Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

What jobs are safe in an automated future?

8 Post author: PuyaSharif 11 January 2012 08:48PM

The trends are clear, more and more work that was previously done by humans are being shifted to automated systems. Factories with thousands of workers has been replaced by highly efficient facilities containing industrial robots and a few human operators, bank tellers by online banking, most parts of any logistics chain by different types of automatic sorting, moving, and sending mechanisms. Offices are run by less and less people as we're handling and processing fewer and fewer physical documents. In any area less people than before are needed to do the same work as before. The world is becoming automated.

These developments are not only here to stay - they are accelerating. Most of what is done by humans today could easily be done by computers in a near future. I would personally guess that most professions existing today could be replaced by affordable automated equivalents within 30 years. My question is: What jobs will be the last ones to go, and why?

Often education is pointed out as safe bet to ensure being needed in the future, and while that is true its not the whole story. First of all, in basically all parts of the world the fraction of the population with an academic degree is growing fast. Higher education will probably not be as good as a differentiator in the future. Second, while degrees in the fields hot in the future is hot in the future there is no guarantee that the degrees hot today will be of any use later on. Third, there is a misconception that highly theoretical tasks done by skilled experts will be among the last to go. But due to their theoretical nature such tasks are fairly easy represent virtually.

Of course as we progress technologically new doors are opening and the hottest job year 2030 might not even exist today. Any suggestions?

Comments (100)

Comment author: jimrandomh 12 January 2012 05:20:30AM *  10 points [-]

My question is: What jobs will be the last ones to go, and why?

If your goal, in asking this question, is to plot a strategy that keeps you employed for as long as possible, then you probably ought to backtrack and look at the goals that's meant to serve. There are better ways to ensure long-term financial security, better ways to maintain a sense of purpose, and better ways to keep busy.

Comment author: roystgnr 12 January 2012 04:04:23PM 7 points [-]

better ways to ensure long-term financial security

Such as? Although I admit this may be likely to change in the future, for most of human history the vast majority of the median's individual's wealth has been the net present value of their expected future earnings.

Comment author: PuyaSharif 12 January 2012 12:31:42PM 1 point [-]

My goal was/is to start a discussion around: 1. Strategies today for maximizing probability of being needed in the future. 2 Even more interesting, what tasks are hard/easy to automate and why? 3 The consequences automation will have on global economy. So far, the comments covers a little bit of all.

Comment author: TheOtherDave 12 January 2012 05:56:37PM 9 points [-]

Politician. Not necessarily in its current form, but I suspect humans will continue to want to feel in some way involved in planning their own futures, and be willing to provide status and power to selected other humans who ostensibly represent them in that process, for as long as there are recognizable humans. (The selected representatives will grow less and less actually relevant to planning the future, but that has almost nothing at all to do with the matter. It's not entirely clear to me how relevant to that they are now.)

Athlete. Again, not necessarily in its current form, but I suspect humans will continue to want to identify with humans who win in competitions of physical skill against other humans. To some degree I put soldier in the same bucket, but there are issues with identifying that as a safe job, and I find it more likely that we'll replace soldiers with more effective machines as they become available than that we'll replace athletes

Reality show contestant. This is a field where observing "people like me" rather than "people who are actually any good at what they're doing" is the whole point, after all. Replacing them with more efficient machines would seem to miss the point altogether.

Parent. Dunno if that counts as a "job" by this post's standards, but I think we will continue having humans raise other humans long past the point where we develop technology that's better at it.

Comment author: David_Gerard 13 January 2012 11:10:20AM *  3 points [-]

Sysadmin. Remember the wise words of Mel Brooks: "Fuck! Even in the future, nothing works!" I am fully confident I will be able to find employment until the singularity if I want to.

Comment author: khafra 13 January 2012 01:38:02PM 3 points [-]

Likewise, I feel secure in computer security. Until computers actually get better at securing and attacking themselves than humans, the value that depends on keeping your system secure will continue to increase as automation does.

Comment author: Solvent 12 January 2012 04:23:36AM 6 points [-]

Well, one thing that humans are superior at is modeling other humans, and their reactions to things. So to the extent that people do anything at all, they'll probably do that.

For example, music composition, writing fiction, and similar artistic endeavors require that the artist know what people enjoy. I think that that will be done by humans for the foreseeable future.

Also, things where you actually want the person doing it to be a person will continue. Counseling? I dunno.

Comment author: PuyaSharif 12 January 2012 08:29:54PM *  3 points [-]

For example, music composition, writing fiction, and similar artistic endeavors require that the artist know what people enjoy. I think that that will be done by humans for the foreseeable future.

Regarding music composition; there are already algorithms being developed for predicting the potential of a song becoming a hit. Next step could be algorithms that creates the songs by themselves. Its all about optimization with positive feedback. Algorithm: Create a piece of art A such that A has a high probability of satisfying the ones experiencing it. Input statistics about human nature + reaction to previous generations + reactions to man made art of the same sort. Most people wouldn't care about how that piece of art was made. (But I guess this will take a while)

Comment author: beoShaffer 12 January 2012 10:19:07PM 1 point [-]
Comment author: Gartonia 06 March 2014 02:43:12PM 1 point [-]

More than that, Philip Parker has created a robot that can write book. They aren't the best, but it's still a huge start.

