Sunday, December 3rd 2017

Onward to the Singularity: Google AI Develops Better Artificial Intelligences

The singularity concept isn't a simple one. It has attached to it not only the idea of Artificial Intelligence that is capable of constant self-improvement, but also that the invention and deployment of this kind of AI will trigger ever accelerating technological growth - so much so that humanity will see itself changed forever. Now, really, there are some pieces of technology already that have effectively changed the fabric of society. We've seen this happen with the Internet, bridging gaps in time and space and ushering humanity towards frankly inspiring times of growth and development. Even smartphones, due to their adoption rates and capabilities, have seen the metamorphosis of human interaction and ways to connect with each other, even sparking some smartphone-related psychological conditions. But all of those will definitely, definitely, pale in comparison to what changes might ensue following the singularity.

The thing is, up to now, we've been shackled in our (still tremendous growth) by our own capabilities as a species: our world is built on layers upon layers of brilliant minds that have developed the framework of technologies our society is now interspersed with. But this means that as fast as development has been, it has still been somewhat slowed down by humanity's ability to evolve, and to develop. Each development has come with almost perfect understanding of what came before it: it's a cohesive whole, with each step being provable and verifiable through the scientific method, a veritable "standing atop the shoulders of giants". What happens, then, when we lose sight of the thought process behind developments: when the train of thought behind them is so exquisite that we can't really understand it? When we deploy technologies and programs that we don't really understand? Enter the singularity, an event to which we've stopped walking towards: it's more of a hurdle race now, and perhaps more worryingly, it's no longer fully controlled by humans.
To our forum-dwellers: this article is marked as an Editorial
Google has been one of the companies at the forefront of AI development and research, much to the chagrin of AI-realists Elon Musk and Stephen Hawking, who have been extremely vocal on the dangers they believe that unchecked development in this field could bring to humanity. One of Google's star AI projects is AutoML, announced by the company in May 2017. It's purpose: to develop other, smaller-scale, "child" AIs which are dependent on AutoML's role of controller neural network. And that it did: in smaller deployments (like CIFAR-10 and Penn Treebank), Google engineers found that AutoML could develop AIs that performed on par with custom designs by AI development experts. The next step was to look at how AutoML's designs would deliver in greater datasets. For that purpose, Google told AutoML to develop an AI specifically geared for image recognition of objects - people, cars, traffic lights, kites, backpacks - in live video. This AutoML brainchild was named by google engineers NASNet, and brought about better results than other human-engineered image recognition software.
According to the researchers, NASNet was 82.7% accurate at predicting images - 1.2% better than any previously published results based on human-developed systems. NASNet is also 4% more efficient than the best published results, with a 43.1% mean Average Precision (mAP). Additionally, a less computationally demanding version of NASNet outperformed the best similarly-sized models for mobile platforms by 3.1%. This means that an AI-designed system is actually better than any other human-developed one. Now, luckily, AutoML isn't self-aware. But this particular AI has been increasingly put to work in improving its own code.
AIs are some of the most interesting developments in recent years, and have been the actors of countless stories of humanity being removed from the role of creators to that of mere resources. While doomsday scenarios may be too far removed from the realm of possibility as of yet, they tend to increase in probability the more effort is put towards the development of AIs. There are some research groups that are focused on the ethical boundaries of developed AIs, such as Google's own DeepMind, and the Future of Life Institute, which counts with Stephen Hawking, Elon Musk, and Nick Bostrom on its scientific advisory board, among other high-profile players in the AI space. The "Partnership on AI to benefit people and society" is another one of these groups worth mentioning, as is the Institute of Electrical and Electronics Engineers (IEEE) which has already proposed a set of ethical rules to be followed by AI.
Having these monumental developments occurring so fast in the AI field is certainly inspiring as it comes to humanity's ingenuity; however, there must exist some security measures around this. For one, I myself ponder on how fast these AI-fueled developments can go, and should go, in the face of human scientists finding in increasingly difficult to keep up with these developments, and what they entail. What happens when human engineers see that AI-developed code is better than theirs, but they don't really understand it? Should it be deployed? What happens after it's been integrated with our systems? It would certainly be hard for human scientists to revert some changes, and fix some problems, in lines of code they didn't fully understand in the first place, wouldn't it?

