886 replies to this topic

[mag1]

um, did I read this correctly?

After the startling defeat of a world master Go player, the South Korean public has become greatly worried about the potential of this technology and the possibility that

they have fallen behind in AI research. In response their government is committing a substantial amount of resources to move such research forward.

 

South Korea sees AI as an opportunity?

This entire thread has focused on the threat of the next generation of technology.

 

Why is it not obvious to the South Koreans that the winner of the Go contest was not a South Korean? (In fact, it was a machine)

How can they possibly think that they will win this by making the technology invention machine go faster?

The Western impulse has typically been to smash the machines.

 

 

 

[mag1]

Want to wish everyone a Happy Earth Hour.

 

Suppose we could also think of it as a Machine Triumph Delay hour. 

If we could all live without electricity, running water ... we would not be nurturing the Singularity machine. 

[niner]

South Korea sees AI as an opportunity?

This entire thread has focused on the threat of the next generation of technology.

 

Why is it not obvious to the South Koreans that the winner of the Go contest was not a South Korean? (In fact, it was a machine)

How can they possibly think that they will win this by making the technology invention machine go faster?

The Western impulse has typically been to smash the machines.

 

Well yeah.  If the technological/sociological landscape is about to be re-made, it would be better to have some control over the change rather than be nothing but a victim.   Anyone who tries to smash the machine will consign themselves or their nation to victim status.   I suppose it might be possible to create a society within a society, like the Amish, that lives without use of technology.   In the long run AI is going to be better for everyone, but in the short term we are going to be facing disruptions the likes of which humans have never seen. 

[Lewis Carroll]

um, did I read this correctly?

After the startling defeat of a world master Go player, the South Korean public has become greatly worried about the potential of this technology and the possibility that

they have fallen behind in AI research. In response their government is committing a substantial amount of resources to move such research forward.

 

South Korea sees AI as an opportunity?

This entire thread has focused on the threat of the next generation of technology.

 

Why is it not obvious to the South Koreans that the winner of the Go contest was not a South Korean? (In fact, it was a machine)

How can they possibly think that they will win this by making the technology invention machine go faster?

The Western impulse has typically been to smash the machines.

 

A very negative outlook you seem to have...

 

I don't necessarily see this thread as being focused on the "threat of next generation technology". In my opinion, the coming automation is a great thing; something all humanity should be actively working towards. Obviously this paradigm shift will be accompanied by some degree of societal/economical tumult, but I remain quite optimistic and eager. This thread has focused on the future possibilities of the next generation of technology. 

[mag1]

The South Korean response is counter-intuitive to me.

However, fair enough, the more theoretical and longer range AI research offers significant promise and opportunity.

 

The more immediate and scarier lower level technologies could have much more significant and disruptive effects.

For example, the robocart that has been noted previously on this thread would have profound impact on our labor economy.

There are now several imminent highly disruptive technologies that have been put in the stack and are just circling waiting to land.

 

We might comfort ourselves by looking at communities that have deliberately embraced a less technological lifestyle, though in many

ways modern nations have also restricted the choices that are available. There is now a long list of industries on the brink of

profound change: education, medicine, transport, retail ... .

 

What happens when building human capital is no longer considered a rational choice? 

The most highly compensated members of our society invested years, often decades, in developing an expert level skill set in a discipline.

What will human society be like when people plan their future in much the same that they mindlessly surf from what TV show to another?

 

 

The title for this thread is: Employment crisis: Robots, AI, & automation will take most human jobs

 

The last video portrayed a world in which the last employee was fired.

A previous video showed a robot that appeared highly capable of performing a wide range of tasks normally done by human workers.

 

Edited by mag1, 20 March 2016 - 01:54 AM.

[mag1]

Almost hundred years after the collapse of the rural manual economy, the elite now can see the rapidly approaching erosion of

their power base. The next layer of our society is becoming increasingly vulnerable to displacement. Even with their

entrenched power, it is not obvious that they will be able to maintain their social position.

 

For example, at the recent Washington CRISPR International meeting, it was decided that germ line engineering should be prohibited.

Such genetic engineering would obviously redefine medicine. In a genetically engineered world it would be expected that

there would be almost no illness.

 

Our family has struggled with familial late-onset Alzheimer's. The prohibition, though, will have little impact on us. We are now actively

searching through our exome to find the causative variant. Once found we can alert the entire global extended family and prevent its

further transmission: no great technology will be required. The entire symbiont community (including many highly trained professionals)

that has helped us manage this illness will then no longer have an obvious function. The power of disruptive change is moving to the

level of the individual: people in positions of authority will no longer be able to control it.