Comment author: D_Alex 12 January 2012 06:10:12AM 1 point [-]

Well, one thing that humans are superior at is modeling other humans, and their reactions to things.

I was thinking along the same lines, and came up with: Cook/Chef, Hairdresser, Fashion Designer and the like as being among the last jobs to go.

Comment author: Solvent 12 January 2012 06:20:05AM 1 point [-]

I was particularly thinking of the creative aspects of these fields being human-suited.

For example, I think it's reasonable that a machine could create food accurately, but a chef would still be employable in deciding what kind of food to make.

Hairdressing again, I feel could be done for most people by a machine. The rich and famous would have their hair done by people, and the hairdressing machines would copy those styles.

Fashion designer I agree with.

Comment author: D_Alex 12 January 2012 06:35:10AM 1 point [-]

I do not disagree with you, apart from one minor point: I think the amount of "creative aspect" in routine, cookbook cooking, hairdressing etc is underestimated. Eg. in cooking the kind and quality of the ingredients must be continuously assessed, and decent cooks taste their food while cooking to ensure a good final result.

"Artificial taste" not only seems rather difficult, but also does not have much development priority... yes, good cooks should be safe for a while yet.

Comment author: Emile 12 January 2012 09:38:40AM *  9 points [-]

Your machine may not be able to "taste", but it can check various properties like color, opacity, pH, viscosity, temperature and other chemical properties ... and what's more, it can check them continuously and track their evolution. Combined with a good enough up-front design (setting ideal ranges for those), and it should be able to get close enough to the way a cook would adapt.

I believe that's pretty close to what happens in food factories making things like ice cream or canned soup etc. - though they probably have tasters too, just to be sure.

Automation isn't about making a machine that does what a human does, the same way. Planes don't flap their wings, cars don't have legs, meat grinders don't have an arm wielding a butcher's knife - machines don't need to be able to taste to replace most of what a cook does. If there are some sub-steps that machines are particularly poor at, there may be a workaround, for example ensuring more homogeneity in the ingredients than a human cook would need, or using a dedicated human to do only that step (tasting, for example).

Comment author: NancyLebovitz 12 January 2012 10:01:24AM 6 points [-]

In a lot of work, people are willing to sacrifice quality to get cheapness and convenience.

Comment author: David_Gerard 12 January 2012 09:52:06AM *  4 points [-]

Eg. in cooking the kind and quality of the ingredients must be continuously assessed, and decent cooks taste their food while cooking to ensure a good final result.

Commercial cooks do not, in fact, taste everything they make. Not even close.

Comment author: perturbation 12 January 2012 02:18:45AM 5 points [-]

All of the research that I've done as an undergrad has basically been data reduction/data analysis; I think that a narrow AI could be developed within 20 years to do all of what I've been doing with much greater efficiency. So much of it is automated/on the computer already that taking the extra step would not be relatively difficult.

Understanding specifically what I was doing and why took some time, and there are still things left for me to learn, but I think that that sort of work is not secure at all from being replaced by automation. If anyone is curious, I'm talking about the field of spectroscopic chemical abundance analysis.

Comment author: buybuydandavis 12 January 2012 01:02:16AM 5 points [-]

Automated systems solutions provider.

Comment author: Dr_Manhattan 12 January 2012 01:25:21AM 11 points [-]
  • Programmer (AI programmer?)

  • Pole dancer

Comment author: James_Miller 12 January 2012 05:03:09AM 8 points [-]

Pole dancer

Sexbots

Comment author: Konkvistador 12 January 2012 09:50:21AM 4 points [-]

"When work is privilege of the rich?" seems sort of relevant to this discussion. Especially:

"The "masses" mostly waste away in cubicles doing makework. The elite do things like install mosquito nets in sunny third world locales"

I assume that anonmouse is referring to activities like the Peace Corps in which young people from well-off families do work in third world countries that would be considered menial low-wage low-status work if done in the United States.

This demonstrates two points.

(1) If people are convinced that a certain activity is high status or in some way “meaningful,” they will gladly do it no matter how objectively stupid it may be.

(2) We currently see jobs becoming a luxury for the rich. Only young people from well-off families can afford to “work” in the Peace Corps.

Comment author: thomblake 11 January 2012 09:48:57PM 9 points [-]

We'll probably redefine 'job' long before we've lost them all. As humans lose the ability to work for a living, we'll stop pretending we do need to work for a living. Then, we'll still have professions for the prestige and whatnot that goes along with it, even if everything truly useful can be done by machines instead.

Comment author: Maelin 12 January 2012 01:31:34AM 16 points [-]

I predict that the transition period between modern capitalism and post-capitalism will be very awkward and uncomfortable, as the number of jobs that it is economical for a human to perform slowly shrinks and shrinks, putting more people in unemployment. Assuming no singularity beforehand, the point at which we collectively decide that everyone gets a good quality of life provided for them without having to work for it will probably come long after the point at which it is warranted.