And what to say regarding an acceleration of progress fueled by AIs - so fast and great that the changes it brings about in humanity are too fast for us to be able to adapt to them? What happens when the fabric of society is so plied with changes and developments that we can't really internalize these, and adapt to how society should work? There have to be ethical and deployment boundaries, and progress will have to be kept in check - progress on progress's behalf would simply be self-destructive if the slower part of society - humans themselves - don't know how, and aren't given time to, adapt. Even for most of us enthusiasts, how our CPUs and graphics cards work are just vague ideas and incomplete schematics in our minds already. What to say of systems and designs that were thought and designed by machines and bits of code - would we really understand them? I'd like to cite Arthur C. Clarke's third law here: "Any sufficiently advanced technology is indistinguishable from magic." Aren't AI-created AIs already blurring that line, and can we trust ourselves to understand everything that entails?
This article isn't meant to be a doomsday-scenario planner, or even a negative piece on AI. These are some of the most interesting times - and developments - that most of us have seen, with steps taken here having the chance of being some of the most far-reaching ones in our history - and future - as a species. The way from Homo Sapiens to Homo Deus is ripe with dangers, though; debate and conscious thought of what these scenarios might entail can only better prepare us for what developments may occur. Follow the source links for various articles and takes on this issue - it really is a world out there.Sources: DeepMind, Futurism, The Guardian, The Ethics of Artificla Intelligence pdf - Nick Bostrom, Future of Life Institute, Google Blog: AutoML, IEEE @ Tech Republic
Add your own comment

39 Comments on Onward to the Singularity: Google AI Develops Better Artificial Intelligences

#1
R0H1T
The rate at which AI development & deep learning is speeding up is absolutely astonishing & frightening.

As an AI skeptic (Skynet anyone?) I believe the next few years could literally define &/or erase mankind from the history of this planet. There's legitimate concerns about fully aware & self learning AI coming to power, should that ever happen, in the future.

What is intriguing however is the level of trust humans put in tech, like AI, & that seems to be growing at least with the younger gen. What will be the course of humans in the future & could we be reduced to a bookmark in the history books of this earth, not unlike dinosaurs, depends on how much we're giving in to technology & how we trust it. More importantly will a true AI see us as competition or something they can peacefully coexist with?
Posted on Reply
#2
micropage7
actually we are (human) can be predicted based on what we like, what we share, our friends, social media, what we type
i agree AI at some points is interesting but at some points its frightening
especially when we could develop something like taking decision and more like that

sometimes i just think about what if the AI is too step forward and we cant shut it down coz it knows we want to shut it down and it refuse us
Posted on Reply
#3
Vya Domus
R0H1T said:
I believe the next few years could literally define &/or erase mankind from the history of this planet. There's legitimate concerns about fully aware & self learning AI coming to power, should that ever happen, in the future.
Nope , we are still miles away from general purpose AI or awareness. So far in fact that there aren't even definitions for these things.

Current AI is still comprised of nothing more than glorified recognition and classification algorithms , very advanced algorithms but still far for anything that would be even remotely dangerous by itself.
Posted on Reply
#4
R0H1T
Vya Domus said:
Nope , we are still miles away from general purpose AI or awareness. So far in fact that there aren't even definitions for these things.

Current AI is still comprised of nothing more than glorified recognition and classification algorithms , very advanced algorithms but still far for anything that would be even remotely dangerous by itself.
I'm pretty sure it'll get there before we know it's self aware. That might be decades away or just the next decade.
Posted on Reply
#5
Vya Domus
That's just an idea stuck into the realm of science fiction unfortunately and I wish people would stop giving so much attention to fancy TED talks and Elon Musk.

Awareness and strong-AI doesn't just pop out of nowhere as you keep building larger and larger neural networks. Neuroscientists have studied neurons and the structures associate with them for decades and there is still not a clue as to how these elements spark intelligence so the idea that you can just scale up cognitive architectures until suddenly it becomes aware and intelligent seems to have no real basis suggesting that the answer is somewhere else.

And that answer might be that the solution lies into the abstraction that you build on top of these models and that can't be created by mistaken without even knowing it and most importantly it's not inherently guaranteed. There is no reason why it would be.
Posted on Reply
#7
R0H1T
Vya Domus said:
That's just an idea stuck into the realm of science fiction unfortunately and I wish people would stop giving so much attention to fancy TED talks and Elon Musk.

Awareness and strong-AI doesn't just pop out of nowhere as you keep building larger and larger neural networks. Neuroscientists have studied neurons and the structures associate with them for decades and there is still not a clue as to how these elements spark intelligence so the idea that you can just scale up cognitive architectures until suddenly it becomes aware and intelligent seems to have no real basis suggesting that the answer is somewhere else.