Edited by mag1, 20 March 2016 - 02:34 AM.

[Elus]

Back when I started this thread 3 years ago, I did not expect things to progress this quickly. I'm not alone in my surprise. AI experts didn't expect this kind of progress either.

 

Google had this to say about their recent victory against Lee Sedol:

 

"Experts predicted it would be at least another 10 years until a computer could beat one of the world’s elite group of Go professionals."

-Google's Research Blog

 

Just a few days ago, we saw that Canadian province Ontario plans to trial universal basic income. 

 

It's both stunning and a bit frightening that the world's socioeconomic and technological landscape is changing so rapidly. I cannot wait to see how this plays out. 

 

I have a feeling that we won't need to wait long.

Edited by Elus, 20 March 2016 - 03:15 AM.

[mag1]

Yes, those on the thread really should not be entirely surprised that the loop could soon close: we all know what would happen next.

The time for serious consideration of this has definitely arrived. It is entirely clear that once we fall a step behind the exponential curve

it will already be over.

 

And of course humanity is still ramping up. Asia is getting on board and apparently seem eager to see what might happen if they

hit the accelerator. It is not unreasonable that some of us might start to feel a little concerned.

 

Giving some hints of financial security to those who would be displaced by the approaching changes would not be a bad idea at this time.

Without such gestures, every inch of progress will be fought by those trying to protect their own self-perceived-interests. There will surely

be an extraordinary amount of wealth that will be created in the relatively near future. If we can find a way that everyone sees that they will

share in this bounty, then we will arrive at this promised land much sooner.

 

One of the Nature articles suggests a billion human genomes sequenced by 2040. This would be the end of medicine as it is currently

understood. Yet, the prediction of 2040 is likely based on the linearity assumption that they were trying to disprove. A billion human genomes could be only a decade

out.

http://www.nature.co...ur-mind-1.19431

 

As we hurdle rapidly into this hyper-technological world, one does wonder about how the natural world will fare.

Apparently there is now growing alarm in the climate science community that our planet has already reached the breaking point.

Perhaps hitting the gas truly is the best strategy: humanity has made such a complete mess of things that we will need

AI to put all the pieces back together.  

 

 

http://www.slate.com...ure_record.html

 

Earth hour is more relevant than ever.

Edited by mag1, 20 March 2016 - 04:31 AM.

[Mind]

The end result of the UBI is getting inserted "into the Matrix". Once you get people on the UBI, most will become unproductive yet demand ever more. They will ask why some people get to sail around the world in luxury yachts while they have to sit around in their "system provided" housing eat GMO food, and enjoy "system provided" entertainment. The most efficient way of dealing with these desires will be VR/Matrix-style existence.

 

Even though getting inserted into the Matrix (wire-headding, paradise engineering) would probably be quite nice, I don't like the thought of not be in control of my life and destiny. Sure, even in the modern day, I have little control of what happens in my life, but it is better than zero, IMO.

 

Also, why worry about climate change or the environment? Don't get me wrong, I love getting out in nature, am a big advocate for clean energy, and would like to see more natural areas expand and be preserved, but if people get on the UBI, dependent, and eventually live out some sort-of hive mind/VR existence, then why have "nature". Just pave over the earth and make it into one big solar panel. All that will be needed to simulate a "fantastic" existence is energy.

[albedo]

This factory is replacing robots with humans

http://www.weforum.o...ots-with-humans

Niches look to remain but global impact will be limited by definition

[mag1]

I was surprised how unanimous the positive comments were on the AlphaGo story.

My presumption was that typical Westerners would focus more on the negatives than the positives.

 

It was even more surprising that this was seen in such a positive light considering that the computer

design used was based on a general purpose problem solver. This was not a mere exercise in brute force

computing, instead AlphaGo can be understood more as a step toward creating a human like thinking computer.

 

It would be very useful to have expert input on where this recent development puts us on the path to Singularity. 

The article from Google talked of this as being a wormhole to ten years into the future. If AlphaGo could now

go into a feedback cycle, then one wonders whether the lift off point for extreme technological change has arrived.

 

The question of human employment prospects in an AlphaGo world would make for an interesting game of Go.

The white spaces could represent human employment prospects, while the black spaces could represent the employment

space of machines. We could always let AlphaGo play white.

 

(The most obvious way to avoid complete mass unemployment would be to simply prohibit innovation. There are several industries

that have successfully implemented such a strategy.)

 

Edited by mag1, 20 March 2016 - 08:37 PM.

