We should start asking us what's the purpose of humans, instead of talking about needs and labor, and jobs, and employment, and productivity, etc. We can automate everything, and augment a few individual to allow them to provide all the necessities for the entire planet, but why? What is the society we want to live in, what do we want to do with our lives? What are we trying to accomplish? Survival for now has mostly been solved, we're all playing a social game, we create conflict from boredom, success is a social construct, glory and fame are false realities. If we don't ask ourselves why? If we don't think about what we want for ourselved, what we want our world to be like, then we're simply a deterministic construct, like animals, we simoly do things we're programmed too, and I that case, we're bound to simply replace us all eventually. Maybe we're just part of evolution, we're the natural event that leads to sentient machines.
Approximately half of the world's population lives on less than $2.50 a day. Some 1 billion children live in poverty. One in nine people on earth suffer chronic undernourishment -- one in four in sub-Saharan Africa. Some 3 million children die annually from undernutrition. Diarrhea kills over 800,000 people annually, and malaria kills more than 450,000.
I know what you mean -- developed nations have made incredible progress -- but we have a long long way to go before we can say this species is sitting pretty.
> Some 3 million children die annually from undernutrition. Diarrhea kills over 800,000 people annually, and malaria kills more than 450,000.
That sounds like great progress. I could easily imagine that, even in absolute numbers, not per-capita numbers, that might be the lowest rate of preventable illness in human history.
imho very good observation. I myself am often astonished that money seems to be the central nexus of life for so many, where it should be: what can I do with it, how can I use it to shape the world to my ideals...
In a sense, replacing or augmenting labor is a false dichotomy. We can’t actually choose one or the other. ("Capital-Labor substitution" is a thing, in economics; it's structurally unavoidable.) Most nations will choose both, and the impact will fall unevenly across industries and demographics. That is, some people and professions will be replaced, and some will be augmented.
And among those people who are replaced, some will adapt well and others poorly. Those who adapt well, generally, are those who are willing and able to learn. I suspect that this group will already be in the habit of learning; they may even be part of a culture of learning inculcated by their parents, which will reproduce class privilege from one generation to the next. So those more adept at learning, the metaskill of acquiring new skills, are best positioned to weather the coming technologic changes, and thrive under new conditions.
If we step back as technologists, we have to look at the aggregate effects. They will be contradictory, and therefore, the narrative of technological change must reflect those contradictions. People will be replaced and augmented. Society will be harmed and helped.
And as members of that society who are pushing change, we need to think hard about how to help our fellows on the receiving end cope with their new reality. We need a gospel of learning, evangelists to preach it, and a church to support the practice.
Learning is great, but we don't actually need more people to do it. Need is a really important term, and we shouldn't use it inappropriately.
What we actually need is to decide whether the people left behind by advances in technology deserve to not be left behind. We can't continue to live in a world where people are increasingly finding themselves with zero options. Either they must become productive with no exceptions (and eliminated if not productive!), or we must support them.
I think the choice is clear. We are not just advancing technology, we're creating innumerable surpluses. The only deficit created by those advances is a construct of profit.
>> So those more adept at learning, the metaskill of acquiring new skills, are best positioned to weather the coming technologic changes, and thrive under new conditions.
Or, some are merely older workers, who by the time they acquire a new relevant skill, their working life would have ended.
Do we have enough care for the seniors and the disabled, enough research for tropical diseases afflicting underdeveloped nations, enough desirable housing for the homeless, and so on and so forth?
Unless the world is so desirable that average people with sufficient training and supervision can't improve it, we should care more about re-purposing wealth than having no more jobs for people to do.
Note: In the long run, if robotics and AI become so advanced that there is little need for average people, then a high level of basic income should be the norm and most people would be free to pursue their hobbies and social life full-time.
>Augmenting people is the same as replacing them.
>The person now has a greater output which means you need less of them.
Attention, i think this is a fallacy in multiple aspects.
Maybe they don't produce more, just higher quality?
On the other hand you have jobs where more is never enough, e.g. Scientists.
Higher quality is the same as higher production. That higher quality reduces the need for all sorts of people in QA roles.
Also, not all scientists are blue-sky researchers. Many do lab work, diagnose disease or provide services to patients. Others perform field tests or public outreach. Or they teach. Augmenting these people does replace people as fewer scientists are needed to perform a service once performed by many.
> This is the true opportunity of technology: it extends human capability. There is way too much handwringing about the possibility of technology eliminating human jobs, and way too little imagining new jobs that could only be done with the help of technology.
This is true.
It's also true that we're about to be at a point where extending the capability of a couple of central operators at a trucking company is going to introduce significant volatility (being unemployed in an economy that no longer values your current skills) to thousands of people's lives.