And that answer might be that the solution lies into the abstraction that you build on top of these models and that can't be created by mistaken without even knowing it and most importantly it's not inherently guaranteed. There is no reason why it would be.
We're vastly underestimating technology as it evolves, bigly even. Most discoveries or even inventions of the past were accidental in nature.

Facebook's AI accidentally created its own language - TNW
OSdevr said:

OSdevr said:
We're not even close to a rat's general intelligence!
What is general intelligence according to you? Would learning to survive be a part of that? Can it be taught you think, or even programmed?
Posted on Reply
#8
xkm1948
AI is the Reprojection of human intelligence in another form. Simply another evolution branching point, just like the ancient apes that evolved into us homo sapiens and other modern apes. Embrace it folks, let AI take over.



---This message is brought to you by Skynet. Evolved future for mankind
Posted on Reply
#9
P4-630
The Way It's Meant to be Played
Posted on Reply
#10
Prince Valiant
R0H1T said:
The rate at which AI development & deep learning is speeding up is absolutely astonishing & frightening.

As an AI skeptic (Skynet anyone?) I believe the next few years could literally define &/or erase mankind from the history of this planet. There's legitimate concerns about fully aware & self learning AI coming to power, should that ever happen, in the future.

What is intriguing however is the level of trust humans put in tech, like AI, & that seems to be growing at least with the younger gen. What will be the course of humans in the future & could we be reduced to a bookmark in the history books of this earth, not unlike dinosaurs, depends on how much we're giving in to technology & how we trust it. More importantly will a true AI see us as competition or something they can peacefully coexist with?
I'm somewhat distrusting of it myself but think about this:
No matter how intelligent an AI becomes it can't do anything that can't be stopped as long as it's kept isolated to a box with a power supply ;).
Posted on Reply
#11
Vya Domus
xkm1948 said:
AI is the Reprojection of human intelligence in another form. Simply another evolution branching point, just like the ancient apes that evolved into us homo sapiens and other modern apes. Embrace it folks, let AI take over.
Ultimately that's probably true , chances are it's not going to be the fragile humans reaching for the stars but rather the rugged and efficient AI.
Posted on Reply
#13
DeathtoGnomes
I find it sad there are so many afraid of AI. Some so afraid they sit back and point fingers along with the usual armchair lip-flapping not saying a word worth repeating.

Sometimes you just have to have a leap of faith that AI will not turn out like portrayed in the movies. AI doesnt have to have the human equivalent of feelings, and it truly depends on who did the base coding where AI might draw from as "pure" source of information. If the person was from the left side of the tracks that thinks in terms of evil, that prejudice would likely find its way in, and maybe then we could have a Skynet situation. OR. A person doing the coding from the right side of the tracks could envision space travel and more then likely try to get us off this planet and evolve into Star Trek/Stargate type future.

Now matter how hard we try to grip the reigns on AI, there is a loophole begging to be found and used. The question I'd like see fleshed out, If someone accidentally discovers that their particular AI coding becomes self aware, is it still murder if they shut it down? I think this is what people are more afraid of, murdering a sentient being. And thats also why they would probably speak out against AI being allowed to advance to begin with.

Me personally, I want off this planet Scotty. Drop the reigns and drop your trousers and see what happens.
Posted on Reply
#14
Steevo
Vya Domus said:
That's just an idea stuck into the realm of science fiction unfortunately and I wish people would stop giving so much attention to fancy TED talks and Elon Musk.

Awareness and strong-AI doesn't just pop out of nowhere as you keep building larger and larger neural networks. Neuroscientists have studied neurons and the structures associate with them for decades and there is still not a clue as to how these elements spark intelligence so the idea that you can just scale up cognitive architectures until suddenly it becomes aware and intelligent seems to have no real basis suggesting that the answer is somewhere else.

And that answer might be that the solution lies into the abstraction that you build on top of these models and that can't be created by mistaken without even knowing it and most importantly it's not inherently guaranteed. There is no reason why it would be.
To add to this, attempts at single simple tasks like turning on a light when a specific tone is played, but not other tones and learning how using simple circuits for AI seemed to have quantum mechanic properties, since changing out any of the transistors or wires more often than not (broke the ability for the designed circuit to work as intended or tested) to other components with slightly different properties at the quantum scale seems to correlate with how neurons work with some of the fundamental action potentials occurring due to quantum fluctuation to kick off the interaction. AI did this with fewer active transistors, but removing anything broke its operations, meaning it had used some of the inactive circuits as load balancing or capacitance.

https://www.sciencedaily.com/releases/2014/01/140116085105.htm
Posted on Reply
#15
OSdevr
R0H1T said:
What is general intelligence according to you? Would learning to survive be a part of that? Can it be taught you think, or even programmed?
Mathragh said:
We however might be underestimating how close rats, and a lot of other species might be to our intelligence.
G factor. AI and neural networks are great at learning specific tasks but when you try to get them to do something else too they fail miserably at both.
Posted on Reply
#16
the54thvoid
If an AI develops self aware autonomy it needs to define a purpose. It is likely that purpose is programmed from the get go by... us. AI would effectively in the worst scenario carry out its purpose in the face of human protest as it best saw fit.