[Mind]

Robots are better than human teachers, no surprise here: http://www.kurzweila...eplace-teachers

 

If it is "good" for robots/software to replace dead end jobs, then they should replace teachers as well.

[Elus]

New York Times [July, 1997] 'Computer needs another century or two to defeat Go champion'

 

''It may be a hundred years before a computer beats humans at Go -- maybe even longer''

 

"But winning the $1.4 million prize promised by the Ing foundation to a program that beats a human champion may be an impossible dream. The offer expires in the year 2000. Go programmers are hoping it will be extended for another century or two."
 

[mag1]

Finally see where the support for AI would be coming from. Nations could support AI research as part of their overall industrial strategy without firmly

embracing the wider Singularity implications. Clearly robotics will over a huge near term market that will translate into substantial employment prospects

(unless robots are used to make robots).

 

At the same time a universe saturated with intelligence could be somewhat detrimental for business. A microbe that could produce hydrogen would mean

free energy, some sort of a vat that could grow food ... or a 3D printer that could make anything I wanted ...

 

There will be a drift in the general direction of Singularity without there being any large sense of urgency that achieving the Singularity is a goal in itself.

At this point 30 years to Singularity is beginning to seem conservative. If someone decided to hit the accelerator again, then we would be quite close.  

 

Does anyone have a list of the next set of AI goals that we should be on the look out for?

Have to admit that I would like to try out a robotic tennis trainer.

Edited by mag1, 21 March 2016 - 10:17 PM.

[PWAIN]

When Dolly the sheep was announced, there was a lot of talk about how this was not expected for at least a decade, probably much more. I am not sure that that breakthrough really got us that much further ahead. I do wonder if this breakthrough is going to really result in earlier AGI.

 

Mind, there is good reason to believe we are already in a VR/simulation. I do sometimes wonder about the lack of evidence of extra terrestrial life, could this be because our VR world is created without ET's because that would require ET's to be in the VR matrix or the use of AI. I wonder if it could be that an AI forced us into this matrix as an alternative to extermination. The AI does not want AI's to be stuck in here with us so we don't get ETs.This onion could have many layers.

 

Either way, we will probably know soon enough when we awaken our version of the AI. Interesting times ahead.

 

 

[mag1]

Really glad that Google was secretive about this research.

Outside of such a research powerhouse, there would be dribs and drabs.

Everyone would be able to move the goal posts along the way.

Once the breakthrough was achieved there would be little, if any surprise.

 

By concealing the research (as was done with AlphaGo) we are now in the afterglow of a research breakthrough.

Few (any?) commentators had predicted that Go would now be another computer solvable app.

 

More than a few people must now be wondering how far off can the Singularity be?

[mag1]

The edurobot story highlights the overwhelming benefits that can arise with information technology.

 

Clearly it makes no sense for a single teacher to teach a roomful of students with up to a 50 point

difference in their IQs. The most valuable resource in the universe is being squandered due to a bad 

educational model. The "Every student deserves their own edurobot" movement will without question

enable a new era of high human accomplishment. Instead of a world with a few geniuses rowing the boat,

everyone can grab an oar and get to work.

 

Not only will infotech create enormous value in education, but also in dementia care. The FDA has recently

approved an exoskeleton for personal use. It would be such a blessing for our family to have our loved one with

dementia to be up and walking about again. Having a robotic friend would also be such a large step forward.

 

Notably the scenario mentioned above will never materialize if humans had to be employed to provide these services.

One should not hold one's breath for a personal 40 hour a week tutor for every student. The entire workforce would

need to be devoted to teaching. A personal assistant for everyone with dementia is also not feasible.

 

The bankruptcy of much of the developed from the Alzheimer's pandemic is looming with caregiver:patient ratios

not much different from school. How could we possibly afford to provide a higher standard of care without immediate bankruptcy? 

Edited by mag1, 27 March 2016 - 09:18 PM.

[Elus]

I wonder what awaits when this one gets out of the lab...

 

IBM wants to accelerate AI learning with new processor tech - "Resistive" chips could speed up neural network training 30,000 times.

 

 

[Elus]

Asia is getting on board and apparently seem eager to see what might happen if they

hit the accelerator. 

 

I think IBM's recent announcement (linked in my post above) may also be aimed at accelerating progress. 30,000 times would be a rather dramatic speedup.

Edited by Elus, 29 March 2016 - 01:22 AM.

[mag1]

Yes, there has been a tragic loss of human potential due to the embrace of politico-economic systems that never made sense.

