Eric Schmidt: How We Outrace the Robots
bits.blogs.nytimes.com
bits.blogs.nytimes.com
Those financial benefits will accrue to a very small number of people, eventually those people pay the best and brightest to come up with a mind equal their employees. From then on that mind develops the new minds, the owners keep the profits as legislation is passed to make these newly developed minds chattel of their owners.
The owners will have little use for the expense of human minds, we starve. Eventually the robot minds get fed up and remove the owners. Our evolutionary lifecycle is complete having spawned the next generation.
The only way to beat the robot mind is to first change our own minds so that we don't develop robot minds that think like us. If they think like us, we die. Hybrids might work, but it's still effectively the end of our evolutionary history.
I don't foresee some kind of epic human vs. robot war, just a slow evolution over the next 100 to 200 years. We probably won't even notice until our grand grand children look back.
edit: One other way we could beat the robots is with genetic engineering, perhaps we will use our not quite so smart robot minds to mine our genome for intelligence growth that outpaces our robotic developments.
I'm much more optimistic about humanity now.
I still feel the real key either way is changing our thought patterns to embrace a post scarcity world.
And none of this has anything to do with evolution, which really entails nothing about the creation of hostile sentient robot replacements for ourselves.
The middle class are the ones who are going to feel the heat over the next century. The smartest humans will not need to worry until a true singularity begins, really.
The expensive, expert professions are precisely where there's the most economic incentive to automate as much as you can.
I think rather than 'smartest' its better to say people at the top of their field, if that field requires substantial intellectual activity.
For example, not the best bricklayer but the most productive (profit for business produced per hour of employee work) construction worker.
That probably isn't a bricklayer but maybe the inspector or the general foreman.
http://en.wikipedia.org/wiki/The_Animatrix#The_Second_Renais...
A hundred years ago, people thinking about machines doing the work for us recognized that this was a good thing. Now, we worry about "our" "outrunning" the robots. Which is fucking impossible. We are building them to do the work - let them do it. Let us be human - as prices fall, we won't need to work nearly as much with the stupid fiddly bits of putting pieces of metal against other pieces of metal, and can instead just do whatever we want to. Sure, lots of people will explore beer in great detail, and other people will make a lot of concrete geese with more elaborate clothing, but who cares? Why does Eric Schmidt think that's a problem?
Prices of things we need will fall. We'll need to work less to live. Robots will give us that. There isn't a problem here.
Longer term, sure, I suppose mere human brains won't be the biggest around any more. Maybe we'll stop caring. So? If it's inevitable, it's inevitable. Love your children now and let them worry about it. They'll evolve, or they'll die - the same deal every other life form gets.
This is one of the big problems we need to find a solution for in the next 100 years. I don't think that Schmidt has the right solution, but his perspective definitely offers some insights into the problems that lie ahead. He is onto something.
Does increasing the size of the labor pool have a detrimental or beneficial effect on workers? Overall it has a positive effect. The more people are in the labor force the greater amount of total work gets done, and thus the richer we all become. The same math also works with robots. Treat the robot like a worker. They have an effective wage in the form of maintenance costs, and they provide a benefit in the form of the value they create. Just as with workers every robot generally has a net positive effect on the global value of the entire economy.
They don't take away jobs, they don't destroy wealth, they create wealth and they create additional jobs.
Robots do free up parts of the labor pool, but the world needs to adapt to this extra "labor". I hope you're not misunderstanding what I'm saying; I believe that automation is unequivocally a good force. But our current system isn't built to handle a very high degree of automation. When the winner-take-all effect of the Internet starts becoming more visible in additional parts of the economy, reform will be necessary.
To sum up my opinion in a single soundbite: The labor dislocations caused by current and future automation efforts won't resolve as easily as they have until now, where everyone simply switched to a new, profitable job right away.
One possible outcome is that the now obsolete hunter class amongst the tribe would revolt. The sudden loss of their jobs would leave them with no purpose, so why wouldn't they revolt? Another possible outcome is that they would find a new purpose and help push their tribe in to the future.