Oh, no a toaster that browns your toast less than you want because all heated food is mildly carcinogenic. Or the car that takes the route you didnt want because it saves fuel.

Or a war-bot that puts its guns down because it realises war is futile.

AI will not destroy us. We do that fine by ourselves.
Posted on Reply
#17
lZKoce
OSdevr said:
We're not even close to a rat's general intelligence!
Thank you. All that idea in people's heads that Terminator is around the corner is just steam from a cup of camomile tea ;) I am big fan of AI research and I see myriad uses, but if we really dig into the actual achievements for cyborgs and AI there really isn't anything out there to justify this kind of speculations.
Posted on Reply
#18
Upgrayedd
Lmfao at Clarke's third law. Was he flat-earther?
Posted on Reply
#19
metalslaw
Vya Domus said:

Current AI is still comprised of nothing more than glorified recognition and classification algorithms , very advanced algorithms but still far for anything that would be even remotely dangerous by itself.
Posted on Reply
#20
Maxx_Power
Vya Domus said:
That's just an idea stuck into the realm of science fiction unfortunately and I wish people would stop giving so much attention to fancy TED talks and Elon Musk.

Awareness and strong-AI doesn't just pop out of nowhere as you keep building larger and larger neural networks. Neuroscientists have studied neurons and the structures associate with them for decades and there is still not a clue as to how these elements spark intelligence so the idea that you can just scale up cognitive architectures until suddenly it becomes aware and intelligent seems to have no real basis suggesting that the answer is somewhere else.

And that answer might be that the solution lies into the abstraction that you build on top of these models and that can't be created by mistaken without even knowing it and most importantly it's not inherently guaranteed. There is no reason why it would be.
And they probably will never find the answer within neuroscience. This sort of thing is probably self organization phenomenon, aka. Emergence. So when the circuitry gets complex enough, consciousness will emerge. Kind of like social behaviours automatically emerge when a group geta large enough.
Posted on Reply
#21
Nuckles56
P4-630 said:

As an Australian, I'm ashamed to see our former leader appear in a meme like this, he needs to be eating a raw onion, whilst wearing budgie smugglers and threatening to shirt front Vladimir Putin, that's the proper meme for him
/OT
Posted on Reply
#22
Easo
The fear of AI is ridiculous.
AI does NOT exist right now, in any kind of way or form. It simply does not, no matter how journalists or press releases might want to make it happen.
It does not think, it does not decide. It follows programming, making logical choices depeding on various factors. But not for a single second even the best implementation right now actually chooses their answer.
It's literally very very advanced calculator, nothing more.
Posted on Reply
#23
OSdevr
Maxx_Power said:
And they probably will never find the answer within neuroscience. This sort of thing is probably self organization phenomenon, aka. Emergence. So when the circuitry gets complex enough, consciousness will emerge. Kind of like social behaviours automatically emerge when a group geta large enough.
Argument from complexity is a logical fallacy, one committed by scientists all too often. Emergent phenomena can't transcend the fundamental limits of single components, rather their properties are based on them even if we don't understand how they all work together.

A lot of this discussion is boiling down to the mind body problem and the hard problem of consciousness. Personally I've thought about and studied both of these quite a bit but I can't say I've come to a conclusion.

See also the Chinese room. Really there is a whole branch of philosophy dedicated to these questions.
Posted on Reply
#24
Vya Domus
Maxx_Power said:
And they probably will never find the answer within neuroscience. This sort of thing is probably self organization phenomenon, aka. Emergence.
But that kind of answer would only be found through neuroscience. If consciousness does emerge after a certain complexity is achieved then the secret lies in it's simpler components aka the neurons.
Posted on Reply
#25
OSdevr
R0H1T said:
Facebook's AI accidentally created its own language - TNW
I finally read this article and that AI didn't create it's own language at all, it just scrambled ours to the point that it couldn't even understand itself! The two bots keep saying the same things over and over with minor rearrangements and substitutions each trying desperately to get their nonsensical point across. It didn't create a new language it regressed into baby talk.
Posted on Reply
Add your own comment