Comment author: thomblake 12 January 2012 03:00:13PM 3 points [-]

Assuming no singularity beforehand, the point at which we collectively decide that everyone gets a good quality of life provided for them without having to work for it will probably come long after the point at which it is warranted.

Indeed, there are plenty of people who think we're already way past the point at which it's warranted.

Comment author: mwengler 12 January 2012 02:42:13PM 3 points [-]

In my mind the question comes down to what do humans value that other humans do that will be hard to do or hard to value if it comes from a machine?

Art in all forms. Live music performance. Painting, personal photography. Interior design. Even industrial design. For a long time the parts of this that can be automated will be used by experts to improve the quality of their performances.

SImilarly with engineering, I think it will be a long time before the tools replace the tool-bearer. Software will be easier and easier to write because the level at which it is being written will be higher and higher, and more and more (exponentially) cpucycles will be used to compile the software, but that "prime mover" putting it all in place will exist for a very long time. Similarly when math is automated there will still be physicists and mathematicians using the tools to develop theories, they will just develop them faster and more accurately than they used to.

The ownership idea is strong and will take stand you well for a long while (generations more I think). In the same way that patents and copyrights fade and expire, I would expect an increasingly large portion of basic production to be deprivatized in some fashion or another. When there is no human input into changing productivity, no human input in to managing the deployment of capital, it will simply not make sense to attribute that capital (and its earnings) to some particular owner anymore. HOW that transition takes place may be wierd and slow, but natural selection dictates that it will take place in some form in some facets of the human world, and that its greater efficiency over private ownership will make it grow.

But in the mean time, the arts, high end of science/math/engineering.

Comment author: torekp 12 April 2012 12:36:57AM 0 points [-]

In my mind the question comes down to what ... will be hard ... to value if it comes from a machine?

As atucker mentioned, personal servants will still be in demand, the more personal the better. The "oldest profession" may also be one of the last surviving. Sycophants, courtiers, jesters ...

Comment author: Thomas 12 January 2012 08:24:11AM *  10 points [-]

The safest job, the only one safe job - is to be the owner. You can't milk the cow better than a machine, you can't do a thing better than a machine on the long run. But you can own a farm and take the dividends.

Comment author: jhuffman 12 January 2012 03:27:37PM 6 points [-]

I think that may the only job that could be safe for a given length of time. Eventually what you own will be superseded by something owned by someone else, and what you own will be worthless. If you are constantly investing in different products, markets and technologies you might stay ahead of it for a long time but that isn't what most people think of as a "safe job". I think what people are asking for in a "safe job" does not and will not exist.

Comment author: DanielVarga 12 January 2012 10:39:48AM *  6 points [-]

But you can own a farm and take the dividends.

I don't know, taking dividends sounds like something very easy to automate. My piggy bank could do that.

More seriously, is being an owner even a job? If by job we mean "source of income", then the first big question is, is an economy sustainable if the only source of income for humans is ownership? (And probably redistribution, like a basic income guarantee.) My guess is yes, but I am not sure. (And the transition to this economy will be surely painful.) The second big question is, will the humans be happy about all their free time? My guess is yes, but I am even less sure.

Comment author: Thomas 12 January 2012 02:45:17PM -1 points [-]

being an owner even a job?

Even if it isn't, it should bring you the money, what is the most important part of the "job business". (The job you are doing for fun is a play.)

Comment author: Viliam_Bur 12 January 2012 10:18:38AM 6 points [-]

The safest job, the only one safe job - is to be the owner.

Depends on what kind of society you expect. In a communist revolution, being owner brings you a bullet in head.

I don't know what kind of society will emerge when most people will not have jobs because of automation... a good long term strategy should investigate this too.

Comment author: DanielLC 12 January 2012 02:14:22AM 6 points [-]

Jobs are not a scarce resource. Employees are. We do not have approximately 100% employment because the number of people is about the same as the number of jobs. We do because as long as there are people available we will be willing to hire them for some amount of money. If they have nothing better to do, they'll take it. If there's no jobs worth doing, they won't look for them, and won't be counted towards the statistic.

As what some longer living jobs might be:

  • AI programmer would be an obvious example. The last jobs will go when they finish their job.
  • Engineer. Someone has to design the machines that take everyone's jobs.
  • Entrepreneur.
  • Maid. Hard for machines to do, and could single-handedly soak up half the workforce, at least. (I'm not sure how maids' maids would work.)
  • Any job that machines can do, but targeted towards upper classes.
Comment author: roystgnr 12 January 2012 04:23:10PM 6 points [-]

Historically your thesis has been frequently disputed, and the people disputing it have always been proven wrong. Population increases 10-fold? That doesn't mean 90% unemployment, because now there are also 10 times as much demand for the products that new jobs can create. Automation puts 90% of farmers out of work? Fine, they'll just eventually go back to work producing the products newly demanded by the remaining much-more-productive farmers and each other.

There do seem to be a couple limits on how far that process can go, though.

One is the Malthusian limit. Recast in economic terms, there is a finite amount of capital around, and as the population of labor increases past it, the price of the latter in terms of the former can be expected to drop below subsistence level. This keeps conspicuously not happening to humans, mostly because technology keeps improving the capital value of existing material goods, but it seems unwise to count on it not happening forever. This limit happens to other animal species all the time when their expansion into new territory hits its carrying capacity, and the results aren't something we'd be happy with for humanity.