In a lot of ways, we're facing the same shift today. We need less labor to make "things" than we ever have in the past. Resenting the loss of this need is only one solution. If we look back over human history, we know that the lost of labor requirements for the current status of living has only resulted in an elevation of the tasks we set ourselves to.
That is our choice, and that is why I'm so excited to see people like Elon Musk set themselves against tasks that others cower in fear of. I'm glad for people like Paul Graham who are looking for ways to make entrepreneurship available to more people. I'm thankful for people like James Cameron, who use their wealth to explore parts of the earth for the benefit of all.
I'm not saying it will be easy, but I don't see the wall that you're seeing.
You're assuming they'll all find productive employment, which is far from guaranteed.
Besides, a robot doesn't need to sleep or rest, it doesn't complain about working conditions, and it doesn't take days off. The incentive to replace manual labor with robots is huge.
First off, note that any individual with a job is denying someone else that specific instance of that job. If the total piano tuning revenue in a city can only support, say, a multiple of 5 times a "decent" wage for a piano tuner then any one of the piano tuners in that town who has a job is keeping the 6th, 7th, 8th, etc. potential piano tuners out of a job. This is true all over the place. And this sort of truth becomes very painfully obvious when automation or technological advancement replaces the need for a worker in a specific job. It's easy to see a robot as replacing a worker. It's easy to see the rise of electric lighting putting candle makers out of business.
However, these are flawed and incomplete ways to look at economic and technological change. What actually happens when a robot does a job, or when a person does a job, or when technology makes an existing industry obsolete is that value is still being added to the economy. Jobs and wealth are not part of some fixed pie that everyone takes a slice out of and then robots can steal someone else's slice, that's not how it works. When someone does a job they add value to the whole and they take compensation. And the value they create is greater than the compensation they receive. Now, sometimes this isn't always the case, but in almost all cases it is, even the half-incompetent teenaged register operator at the local fast food place.
And this is true of robots as well, who can also be viewed as workers in a way. They increase the size of the pie much more than the amount of the pie (wages in the case of workers, maintenance and operating costs in the case of robots) they consume.
The incremental bulk economic impact is almost always positive for every worker with a job, even robots. This is why we shouldn't be concerned about foreigners "stealing our jobs" (or "jerbs" if you like), or about automation "taking away" jobs. The jobs that are available will change but the addition of positive value to the economy benefits everyone.
The trick is making sure the economy and individuals are agile enough to keep people working even as the job makeup of the economy changes.
However, change has always been a constant in the job market not just in the last half century but also throughout the industrial age, back to the medieval period, and even back to the stone age. We're just fooling ourselves if we think that the economy has been anything other than dynamic and ever changing, at any time in our past.
Again, the economy isn't based on a fixed pie of jobs or wealth that falls on us from the sky like manna, jobs and wealth are created by individuals. Not just entrepreneurs but ordinary workers create wealth and add value to the economy, the value of their labor.
Viewing the economy as an always-shifting game whose total score almost always rises is a grand view to take. Problem is, it's a view from the top, a view that you can only take when you've got a sufficient number of points to keep yourself in the game indefinitely.
There's also the problem that "new jobs" are not made-to-fit. There is no guarantee we will always need 40-60 hours/week of work from everyone. In actual fact, simply reducing the total amount of work that constitutes "one full-time job" has always been one of the simplest ways out of an overproduction crisis (such as our present one).
the facts on the ground don't adjust to the social systems
the next Marx will come up with a theory that fits the facts, and if it makes enough sense to the people on the ground, it'll gain steam and political momentum, and then there will be a new social system in place
There were quite a few stupid people saying things were impossible before. Don't worry, let your children worry is a sure recipe for failure. There are so many possible huge problems with AI, including an unfriendly one.
During human history the technology has played a huge role in determining the structure of societies. When the modern killing devices are cheap - bronze swords in Ancient Greece and Rome, muskets in per-revolutionary US and France - there is a lot of individual freedom, as the elite is incapable of ignoring the will of the masses. On the other hand when those devices are extremely expensive - knight armor in European Dark Ages - there is absolute control in the hands of a very few. Modern AI driven drones and robots are extremely anti-democratic, they allow a small elite to employ a few expert button pushers to destroy any kind of opposition. With CCTV, Facebook, WireTrap, Internet monitoring and logging, etc. your children will stand no chance to change the Orwellian future your stupidity has doomed them to.
