How is AI any different from other technology improvement such as electricity, automation, personal computers, steam engines, or the cotton gin?
All of them destroyed some jobs, and created new ones.
Are you sure that the impact of this one thing will be larger than other events, such as climate change, space travel, video games or war?
The only thing that is likely from AI of is perhaps increased productivity, which, on its own, will not create chaos. And, we have had several false starts in the past on that with AI, so even that is unclear.
My suggestion is to continue living your life in the best way you think possible.
That is the primary difference. If a founder creates a technology from Singapore and sells it worldwide it can replace millions of workers without those workers or their countries seeing any gain, relative or otherwise.
A secondary difference is that it may fully obsolete some humans, just as cars fully obsoleted horse drawn carriages. In the past, farmers could move to factories, but once robots are able to understand simple commands and enact them with human-like limbs at a all-in cost lower than minimum wage, then a good part of human society becomes unable to earn a living. Sure, some things only humans can do (those that require a belief that the job must be done by a human: therapist, tour guide, politician, etc) but it doesn't seem especially likely that this will absorb the billions of low-skill workers.
cotton gin didn't replace dr's, lawyers, teachers, truck drivers, surgeons, taxi drivers, delivery drivers, writers (many news agencies already use AI to write articles) customer service / tech support, and even coders (there's a project to teach AI to code). The end goal of ai/automation is a society where nobody has to work at all but can live like kings because the robots take care of everyone..
The real question then becomes one of what will the masses do? Hedonism is a short trip, next sets in boredom. What we need is scientists and explorers in space. Ultimately, a risky adventure.
That takes even more science in areas that are much more undeveloped than AI: psychology, sociology, science-based politics and definitely biotech (which is on the cusp of important breakthroughs ) and space travel. (Elon Musk plan in essence)
The meantime would require a massive shift in thinking of elites from greed to charity and cooperation.
Doing all the things that fall outside the realm of economy:
- Caring for their children and elderly relatives.
- Pursuing artistic and scientific endeavors.
- Volunteering in their local community.
All of this would certainly keep me busy. I honestly don't know if this reasoning applies to 100% of people; that's a hard prediction.
Those things also created massive waves of unemployment that took multiple generations to adapt to, so no difference is needed.
Furthermore, there is considerable worry that the waves of innovation are accelerating and our social/educational models will need to change massively to keep up with the pace, otherwise the waves of innovation will occur faster than we adapt to them.
History suggests that the massive devastation and annihilation of labor resulting from a multi-continental war is extremely economically stimulating to the survivors. So many opportunities for jobs rebuilding and restoring the things and people that were destroyed!
(I am, of course, being sarcastic, but it's useful in these contemplations sometimes to consider the stakes by putting a null hypothesis of sorts on the table).
I don't see any sarcasm here. Two previous macroeconomic transitions (industrial revolution, post-industrial white collar economy) were indeed resolved by world wars.
I considered this scenario extremely likely to be repeated.
See, the main difference is the scale of the destruction. WW2 exactly ended because it escalated to really horrendous levels. (If the Japanese or Germans got the nuke first, we'd be in a very different world now.)
WWI culminated in development of chemical weapons. They were used in wartime, assessed, and stockpiled. But everybody was scared of ever using them again. Chemical weapons were not used in WWII.
The same thing happened with nuclear weapons in WWII. They were used, evaluated, and then stockpiled. But everybody was scared of ever using them again, and they weren't. Nuclear weapons most probably will not be used in WWIII.
Plenty of able-bodied people today don't work. They live off investments, the government, and/or the largesse of friends and family. Many lead lifestyles you or I might find very limiting, but they're Americans and they vote. They'd become available to work if something matching their work ethic, life situation, and BATNA was to come along. For many, nothing ever will.
Our world is not really in very much chaos over this, compared to the chaos of the past 100 years from other sources.
Futurists once predicted that the 40-hour workweek would become a thing of the past as automation made leisure time more available. I submit that we're actually seeing that, though maybe we don't recognize it: many of us working more than 40 because we have jobs that require it, and with those of us working single digits being pretty quiet about it, the aggregate average is slipping further below 40 every year.
A society where 10% of the people work really hard and 90% of the people barely work at all sounds more like the popular concept of a utopia than a dystopia, at least to me.
Revive unique ornate embellished design. Centuries ago, every product in the home of the wealthy was its own miniature work of art; almost necessarily, because it was all hand-made. Certainly kept artisans busy.