The other is the "robot world" limit. Although fears of "androids will take all our jobs" have mostly been replaced by "computers and industrial machines will work alongside us and make our jobs way more productive", something more like the former could still happen eventually. This one has also already been observed happening to other species, surprisingly. Technologies like saddles, horseshoes, horse collars, carts, etc. made horses more and more productive, more and more popular... right up until the invention of the internal combustion engine, at which point most horses were no longer worth the cost of boarding them.

In either case, the safest job is "person with capital". That either means enough economic capital to be self-sufficient via (possibly collective) production and trade without asymmetric employment, or enough political capital to convince the angry unemployed mob to give you a big share of the spoils when they fleece the people with economic capital. But now we're veering into politics and I'll stop.

Comment author: DanielLC 12 January 2012 06:40:10PM *  4 points [-]

One is the Malthusian limit.

There will be enough jobs. Whether or not you can live on the pay isn't guaranteed. If there's too many people, or even if the gap between the rich and poor is too large, the pay of the lowest jobs can fall below subsistence level.

This keeps conspicuously not happening to humans

The general version hasn't. I think there has been times when the wage for untrained labor has fallen below subsistence.

The other is the "robot world" limit.

If robots do every kind of manual labor for us, there will still be jobs. The only limit is when they do the manual labor and the mental labor. It's not the robot world limit. It's the singularity.

That still very well might not get rid of jobs. It's unlikely that people as we know them are the best at any job, but as long as you have them around, and they're going to have those pretend MMORPG jobs anyway, it might be best to use some of them.

In either case, the safest job is "person with capital".

Strictly speaking, that's the only kind of job. Being good at something is considered human capital.

I'd agree that "person with monetary capital" will be a job that will last up until the end.

This isn't really relevant, but it's worth commenting on. I can't shake the feeling that I'm arguing with you, but we seem to be in agreement.

Comment author: mwengler 13 January 2012 03:57:03PM 2 points [-]

The malthusian limit used to happen to people all the time. You can graph population against various supporting resources and watch it rise and fall in large regions historically. What is confusing the issue is that the industrial age has exploded productivity so much that we have not had a modern malthusian disaster. We came within a few inventions of having one in the 1960s or 1970s but the green revolution inventions aced that one out.

Current trends suggest the mid-term future of earth's population is maxing at about 10 billion and then falling slowly. This will put off any malthusian disasters through the mid-term.

I HIGHLY recommend "The Rational Optimist" by Matt Ridley. Most amazing collection of useful points of view and information I have seen in a long time. Collapse by Jared Diamond is also a fun compendium of local malthusian disasters over the millenia.

Comment author: atucker 12 January 2012 04:56:34AM *  2 points [-]
  • Maid. Hard for machines to do, and could single-handedly soak up half the workforce, at least. (I'm not sure how maids' maids would work.)
  • Any job that machines can do, but targeted towards upper classes.

I find it pretty easy to imagine a future where rich/still-employed people regularly employ groups of servants, or maintain an entourage. It has a lot of historical precedent, and is already the case in many parts of the world.

Comment author: mwengler 13 January 2012 04:03:19PM 2 points [-]

Good point about the rich. Even now, a hand-painted painting by the original artist costs way more than even an extremely good copy made by a hired painter with great skill, which costs more than signed limited-edition reproductions, which cost more than unsigned reproductions. We only need a few zillionaires who value original art to keep a number of people employed as painters.

I suspect in the real world, production will tend to distribute across people in such a way that any system with more production will tend to have a higher amount of stuff owned by even its poorest, no matter how good the machines. If a trivial picture of capitalism puts all of a tremendous amount of wealth in a minority's hands, that minority will have to be willing to slaughter viciously parts of the marjority to "protect" its property rights. If you are that rich, why would you be willing to do something so distasteful? Better to just let things get reorganized and only have a tenth or a hundredth of the zillions of dollars you theoretically would have had under the old organization.

Comment author: Viliam_Bur 12 January 2012 10:25:37AM *  2 points [-]

Some people value "hand made" things, so there will be some market.

I would use machines to help me make those "hand made" things faster. Seriously, even with laws, some use of machines would be allowed; I just need to find a way to use it to maximize my total productivity. For example, even if the product must be done by my own hands, is it OK if a machine guides my hands, provided that the machine parts will not touch the product? I could watch some movie while the machine moves my hands to do the work.

Comment author: TheOtherDave 12 January 2012 03:19:05PM 1 point [-]

Perhaps I don't even have to be in the room... either cut my hands off and replace them with superior prosthetics, or clone additional hands, and attach them to the machine while I take a nap.

Comment author: Viliam_Bur 13 January 2012 09:31:22AM 2 points [-]

In that case I would clone all my body (except the brain), put it in the other room and put a webcam there, so my clients will have real-time proof that I am not cheating.

After some more thinking... perhaps it would be easier to generate a false webcam output. :-D

Comment author: CronoDAS 13 January 2012 04:28:51AM 1 point [-]

Jobs are indeed a scarce resource when you can't get one.