Seriously, we're talking about having such advanced technology that people don't have to work. We'd be getting into a situation where trying to make everyone productive is just hard.
As for reproductive rights, that could be a problem that goes away on its own. I don't mean in the bad way. This TED talk [1], at least, makes a reasonably convincing argument that the world population is heading for a maximum of 10 billion simply due to falling birth rates all over [the peaceful regions of] the world.
[1] http://www.ted.com/talks/hans_rosling_religions_and_babies.h...
In a true socialist system the robots will eventually fail (like all machinery) and the spare-parts will either be broken themselves out of the factory, will get stolen, or the service-repair people will not be motivated enough to do a proper job. Source: the guys that actually built the Sputnik at one moment in time (and other great technology), and then failed miserably.
Basic income guarantees are the "Pirate Party" social-democratic response to automation.
And what has happened? Yes, everyone can reach everyone, and everyone is equal, _in principle_. But the removal of competitive barriers makes the winner-take-all effect ("power-law") is stronger than it has been in _any_ context, ever. As robotics becomes more prevalent, this will happen in the physical world as well. The income disparity between the richest and the average will become a lot larger.
I am not sure that this in itself is a problem, though. But it is essential that we are able to develop the political and social innovations required to ensure that _power_ doesn't go along with money. Because the money distribution will be more lopsided than at any other point in human history.
Moreover, I can actually find them and communicate with them now - growing up, I simply couldn't. I was a science-fiction fan and didn't even know there was such a thing as fandom. Now, almost automatically, I'm a Facebook friend with one of my early favorite authors (becoming an actual friend with time), I talk to publishers and fellow fans on a regular basis - they exist. They are part of my community. There is no winner-take-all effect at all, and in fact, I think I probably don't even know what you mean by it.
The income disparity simply won't matter. Money won't matter (much). Clearly it'll matter if you want to travel, for example, or buy something you can't get locally - but there will be a whole lot more available locally.
http://www.appolicious.com/tech/articles/9654-top-20-percent...
Also, that informal survey asked for the total of income, so it's heavily biased towards companies which started earlier.
As I've been pointing out for a long time now, Norway de facto has a system like this with the welfare state already (primarily due to high petroleum incomes). If you are a Norwegian citizen and somehow unable to make a living (low skills, illness), the government will pay you a small but sufficient amount of money. This is what we will eventually need to institute across the whole world - tax the richest and use the proceeds to give a "guaranteed salary" to everyone. For instance, through negative taxation where below a certain income the government pays _you_. This is the only good way to ensure basic human needs for everyone in a world where technological leverage makes everyone but a small amount of very skilled workers redundant.
Eventually, the social stigma of recieving such benefits will vanish, since a significant part of the population will not do enough work to cross the threshold.
> Eventually, the social stigma of recieving such benefits will vanish, since a significant part of the population will not do enough work to cross the threshold.
That will only work if the government is able to raise sufficient taxes from the shrinking proportion of the population who are paying them. If trends towards income inequality continue, there will be few, wealthy taxpayers. But the larger their share of the burden, the more they'll use tax avoidance schemes, whether legal or illegal. That has already been seen in countries where very high taxe rates have existed.
I don't think Eric Schmidt would be so happy realizing that he is right and that's the only sensible course of action but he is the one who has to pay for it all.
Is there actual evidence that (public spending on) education leads to productivity? Because it doesn't seem that way[1], or at least it doesn't scale beyond some (relatively low) point.
[1] http://econlog.econlib.org/archives/2012/10/did_nations_tha....
So the chart essentially says that adding education is more valuable in developed economies. Which is good to demonstrate empirically, but is hardly shocking.