Today, flat design is considered desirable – by definition, devoid of embellishment. Embellishment is considered tacky. This trend is no doubt encouraged by mass industry, which benefits from more uniform design: it is easier to produce, and easier to mass-market.
A market for unique embellishment could provide employment for many. By definition, unique embellishment cannot be scaled, else it would not be unique. So-called "hipster" culture has already started going this route, with hand-made "artisanal" goods and services (food) becoming trendy and employing many producing what would otherwise be a fraction of a large factory's output.
Of course, the problem is maintaining the key trait – uniqueness – as desirable. Ornate embellishment can be mass-produced. The trick is convincing people to spend more to buy something that was inefficiently produced, as opposed to something that looks like it was inefficiently produced.
This is a cool idea, but isn't it a chicken or egg scenario? I need lots of disposable income to buy unique, handmade things. If not enough people have money to buy these things, then the market won't be able to support many artisans. I'm not sure if I'm missing something here.
You are right. It is a practice that has to start from the wealthiest. Of course, this is why artist communities thrive near centers of wealth. I'm not sure how to kickstart such a larger trend.
As a society, fund jobs where humans are still best. Caretaking of elderly, daycare, more teachers for more hands-on learning. Unless robots are full on conscious beings, people will always need that human connection for their own proper development and well-being.
Case in point, Norway has a ton of oil money and they could almost live off it as a country. But instead they save as much as they can and invest the rest in a strong social state that takes good care of the young, elderly, poor, disabled.
There's only so many people that can fill those types of jobs..I'm a developer though --so an introvert - I'm never going to want to take care of elderly for sure.. I could def. see myself teaching because - I do like to do that, but I'd probably teach coding, and when there are no coding jobs - when all teaching is about taking care of elderly or becoming a teacher... or a massage therapist - I'll be out of a job there as well.
There's 9 billion people on the earth and if we try and divy up all jobs that require a human touch -- we'd still have about 8.4 billion people left unemployed. Full on conscience robots aren't impossible either - further out for sure (2050-2060 by most estimates) - but w/ AI's help in researching/coding next gen AI that timetable could move up exponentially.
I'm not sure why people continue to ask this question when the threat of massive global food web and ecosystem disruption caused by human-driven environmental damage is arguably far more imminent than job displacement. I say arguably because there are arguments that state the contrary, but they aren't realistic. Drastic job displacement due to automation will take at least 10 more years, while we're already seeing unprecedented droughts, groundwater depletion, pollinator extinction, and massive biodiversity eradication in the oceans. Scientists are literally saying that we are past the point of no return [0]. So why would people think that a massive unemployment wave will strike before an ecological catastrophe does? I'll tell you why: because rich people don't even know what they're doing. If they did, they'd know it would be in their best interests (granted, long-term best interests) to use their massive capital advantage to improve the longevity of the human race, i.e. their customers. And, as time has shown, the only way to wrangle in stupid rich people is by a revolution, and probably a violent one unfortunately. So ironically, as I say that there are more pressing issues at hand, my solution is to have a revolution. Disclaimer: I don't want this to happen any more than the next person, but it will regardless.
> I'm not sure why people continue to ask this question when the threat of massive global food web and ecosystem disruption caused by human-driven environmental damage is arguably far more imminent than job displacement.
Because it tickles your brain more to think about this. (No sarcasm or anything, just neutrally offering a hypothesis.)
>How would you avoid chaos?
Encase myself in a robotic super suit and fend off the hordes one bullet at a time.
In all seriousness, I think the most likely course of action is that a significant part of society will retreat into virtual reality and drug addiction. As we are already seeing in Europe, east Asia, and parts of the US, in the absence of economic opportunity the population drops once people have some experience of modern life and see that it's not all it's cracked up to be.
I know several people in the US that have done this already. They don't really see a reason to participate in the system into which they were thrust more than absolutely necessary. Drugs and video games are how they fill their time.
I don't think we can pretend that most people will be employable, and employ them in make-work jobs. That won't fly even for those who benefit from these jobs - people are too knowledgeable and incredulous now. They will know it is just a game.
The population will fall, and indeed it already is falling. VR and drugs sound like a way to sedate people, not drum up chaos.
Indeed it is one possible model of a surprisingly stable society.
One other intermediate is to cart them off to space. And then next once population is low enough so that only the best remain in the genetic pool, we might see something special. Plato's republic? That was never possible because you could not have every one be a philosopher scientist as the support base was never there to begin with.
Utopia of a kind, probably with people who are no longer quite exactly homo sapiens but a new species even. (super advanced AI could be it)
There are many ways to have a temporary dystopia but how many to have a permanent one?