Comment author: mwengler 13 January 2012 03:50:45PM 1 point [-]

It is virtually always the case that you can get a job, you just can't get one that meets some rather stringent criteria for longer than you expected it to take to get such a job.

Labor is a productive input, out of equilibrium it can be wasted but eventually a system will tend to employ it fairly efficiently , which doesn't leave much of it laying around unused.

Comment author: CronoDAS 14 January 2012 12:43:19AM *  2 points [-]

Jobs shouldn't be a scarce resource, but markets can stay out of equilibrium for a surprisingly long time and economies can fail to produce as much as they are capable of doing. Right now, in the U.S., there are about five times as many unemployed workers as there are job openings, and there have been times when things were worse, most notably during the Great Depression of the 1930s.

Even McDonalds is getting more far more applicants than it wants to hire.

Comment author: wedrifid 14 January 2012 01:05:33AM -1 points [-]

Jobs shouldn't be a scarce resource, but markets can stay out of equilibrium for a surprisingly long time

Especially when there are laws involved!

Comment author: mwengler 16 January 2012 05:53:26PM 0 points [-]

Especially when there are laws involved!

Lawless areas seem to have gigantically lower productivity than lawful areas. ALL the high per capita productivity regions of the world have overwhelmingly strong governments providing physical security to lives and property at extraordinarily high levels.

Do you have any evidence that lawless areas do economically better in ANY measure than heavily regulated areas?

Comment author: wedrifid 16 January 2012 10:09:16PM *  1 point [-]

Lawless areas seem to have gigantically lower productivity than lawful areas. ALL the high per capita productivity regions of the world have overwhelmingly strong governments providing physical security to lives and property at extraordinarily high levels.

Do you have any evidence that lawless areas do economically better in ANY measure than heavily regulated areas?

That would be a rather bizarre thing for me to be try to provide evidence for. I have not and would not assert any similar thing. That's the most surprising thing I've had appear in my inbox for ages!

My comment was merely an agreement with CronoDAS that in practice markets are not perfectly efficient and acknowledgement that with respect to job markets in particular things like minimum wage laws contribute to this. And you know what? Without claiming much expertise and so with rather low confidence I say minimum wage laws are a good thing. Ours (in Aus.) are comparatively high (compared to, say, the US) and it seems to work fine.

Comment author: James_Miller 11 January 2012 10:06:24PM 18 points [-]

Government jobs protected by politically powerful unions.

Comment author: timtyler 11 January 2012 11:41:39PM 5 points [-]

What jobs will be the last ones to go, and why?

Probably management - because they are in charge of hiring and firing everyone else.

Comment author: cousin_it 11 January 2012 11:55:54PM *  6 points [-]

In Manna, some managerial jobs are the first to be made obsolete.

Comment author: timtyler 12 January 2012 01:37:22AM 6 points [-]

The early chapters of Manna are good - but it's still fiction.

We already know what the first things to be automated are, and management doesn't seem to be too badly hit.

Comment author: shminux 11 January 2012 09:29:56PM *  12 points [-]

Downvoted for groundless assumptions and for failing to google the basics. There are more, not fewer physical documents produced, because it is easier to produce them, the number of bank tellers has actually increased, etc.

Third, there is a misconception that highly theoretical tasks done by skilled experts will be among the last to go. But due to their theoretical nature such tasks are fairly easy represent virtually.

Name three.

Comment author: Konkvistador 11 January 2012 10:06:16PM *  21 points [-]

Third, there is a misconception that highly theoretical tasks done by skilled experts will be among the last to go. But due to their theoretical nature such tasks are fairly easy represent virtually.

Actually I think he may be right, since this is basically a consequence of Moravec's paradox.

"The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard. The mental abilities of a four-year-old that we take for granted – recognizing a face, lifting a pencil, walking across a room, answering a question – in fact solve some of the hardest engineering problems ever conceived.... As the new generation of intelligent devices appears, it will be the stock analysts and petrochemical engineers and parole board members who are in danger of being replaced by machines. The gardeners, receptionists, and cooks are secure in their jobs for decades to come. "[2]

But why might this be so?

“Encoded in the large, highly evolved sensory and motor portions of the human brain is a billion years of experience about the nature of the world and how to survive in it. The deliberate process we call reasoning is, I believe, the thinnest veneer of human thought, effective only because it is supported by this much older and much powerful, though usually unconscious, sensorimotor knowledge. We are all prodigious olympians in perceptual and motor areas, so good that we make the difficult look easy. Abstract thought, though, is a new trick, perhaps less than 100 thousand years old. We have not yet mastered it. It is not all that intrinsically difficult; it just seems so when we do it.”[4]

Comment author: Dr_Manhattan 12 January 2012 03:32:38AM *  4 points [-]

I agree with the theory, but not with the practical conclusions. Yes, we invented automatic provers before automatic gardeners, because there are were some harder problems involved. But at this point they both seem with reach - for example, take a look at progress in self-driving cars just over the last 6 years. Driving is very similar to other natural problems. Cooking example is just silly in the first place. Well, maybe creative cooking will be harder, but cooking by emulation should be really easy for machines.