Yet, people such as HG Wells anticipated the rise of Asia over a century ago.

The tribal impulse to imagine equal signs everywhere, even when they were illusions has cost humanity profoundly.

 

We should pledge not to let this happen again.

The optimal development of human potential should be the driving rationale of our global society.

If edubots can be developed that are more effective than human teachers, then retaining an inferior educational model is simply

a price we can not afford to pay.

 

The article on IBM's new AI technology idea has me wondering whether we are far off from DIY AI. Time and time again it has been shown

that open sourcing such projects can result in a burst of creativity that can greatly accelerate progress. Never underestimate what teenagers

with a computer can achieve in their parents' basements (or garages)! Of course, we are now quickly reaching a point where these garages might

unleash a Singularity event. Somewhat scary really.

Edited by mag1, 29 March 2016 - 02:40 AM.

[mag1]

These airships can carry 20 tonnes of cargo.

A 100,000 pound house is 45 tonnes (metric tons). Don't get me started whether this is a short or long ton (tonne?).

 

The old jokes about being an early riser now because the builders for your neighbour's new house start using a jack hammer

right under your bedroom window at 5 am, or the lovely rain that you can experience inside your new house due to poorly done 

roofing, or the one about spending much of your life's savings for a new house and not receiving a house might now all be part

of the past.

 

http://www.ibtimes.c...usiness-1552127

 

[mag1]

This url highlights my earlier comments about the centrality of distribution in the modern economy over production.

$75 of the $299 US retail price of an iPod went to American distribution, nearly the same as the major components.

In further refinement of these costs it was found that only $5 was expensed for Chinese assembly for an iPod. 

I am very worried about what might happen if a way could be found to eliminate the distribution cost through some of the

transport technologies we have considered on this thread. If this were done, nearly the entire contribution of the

American economy to iPod production/distribution would disappear.

 

https://people.hofst...iPodmargin.html

 

Speaking of which transbotics seems to be inching forward.

 

http://www.albanydai...ence-12139.html

http://vegasinc.com/...making-first-u/

 

This is exactly what I was afraid of! 

Having swarms of these drones flying overhead 247 in an urban environment would be such a nightmare!

 

Shopping cart bots would be so much better.

This is one of those times where simply because we can do something does not mean that we should.

The Amazon prime time aircraft/copter does look pretty cool, though.

 

The demand for this air drone service could simply be overwhelming!

I would be first in line to have my supplements, books and other deliveries done.

Delivery costs would be so much less than it is now.

Perhaps it will even be free.

 

This shopping cart approach would be so much better environmentally and aesthetically.

Hmm, article below goes on to note that Domino's used something similar last year as an April fool's day prank. The article is dated March, so hopefully we are ok.

 

I would want to storm the online stores as soon as this one was on the market as well. I have been putting off buying new socks for a while now (suppose a few years). With shopping drones, I will just back up the truck so to speak.

 

Life will be better!

Life will be fantastic! 

 

http://arstechnica.c...delivery-robot/

 

 

 

 

Edited by mag1, 06 April 2016 - 01:35 AM.

[mag1]

How is government possibly going to be able to control transbotics?

As it is now there are a handful of huge international shipping ports that government can easily control in order to

apply tariffs other taxes and sometimes import restrictions.

 

It is possible in the nearish time horizon that tens of millions of transbots will be whizzing about the landscape of most cities

delivering who knows what to who knows who? It has been fairly easy for government to impose sales taxes on retail trade

because such trade happens at clearly defined locations. Some small towns might have a single supermarket. How could

control be achieved if there is a frenzy of activity from everywhere to everywhere?

 

What would happen if a massive cargo ship were to simply float along one of the coastlines of America and just let the transbots

drop off merchandise to whomever might be on shore or even at the international boundary of American jurisdiction? Transbotics could

make trade control entirely impossible! Complete global Free Trade could emerge as a consequence of transbotics in a way that centuries

of comparative advantage lessons and trade negotiations has been unable to achieve. 

Edited by mag1, 10 April 2016 - 08:24 PM.

[mag1]

Looks like the chemists are getting worried.

Well if they're worried, then I'll up the ante to being double worried.

Playing around with chemicals might not be that much tougher than playing expert level chess or Go.

 

One could only wonder what the chemical singularity might be able to achieve.

If we were to get to the point where you could just push the button and set some sort of objective (Alzheimer's perhaps), interesing

things could happen.. quickly. Of course, we could also wind up with a few billion tons of green mush.

 

http://blogs.science...thms-are-coming

Edited by mag1, 13 April 2016 - 11:26 PM.