Huxley foresaw specialising humans to the kind of work they are supposed to do, but he never foresaw the tools we now have.
Redesign the monetary system from the ground up with modern goals in mind, as opposed to the legacy goals stated in the 1913 Federal Reserve Act.
As it stands, 2 of the 3 explicitly stated goals of the Federal Reserve system are:
1. achieve maximum employment
2. maintain stable prices
Exponential growth of computing and automation makes #1 a sadomasochistic, obsolete, and counterproductive goal as putting everyone to work would mean us having "too many cooks in the kitchen".
Stable prices is also counterproductive and counterintuitive, as automation is making goods and services better, faster, and cheaper. Prices should tend toward zero in an increasingly automated labor market.
Live a lifestyle by then which is very inexpensive and build up savings.
Meaning: Own your own property. Reduce or eliminate long term contracts (vehicle loans, two year commitments, etc). Reduce overall spending (so your savings last longer).
There's a reason why, in 2008, retirees as a group were least impacted. They had more stable lifestyles that didn't depend on an active job. Plus even if the AI revolution never happens, this strategy has benefits for the next recession, illness, or other unemployment.
Culture itself has to change. A winner-takes-all global capitalist system is already on track to pile loads of wealth upon the wealthy and hollow out the middle class into a modern serfdom.
We as a people need to better value kindness, volunteering, or even the basic dignity of life itself.
If we are lucky, we will achieve these things in time for the machines to learn them from us. If we end up creating super-intelligent machines in the mould of Ayn Rand, I suspect our future will be harsh and unkind.
> AI will lead us to the largest unemployment wave in history.
Citation needed.
But let's assume you're right. I think the last thing we need to do is hand more power to governments.
For instance, if AI bring massive productivity increases across many sectors, than price pressure will be downward. Anti-market solutions like price floors (to support business) or minimum wages (to support workers), will not allow us to fully enjoy the improved standard of living brought about by falling prices.
Some localities or states will outlaw self-driving cars (as they hurt driver jobs), but that will simply prevent delivery prices from falling in that locality, hurting the many for the benefit of the few.
Imagine if the cost of living dropped to 10% or 1% of what it is today, while at the same time quality and quantity of goods increased. We wouldn't need government to assure some income based on today's cost of living. Instead, a meager amount of savings, a small amount of charity, or a small amount of work could sustain everyone's existence. Propping up current wage levels with UBI would be fighting the natural economic forces driving cost of living down.
So, AI brings massive productivity, which drops cost of living... and then what happens to the industries whose costs were reduced as part of that?
What if the impact of AI reducing costs looks like the impact of on textile industries in countries that receive mountains of donated clothes?
"The result of this harmful practice is increased dependency on foreign aid in countries like Kenya and Uganda. Local industry can't compete, so factories close down, taking priceless jobs with them. It's a huge problem."
The main problem with robotics is the very high barrier of entry to making silicone electronics. This obviously concentrates power in the hands of the very few.
I think you meant the industries' prices instead of costs.
Take the hypothetical of costs declining massively on everything:
Instead of donating tons of clothing, we donate tons of food, building materials, cars, etc, since they are dirt cheap.
Eventually this leads toward the techno-utopia of having all earthly needs provided for by technology for free. So the jobs in Uganda are no longer priceless, they're unnecessary.
I don't believe in Utopias, but you get the point.
Enact a tax on automation. Don't do it immediately, do it when the tech hits the knee of the adoption curve. Use the income to implement a guaranteed basic income. Otherwise it's going to end in dystopia.
That sounds good in theory but what counts as automation and how do you put a dollar value on it? For example, if I use AWS to provision servers instead of hiring sysadmins to manage my own hardware, should I be taxed? If I use computer vision software to detect defects in products I manufacture, should I be taxed? By how much?
Historically, technological advances do not lead to widespread unemployment, though they do generate widespread speculation about future unemployment.
If this time is different, then I think the answer will be in something like universal basic income. Preparing for that and avoiding chaos means introducing social structures and ideas that can lead us out of capitalism.
To me this is like asking how do we crack the littering problem from tourists on Olympus Mons?
I still think we are really far from this. Siri, Alexa and Google Assistant aren't anywhere near good enough to replace a human assistant. Siri has been out for 5 years and look at how slowly the improvement in digital assistants has been since then. At the rate seen over the last 5 years we're at least a decade or more away from digital assistants putting human assistants out of work.
Now we talk about AI actually putting jobs that require more training than an assistant (which I am not denigrating, it is a hard job, but easier to train someone on than some other professions) and I think we're 50-100+ years away from AI actually causing some kind of employment crisis.