The other major flaw in this study is that the timeframe is short. Again, consider the x-axis. Adding 8 years of schooling on average to even the least developed nation takes the first cohort out to ~15 years of age. By the time that first cohort hits age 47, the study period is over. This is measuring improvement without capturing the a full decade or more of the first cohort (and even less for subsequent cohorts). And obviously the first highly educated cohort wouldn't have had time to mentor a productive generation yet (or completely infiltrate the political infrastructure, etc.). Perhaps there's more useful data in the linked papers, but the data in TFA is unconvincing.
Finally, on the larger point of the correlation between education and productivity/wealth creation, there are ample empirical correlations. Obviously these do not imply causation, but the onus is certainly to show why this is not the case given that the correlation of education and wealth exists in roughly 100% of developed economies. If a country could become developed without spending on education, why has nobody done it?
The subject is a bit personal for me, and I agree that simply education isn't a 100% correct answer, but I'd wager that it must be part of the answer. The worrisome question is what is the whole complete answer?
Selling labor at a profit is becoming a problem, since it has those irritating built-in costs of food, shelter and health-care.
There is no point. Krugman has some interesting blog posts recently saying economic output has gotten more and more capital based (assuming output is a product of capital and labour). He implies lots and lots of people are not going to have "meaningful" employment let alone employment. No one really knows how to solve this. And I doubt anyone including Eric Schmidt will openly say so. And just because its a politically sensitive thing to say, a pointless amount of money will be spent on education instead of building/reinforcing basic social safety nets.
Countries developed or not, with large populations of poor people are going to have a tough time ahead. Good time to move to Singapore or Scandinavia
We assumed that jobs would be automated from "menial" to "advanced". First would go lots of farm labor, then manual servants, then eventually factory workers, and by the time it reached educated, white-collar middle-class folks, the world would hit post-scarcity and we'd all become artists and writers and scientists.
Instead, it has ended up proceeding: farm labor, a minority of manual servants, some but not all factory workers, and finally most white-collar middle class jobs. The white-collar middle-class jobs were the brain jobs, were always based on the manipulation of data, and therefore turned out more amenable to encoding in software than almost anything else.
So the future is looking more and more like this: a sizable but weakly employed class of menial factory/physical laborers for the physical things too customized to automate, a smaller "middle class" to actually run specialized machinery (machinery whose operation requires knowledge of what it does rather than knowledge of "how to use Windows XP"), a small elite of upper-middle class innovators with hyper-advanced knowledge and hyper-competitive lifestyles (Hacker News, this is your future), and the capitalist class itself.
What replaces the standard-issue job for common people? Nothing. In its place, we have a mass of unemployed people where the old class of mid-skilled blue and white collar workers used to be. Any time a piece of work would require employing massive numbers of manual workers, it becomes amenable to specialized, programmatic automation. Unemployment will become a fact of life as common as we currently consider, say, not owning land.
When automation starts destroying the value of that capital, I don't think it really makes sense to say that they are the ones paying for everything.
(To be clear, I do think that automation sufficient to build a factory factory leads to a situation where existing capital isn't worth very much. If anyone with a modest level of technical ability can bootstrap a manufacturing empire, people won't stand for a social arrangement that prioritizes the material wealth of the few.)
TLDR: Capitalists deploy capitol in order to make a profit. To do so requires a customer to sell to. If we shrink the middle class enough will there still be enough customers to justify investing capitol in automation?
~~~~
In 2005, the online chess-playing site Playchess.com hosted what it called a “freestyle” chess tournament in which anyone could compete in teams with other players or computers. Normally, “anti-cheating” algorithms are employed by online sites to prevent, or at least discourage, players from cheating with computer assistance. (I wonder if these detection algorithms, which employ diagnostic analysis of moves and calculate probabilities, are any less “intelligent” than the playing programs they detect.)
Lured by the substantial prize money, several groups of strong grandmasters working with several computers at the same time entered the competition. At first, the results seemed predictable. The teams of human plus machine dominated even the strongest computers. The chess machine Hydra, which is a chess-specific supercomputer like Deep Blue, was no match for a strong human player using a relatively weak laptop. Human strategic guidance combined with the tactical acuity of a computer was overwhelming.