Comment author: Emile 12 January 2012 09:15:46AM 3 points [-]

... not to mention that a lot of food is being produced by machines, or at least in a heavily automated environment. I've visited food factories, I don't remember seeing many cooks. Even home cooking has been automated to a certain degree.

Comment author: jhuffman 12 January 2012 03:40:18PM *  4 points [-]

All these labor saving devices, even factories, are integrated by humans. While the productivity per worker skyrockets (fewer workers needed per X units of output), there is no factory that runs without people who do generally very easy tasks that are very difficult to automate.

The summer after high school I worked in a spray bottle factory. Yes, we made the spray nozzles like come on a bottle of windex. My job was to keep the bins full of the little parts that fed into the machine that assembled them. I also helped unload the boxes of the parts from the carts and stacked them near the bin where they needed to go. Someone else somewhere had a job to handle the "raw" plastic for machine that melted and molded the parts I needed. Someone else put the different parts in different boxes sorted for the cart driver.

These tasks were of course absurdly easy for any human to do with about five minutes of training. Somehow automating all this together into a single factory chain would have presented enormous challenges though. Because the labor is so cheap I could easily imagine that factory will run the same way for the next fifty years.

Comment author: TheOtherDave 12 January 2012 05:41:49PM 2 points [-]

I suspect the driving forces behind automating that sort of thing will ultimately be, not labor costs, but the relative slowness, messiness, and unreliability of humans.

That said, I also expect that the technology that can do those sorts of jobs more quickly, cleanly, and reliably than humans will be developed for different applications where minimally trained human labor just isn't practical (say, automated underwater mining) and then applied to other industries once it's gotten pretty good.

Comment author: jhuffman 12 January 2012 05:45:59PM *  1 point [-]

Maybe but its still easily 50 years away. People are "messy" but they are so cheap and you need so few of them - there is no capital tied up in them at all its just a month-to-month expense. Even if you lease equipment you are still paying for the cost of the capital tied up in it. The diminishing returns for automating such a small cost will ensure its continuity for quite some time I think.

Comment author: TheOtherDave 12 January 2012 06:23:13PM 3 points [-]

Predictions in years are less and less meaningful to me as I go along. I'd give .6 confidence that we're no more than 5 tech-generations away from being able to build a fully automated mining facility (just to pick a concrete example), and no more than 3 generations from there to being able to build one in a way that would be cost-effective (given current-day labor costs and raw materials prices) for at least some application... perhaps underwater mining of rare earths.

I also expect that along the way, selected raw materials prices will increase enough (in inflation-adjusted currency) that using current-day prices is absurdly conservative. Then again, I also expect that along the way we'll see several failures of such equipment that cause as much as half a commute-year (current-day) of environmental damage, which might set the whole project back by decades. So, who knows?

A commute-year, incidentally, is a measure of risk (e.g., death and property/environmental damage) equal to that caused by people commuting to and from their jobs in a given year. My guess is that half a commute-year is typically more than enough to cause the majority of Americans to insist that a new project is way too dangerous to even consider. (Of course, that doesn't apply to the project of actually commuting to work.)

Comment author: PuyaSharif 11 January 2012 10:09:39PM *  8 points [-]

1 Maybe I should clarify: Are the tasks previously done by bank tellers becoming automated? Yes. The fact that the number bank tellers has increased does not invalidate my statement. If there were no internet banking or ATMs then increase would be much larger right? So its trivial to see that the number of bank tellers can increase at the same time as bank teller jobs are lost to automated systems.

2 I'll give you an extreme one. I am a few steps away of earning a degree in theoretical physics specializing in quantum information theory. Theoretical quantum information theory is nothing but symbol manipulation in a framework on existing theorems of linear algebra. With enough resources pretty much all of the research could be done by computers alone. Algorithms could in principle put mathematical statements together, other algorithms testing the meaningfulness of the output and so on.. but that a discussion interesting enough to have its own thread. I just mean that theoretical work is not immune to automation.

Organize all the known mathematics and physics of 1915 in a computer running the right algorithms, the ask it: 'what is gravity?' Would it output General theory of relativity? I think so.

Comment author: shminux 11 January 2012 10:30:53PM *  8 points [-]

Organize all the known mathematics and physics of 1915 in a computer running the right algorithms, the ask it: 'what is gravity?' Would it output General theory of relativity? I think so.

You have fallen victim to the hindsight bias. The parameter space of the ways of reconciling Special Relativity with Newtonian gravity is quite large, even assuming that this goal would have occurred to anyone but Einstein at that time (well, Hilbert did the math independently, after communicating with Einstein for some time). Rejecting the implicit and unquestionable idea of a fixed background spacetime was an extreme leap of genius. The "right algorithms" would probably have to be the AGI-level ones.

I am a few steps away of earning a degree in theoretical physics specializing in quantum information theory. Theoretical quantum information theory is nothing but symbol manipulation in a framework on existing theorems of linear algebra.

"Theoretical quantum information theory" is math, not a natural science, and math is potentially easier to automate. Still, feel free to research the advances in automated theorem proving, and, more importantly, in automated theorem stating, a much harder task. How would a computer know what theorems are interesting?

