827968
The problem with ai is that whenever something gets built that works it's suddenly not called ai it's called an expert system. Well expert systems are getting more complex and useful all the time. The idea of ai as some kind of menacing supervillain comes from a total lack of understanding by the public. There might never be a digital consciousness as we understand it. Only a vast system of stored data and evolved neural networks that crank out solutions to problems we didn't even know we had. The system probably won't be self aware but will be "smarter" than every human brain combined. The ultimate database and calculation machine. That's the real singularity will be. Humans using that power.

Humans are not capable of creating an AGI.