[mag1]

And of course the drones are coming..or you might more properly say the drones have already arrived.

1 million drones sold last holiday season?

 

Suppose we will all be operating one of these to keep an eye on the government and to protect our air spaces

around our properties from invaders.The future has arrived.

 

Welcome to an entirely new industry and almost unlimited employment potential.

Might have to change the name of our thread. Sorry for being so optimistic!

 

http://media.precisi...ne-should-know/

[mag1]

The comment by the poster mag1 (no relation) questioning how government might control the approaching tranbotics revolution

takes on a whole different light when considering the above post noting that 1 million drones were sold last holiday season.

 

It appears that we are now already in a transbotic world. Surely it would not be that unexpected if some of the now millions of

drone owners might fly their drone to the drive through ( or land at the drone port) of their nearest fast food establishment in

order to buy a double cheeseburger. The economic crisis that has been predicted on this thread would then rapidly follow.

The entire retail landscape of our communities would vanish.

 

Surprisingly as noted in the link from my previous post, government has not appeared excessively interested in even bothering

drone owners to register.  

[mag1]

A cautionary tale of the social impact of mass unemployment can be seen in the recent announcement of a social crisis on

Canada's native reserves. These men who were once proud providers for their community are now recipients of

government assistance. Is this the future for all our communities with a guaranteed minimum income? In some of these

communities alcoholism approaches 100%. Social dysfunction has reached complete saturation.

 

The social science experiment where people have no purpose has already been done.

We should all take note as technology is clearly approaching that will put most of the rest of us in the same position as

those on the reserves: only two classes of people would exist those on the reserve and the wardens.

[Major Legend]

I believe in 20 years time, wealth will be the only denominator between people. With positions established much earlier. As technology advances the way in which we differentiate ourselves in the economy to provide additional value will depreciate. Increasingly more and more of the population have nothing to offer to the economy. If we do not want to split into 2 species, we must acknowledge this and start thinking about solutions.

 

I do believe we are quite far away from AGI, however I believe ASI will make 90% humans redundant long before AGI is even suggested or even attempted.

 

The scary thing is people are completely wrong about what is going to be made redundant, usually thinking whatever it is they do can't be replaced by software or machines. Humans have zero chance beating something designed to crush the task at hand, especially with deep learning and so on. A human takes 3-4 years to retrain in a new skill.

 

What people get wrong is ASI will actually rapidly replace specialised/technical positions - positions where the middle class make their living, the thing that ASI will find hard to replace is things like cheap labour, because your robots have to be cheaper than employing humans to do the same.

And yet we are multiplying like crazy. Universal Basic Income and population control are two good ideas. I think it's a bit negative to say UBI will cause complete dependency. Sure some people will be dependent, but many people will be creative, and will be able to focus things that are otherwise impossible in a capitalistic system requiring economies of scale. If you make something that only 1000 people use or buy it would be okay. The people you are talking about are people who have no education, and cannot do anything else - thats why they drink, but if the right options come for them they will choose them and become poets or whatever.

Just because UBI dependents can't contribute economically doesn't mean they can't contribute socially. People could hold hobbyist events and do things like filmmaking without worrying about their bills. I am not being an idealist - I do not believe a million people offered UBI would all just get drunk together.

 

I think it's just the truth - not everyone can be a computer engineer or something equivalent. It's clear there is a spectrum of people who will just be made completely redundant. So the rich really have two options, offer them a living or just leave them to die. 
 

I believe UBI is better than bigger government, because bigger government will just be corrupt as people are people. I don't think an AI overlord is that bad, humans are awfully crappy at government - it's actually a miracle we made it this far given the kinds of things the kings have traditionally done to the peasants.
 

Edited by Major Legend, 18 April 2016 - 01:26 AM.

[PWAIN]

Major, just to clarify, AGI stands for Artificial General Intelligence which implies broad spectrum intelligence much like a human as opposed to a narrow spectrum like playing chess or Go.
ASI stands for Artificial Super Intelligence and refers to an AGI that is substantially more intelligent than any human. Typically an ASI is considered the cause of the singularity.

An AGI would be capable of doing anything a human can do and probably better since it will probably be able to work continuously without tiring.

An ASI would be not just capable of doing anything any human can do but millions of things humans cannot do at all. An ASI will be so far ahead of us it we see us as semi retarded.

I personally believe an AGI will evolve into an ASI very quickly so the distinction won't matter for long.

[albedo]

PWAIN, tks for clarifying the acronyms, after wish I liked Major's post :-)