Comment author: PuyaSharif 11 January 2012 11:05:26PM 1 point [-]

1 Hindsight bias? Quite a diagnosis there. I never specified the level of those algorithms.

2 Which part of theoretical physics is not math? Experiments confirm or reject theoretical conclusions and points theoretical work in different directions. But that theoretical work is in the end symbol processing - something that computers are pretty good at. There could be a variety of ways for a computer to decide if a theorem is interesting just as for a human. Scope, generality and computability of the theorems could be factors. Input Newtonian mechanics and the mathematics of 1850 and output Hamiltonian mechanics just based on the generality of that framework.

Comment author: shminux 12 January 2012 12:09:18AM *  3 points [-]

1 Hindsight bias? Quite a diagnosis there. I never specified the level of those algorithms.

I have, in my reply: probably AGI-level, i.e. too far into the haze of the future to be considered seriously.

2 Which part of theoretical physics is not math?

Probably the 1% that counts the most (I agree, 99% of theoretical physics is math, as I found out the hard way). It's finding the models that make the old experiments make sense and that make new interesting predictions that turn out to be right that is the mysterious part. How would you program a computer that can decide, on its own, that adding the gauge freedom to the Maxwell equations would actually make them simpler and lay foundations for nearly all of modern high-energy physics? That the Landau pole is not an insurmountable obstacle, despite all the infinities? That 2D models like graphene are worth studying? That can resolve the current mysteries, like the High Tc superconductivity, the still mysterious foundations of QM, the cosmological mysteries of dark matter and dark energy, the many problems in chemistry, biology, society etc.? Sure, it is all "symbol manipulation", but so is everything humans do, if you agree that we are (somewhat complicated) Turing machines and Markov chains. If you assert that it is possible to do all this with anything below an AGI-level complexity, I hope that you are right, but I am extremely skeptical.

Comment author: Douglas_Knight 12 January 2012 04:41:04AM 5 points [-]

Mathematicians would probably call much less of what physicists do "math" than the physicists. Let me focus on statistical mechanics. A century ago, physicists made assertions that mathematicians could understand, like the central limit theorem and ergodicity. There was debate about whether these were mathematical or physical truths, but it is fine to take them as assumptions and do mathematics. This happens today with spin glasses. But physicists also talk about universality. I suppose that's a precise claim, though rather strong, but the typical prototype of a universality class is a conformal field theory and mathematicians can't make heads or tails of that. The calculations about CFT may look like math, but the rules aren't formal.

PS - bank tellers per capita fell from 1998 to 2008, though not much.

Comment author: PuyaSharif 12 January 2012 09:39:14PM *  2 points [-]

I agree that the conceptual (non-simply-symbol-processing) part of theoretical physics is the tricky part to automate, and even if I am willing to accept that that last 1% will be kept in the monopoly of human beings, but then that's it; theoretical physics will asymptotically reduce to that 1% and stay there until AGI arrives. Its not bound to change over night, but the change will be the product of many small changes where computers start to aid us not by just doing the calculations and simulations but more advanced tasks where we can input sets of equations from two different sub-field and letting the computers using evolutionary algorithms try different combinations, operate on them and so on and find links. The process could end where a joint theory in a common mathematical framework succeeds to derive the phenomena in both sub fields.

EDIT: Have to add that it feels a bit awkward to argue against the future necessity of my "profession"..

Comment author: prase 12 January 2012 12:40:33AM 1 point [-]

Organize all the known mathematics and physics of 1915 in a computer running the right algorithms, the ask it: 'what is gravity?' Would it output General theory of relativity? I think so.

Do you have any idea how to do it? What does organising the mathematics and physics actually look like? What are the right algorithms?

Nobody probably doubts that theoretical work can be done by machines. But your original claim was stronger: that these tasks are "fairly easy to represent virtually" and that the conjecture that they are going to be the last to go is a misconception.

Comment author: syllogism 13 January 2012 12:03:47PM *  0 points [-]

Are you a grad student? Because I don't know much about theoretical physics, but I find it very hard to believe much academic research could be automated.

I'm a post-doc doing research on computational linguistics. I can't imagine automating my work.

Comment author: PuyaSharif 13 January 2012 02:02:22PM *  1 point [-]

Yes I am, and I'll soon start looking for PhD-positions either in physics or some interdisciplinary field of interest. I know I seem a bit over-optimistic, and that such radical changes may take maybe at least 30-50 years, but I'd guess most of us will be alive by then so its still relevant. My main point is that step by step theoretical tasks will move to the space of computation and the job of the theoretician will evolve to something else. If one day our computers in our computer aided research starts to output suggestions for models, or links between sets of data we haven't thought about comparing wouldn't those results actually be a collaboration between us and that system? You maybe cant imagine automating everything you do, but I'm sure you can imagine parts of your research being automated. That would allow you to use more mental resources for the conceptual and creative part of the research and so on..

Comment author: paper-machine 13 January 2012 12:18:02PM 0 points [-]

When do you believe the first human-equivalent GAI will be created?