Whether or not that is a long term issue or a short term instability is irrelevant (though other than some historical hand waving -- it didn't happen last time! -- there's little evidence that it will be particularly short term), we still need policy to support those people.
The entrenched elite can always legislate protectionist policy and scale human effort horizontally through socialist interventions to make humans cheaper than state-of-the-art AI. This is especially true in the long interim between now and holy grail breakthroughs in object recognition.
Charlatans and predators can always rally the surplus humans into being political forces to siphon off a percentage of productivity gain made by "good enough" AI.
As an autonomous labor theorist, I'm kept up at night often about what to do with the surplus humans. On one hand, hard AI problems can be augmented with remote assistance and intervention. (A cleaner bot discovers spray paint on a wall.. is it graffiti, an advertisement, or part of a mural? Better summon a human eye to make sure.)
On the other hand, "good enough" AI will allow total war scenarios to arise from the ashes as manufacturing millions of murder drones 1.) helps create employment for a nation and 2.) allows the extermination of the surplus for cheap. A human death squad takes about 16-18 years of investment to create. A "good enough" autonomous drone death squad takes about a week.
The most immediate gains of autonomous labor will allow expansion into domains that are currently too expensive to get a foothold in or scale into. These opportunities are where many of the remote assistance jobs will be created. (Deep mining, space construction, etc)
All eyes are on navigating the political climate until those price points are reached, which is why the cry for universal basic income is so tantalizing. Helicopter money on the plebs so they won't revolt and destabilize the global order that makes advanced robotic development possible. If they fail, then the attractiveness of cheap death squads to resolve the human surplus question will most certainly win.
So ignorance, just playing fast and loose, full understanding but actually caring about this BS metric? None of those inspire much confidence in someone making predictions about a complex issue.
Wouldn't another solution be to make sure human built products are seen as luxury goods compared to ones built by machines? You wouldn't even need to 'augment' anyone if you could sell 'human made' products in the same way as 'organic' produce or 'free range' eggs or those local goods that get certified as built in one specific part of the world.
I'm sure there would be a certain audience that would pay a premium for goods not made through automation if the label said it'd help the workers stay employed.
There already is such a market and the audience is paying premium. Yes, handcrafted products are a luxury good and yes it is justified by the creativity which went into the crafting. Am I missing your point?
First thing coming to my mind: https://en.wikipedia.org/wiki/Cuckoo_clock
No, just saying that this could possibly avoid some of the issues with automation. Might see more handcrafted websites and apps and other stuff as well as traditional cuckoo clocks and what not.
Interesting. The author is proposing we team humans with machines. It will certainly happen, but it doesn't solve the deeper problem of technological unemployment when one person with one robot can do the work of 20.
I think a solution for the problem is for groups of people to associate and create non-profit "caregiving corporations" which will have the purpose to provide for their members. Such a corporation could buy land and cultivate for food, have repair shops for cars and other equipment, a hospital, a school, could invest and buy properties around the world for tourism (for members), and many other services.
As robots begin to replace humans, caregiving corporations should invest in robotics to capitalize on this technology in the name of the people. That's how people could have easier access to the benefits of the new technologies.
Otherwise, there will be big multinational corporations with big automated factories and people will be at the mercy of the government, having nothing to offer to employers. At least in the past everyone had his work to offer in exchange for a living wage. Not any more. So regular people need to own capital instead.
Horses don't strive to be employed and humans don't worry about the horse employment rate, either. I don't understand why the comparison to horse-unemployment is so widespread, a horse is an outdated technology in the current context, not an unemployed worker.
The future of IA is a surplus of people with no job, a wall that condemn those to starvation, more to come, as in the brexit, other will be thrown out of fronteers. In the end only a little group of people will remain, the rest will be in the other side, the wall describe your future. A wall for AI and elites against humans. </end panic>
Edited added: the winner takes it all will happen in all fields, only one social network, one big bank, only one country, only one language will be allowed. Slow runner will be taken out the the race, don't complain, the new AI has shown our world is a one dimension manifold, a total order will be established, noise will be eliminated, power and glory at the top, the day in which the market get all the wealth and the humans are slaves, the triumph of technology and the dismiss of humanity, who care about the future? noone can look back or you would be petrified.
One example related to software is augmenting decision-making before trying to replace it. PayPal's first major fraud detection algorithm worked by automatically flagging/delaying a small number of cases and having humans review them. I've tried to do something similar in my jobs when working on cases that are trivial for humans to judge but very difficult to completely specify automatically. Of course you can generally improve these systems with machine learning over time, so that a human reviews 1 in 10,000 cases instead of 1 in 100, but having a dual human-computer system is generally way more efficient at the beginning.
Really, why not replace them? Its inevitable they will be replaced, the company which does replace humans at a low cost will beat the one with humans in it.