Implement a Georgist tax regime[1], take the money production facility away from the banks (direct issuance by the goverment for infrastructure, profit share the results), implement Distributist[2] policies to spread capital ownership as broadly as possible, and implement a gently eugenic basic income regime. (Not happy about that last one, but it has to happen to make things sustainable.)
I have no idea what a "gently eugenic" regime would be. Or, if it's what I'm imagining, why it would be necessary for sustainability. Mind elaborating?
An example would be to make basic income only apply to citizens over 18, so having additional children would not increase payments to a family, and then making children a large tax break, encouraging people who can afford them to have children.
I'm ambivalent about it, particularly since I'm catholic, but I'm afraid the alternative is idiocracy and, eventually, far more human suffering.
Don't let the emotional impact of the word "eugenic" cloud your thinking. Almost all social policies have effects on fertility and are therefore, either eugenic or dysgenic in practice. A basic income that isn't designed properly would be extremely dysgenic and, therefore, unsustainable. See my sibling comment about what a gently eugenic basic income regime might look like.
AI? 10 or 20 years? You're out by at least an order of magnitude.
The upcoming unemployment wave will be caused by the automation of middle class jobs exacerbated by the failing economy caused, ironically, by the decreasing population.
I believe the next step is expanding our footprint out of earth and colonizing other planets. That will provide people who aren't the beneficiaries of the current economy a new start. There will be fringe planets that are 'uninteresting' but still viable for human life.
I don't believe that a basic income would be the right way forward. It might be a good stop gap measure while we try and figure out super-fast travel and terraforming - but it will marginalize a huge portion of the society.
Historically no society has been kind to 'freeloaders' and I don't expect that to change anytime. (Please note the quotes - it is not my opinion, it is just what the opposition thinks). We don't need UBI, we need universal opportunity for jobs for everyone so that they can feel a measure of dignity and go about their life instead of looking at a life of netflix and chill.
The other choice I can think of happening is androidization of people. Once you are able to upload your consciousness into a network, there won't be any shortage of space. I think this is even further out though sadly :(.
Yeah I know. I'm about to say something that is sacrilege.
In nature, the population and the available resources tend to equalize. I am a bleeding heart libertarian (let people do what they want to do as long as it doesn't affect others but I don't mind paying taxes to provide a safety net), so I'm not saying people should die on the street, but maybe we should stop having as many children and things will stabilize.
I even think it would be cheaper long term if we subsidized birth control for the people who wanted it.
That's why I qualified it as a "bleeding heart liberterian". I don't believe the government should mettle to much in what private corporations do with a few exceptions, but I do believe its first responsibility is to its citizens and to provide a safety net. I don't mind my taxes going to help others who got left behind
As far as exceptions to when the government should mettle: it shouldn't allow companies to actively harm people (pollution), it should force companies to give consumers all relevant information on whatever they are trying to sell, and they shouldn't allow mergers that decrease competition.
The health and social systems desperately need people to work there. I don't think truck drivers want to start a career in nursing, but well, many people driving taxis today didn't think they switch to driving taxis when they lost their jobs.
Another way would be pumping money in (re-)education of all these people.
I'm don't think every one of them can be changed into a lawyer or physicist, but I saw a bunch of "street-smart" entrepreneurs whos businesses at least made enough money for them.
Conveniently, nuclear fusion is also 10-20 years away. With the energy problem going away at the same time, we won't need to worry about working anymore.
Actually it has been progressing nicely linearly so the original predictions were way too optimistic, but it is at break even now. In the meantime, there are major improvements to be made in fission.
Making it progress faster would take more genius scientists plus somewhat more funding. Not a lot more though. But where do you find and view do you cultivate such physicists and material scientists?
The biggest problem is the backward compatibility with the current system.
Negative income tax with a government backed efficient job market where to overhead of hiring someone for 2h is minimal. As a result, someone working 20h a week would have the same quality of life as someone working 40h today for a minimum wage. Worker can spend the extra 20h on education, leisure or more work.
That's a good personal solution but it does not scale. But by all means, do that; we need more people who understand the capabilities and limitations of the technology in the space.
All of them destroyed some jobs, and created new ones.
Are you sure that the impact of this one thing will be larger than other events, such as climate change, space travel, video games or war?
The only thing that is likely from AI of is perhaps increased productivity, which, on its own, will not create chaos. And, we have had several false starts in the past on that with AI, so even that is unclear.
My suggestion is to continue living your life in the best way you think possible.