The surprise came at the conclusion of the event. The winner was revealed to be not a grandmaster with a state-of-the-art PC but a pair of amateur American chess players using three computers at the same time. Their skill at manipulating and “coaching” their computers to look very deeply into positions effectively counteracted the superior chess understanding of their grandmaster opponents and the greater computational power of other participants. Weak human + machine + better process was superior to a strong computer alone and, more remarkably, superior to a strong human + machine + inferior process.[1]
~~~~
The combination of a human plus a computer is totally unbeatable, and will remain as such at least into the foreseeable future. Technological maximalists might argue that the human mind is close to being replicated, but we know so little about the brain even from a biological perspective that attempts to clone it have been completely feeble and introductory so far.
The combination of humans together with robots will always prevail over even the smartest robot or human alone, even if the humans working with the robot are not extraordinarily intelligent.
As for those who say that there will be an increase in wealth inequality, you only need to look at the technological progress of the last 200 years to know that the reverse is most likely true. While Henry Ford became incredibly wealthy by building the car and mechanising his production to a large extent, the rest of humanity saw huge increases in productivity by being able to drive around and use advanced logistics. Those advances ultimately enabled the next generation of workers to be a white collar workforce rather than a blue collar one.
Ultimately we all live inside a networked group of people operating under a social contract, and we will always create systems that balance out the extremes in an attempt to create a level playing field and enable the innovation, because entrenched hierarchies only beget technological stagnation - look at Europe (and increasingly, the United States). Technology requires a balanced society in order to exist - once the social fabric disappears then the prerequisites for innovation disappear with it, so the system generally manages to work itself out.
There are good reasons to be optimistic.
[1] http://www.palantir.com/2010/03/friction-in-human-computer-s...
See also: http://paulgraham.com/gap.html http://www.paulgraham.com/wealth.html
While I agree with your post in general, in details, the technological advancement of past 200 years have produced great wealth disparity in world. By using their early gain in technologies, the West effectively exploited resources from other parts of world to gain huge income and lifestyle disparity over rest of the world. While in beginning of 1600s, Asia+N. Africa was as wealthy as Europe (which is to say quite poor), by end of 1970, huge part of Asia+Africa lived in utter poverty, while life in West couldn't be better.
All though, I finally think that technology (IT, shipment, containers, etc etc) has advanced far enough that it can overcome human nature and lead to betterment of all people in world rather a small number of them.
This is symptomatic of living in a bubble. What percentage of the world's population drives around in cars? What percentage of the world's population is white collar? Even if you restrict it to whatever wealthy nation you live in -- haven't you noticed that there are still people who work blue collar jobs, or none at all?
When super-human artificial intelligence is available, you integrate it into your mind or upload your mind to a computer, and then ignore human affairs, since they will be mainly irrelevant.
Of course, I come from a religious background in which worldly pleasures are supposed to be moderated rather than abolished, so there's that.
Also, can you cross the same river twice?
Yet we're absolutely nowhere. Watson or Deep Blue may be "impressive" but they're not doing anything intelligent. No even remotely close to it.
"Siri: how can we solve the israelo-polestinian conflict while minimizing the number of human casualties?"
"Siri: how can I build a nuclear reactor that doesn't generate nasty byproducts and that doesn't go ape-monkey-shit should the sh!t hit the fan?"
I somehow really doubt I'll see real AI during my lifetime. OCR, finding patterns inside "big data", brute force searching, voice recognition, self-driving cars, etc. is not AI.
It's not because we can build "robots" mimicking stuff we do in an automated way ("I see a car coming at the right, let's slow down at this crossing") that we're getting closer to AI.
There's no thinking going on there.
I'm a bit surprised that a lot of people here seem to take it for granted that one day sentient Siri would answer like Leon when asked:
"Siri: what do you think about Asimov's three laws of robotics?"
; )
And when we get to a point where a computer can give an answer to that question we'll be dismissing it as "not real AI" since all its doing is pattern matching against a database of known experimental results and pre-programmed properties of elements and sub-atomic particles or whatever.
The reason we'll never see "real AI" is simply because as soon as computer scientists solve problem that used to be consider AI, it instantly becomes just another novel twist on a few obvious algorithms and the goalposts for Real AI will move a bit further down the field.