Comment author: PuyaSharif 13 January 2012 02:11:06PM 0 points [-]

By human-equivalent i'd guess you mean equivalent in if not all, but in many different aspects of human intelligence. I wouldn't dare to have an opinion at the moment.

Anyone else?

Comment author: Thomas 13 January 2012 02:21:47PM 0 points [-]

I can't imagine automating my work.

You can't. Most can't, probably.

You have to come with some good reasons WHY that would be impossible.

Comment author: syllogism 14 January 2012 07:47:16AM 2 points [-]

Well I'm working on a sub-problem of artificial intelligence. So the task of designing a system to do my work is harder than the end-goal my work is trying to achieve!

Comment author: jdinkum 13 January 2012 04:53:46AM 2 points [-]

The US government has listed some trends here: http://www.bls.gov/emp/ep_table_203.htm

Manufacturing is clearly out. Services that involve interacting with other humans (health care, teaching) are in.

Personally, I think we've plucked a lot of low hanging fruit. I predict the maturation of some technologies that will make a few fields move towards obsolescence. I expect at major ( >%50) decline in employment for long haul truckers and tax preparers under that maturation. Retail sales will decline in favor of ship-to-home, and obviously some jobs are clearly disappearing altogether ( running a newspaper printing press,camera loader on a movie set, working at a video rental store, telephone line repairman).

It's easier to predict what will decline rather than what will thrive. Engineering is generally a safe bet though.

Comment author: mwengler 13 January 2012 03:39:57PM 0 points [-]

How does retail falling in favor of ship to home lower employment for long-haul truckers? You still have to get the stuff from the ports to the distribution centers. The only thing that cuts down employment for long haul truckers is automated trucks, and that is still a long long way away.

Comment author: jdinkum 14 January 2012 09:59:37PM 3 points [-]

Sorry, I should have been clearer.

I do mean automated trucking. I think that development is less than 30 years away. Self-driving cars are a functioning technology now, the main hurdles (not insignificant ones) being social and legal.

Like most human-replacing technology, where the industry still exists, but is now automated ( modern car factories compared to buggy whip manufacturers), I don't expect trucking to be 100% automated, but the human factor will be reduced significantly the next 30 years.

Vehicles will still need humans to gas, service, deal with theft and breakdowns, etc. It will likely start with truckers being able to sleep during trips, still one trucker per vehicle which decreases the down time of cargo by 30%. Then there will likely be one trucker per several vehicles travelling in a convoy. Potentially there could be "truckers" based in areas that handle vehicles as they pass through, rather than travelling with them.

Of course, as the costs of trucking goes down, the attractiveness goes up, so while it may take 1/4 of the manpower to handle the same cargo, the amount of cargo will increase.

I'm estimating a 50% decrease in trucking employment in the next 30 years, even as the amount of cargo doubles.

Comment author: Kevin 16 January 2012 01:24:47AM *  1 point [-]

Lobbyist.

Robotics sales engineer. (A former computer programmer or roboticist versed in the selling of site specific robotics implementation solutions to industrial conglomerates, super heroes, and rogue agents, not a robot acting as a sales engineer)

Comment author: Larks 11 January 2012 11:36:45PM 1 point [-]

First of all, in basically all parts of the world the fraction of the population with an academic degree is growing fast. Higher education will probably not be as good as a differentiator in the future.

In that case people will just need to get Masters degrees to differentiate themselves. A world in which more people are educated (ignoring concerns about the automation of learning) will probably be one with more teachers and lecturers.

Comment author: Konkvistador 12 January 2012 07:19:52AM 3 points [-]

A world in which more people are educated (ignoring concerns about the automation of learning) will probably be one with more teachers and lecturers.

Not necessarily.

Comment author: Larks 12 January 2012 07:50:39AM 2 points [-]

Obviously not necessarily, but true nonetheless.

I'm happy to make any reasonable bet/prediction-book entry on the subject, provided the term is short enough we'll remember.

Comment author: Konkvistador 12 January 2012 08:17:06AM *  3 points [-]

I'm happy to make any reasonable bet/prediction-book entry on the subject, provided the term is short enough we'll remember.

Same here. Let us put it on prediction book and assign odds once we agree. We could even make more than one prediction.

Unfortunately for the world, but fortunately for me, the human capital model is greatly overrated. Education is not primarily about teaching concrete skills. It's a stably wasteful way to sort people according to their intelligence, conscientiousness, conformity, etc.

I think this is true.

However as cognitive science and genetics advances we may yet see say the legalisation of the use of IQ tests or other tools for employment screening. This would eliminate much of the services (and waste) associated with education. If a place like Singapore did this, it just might be a big enough economic boost that other nations might follow.

Regardless many many people who are employed in "education" are employed in primary and secondary education, very little signalling benefits comes from these two. I could easily see say the number of primary school teachers per capita dropping by more than a half in the next 30 years.

Also lectures in the English speaking world might easily be completely outsourced to just a few professors in the next decade or so. Why have 300 lecturers do the same lecture every year? Why not reuse it for a few years, perhaps even decades on basic freshman courses. Why limit oneself to questions and answers from one classroom instead of everyone doing the lectures that year as well as questions from previous years? The opportunities to use user data to optimize the lectures far more than is humanly possible even for a top performer. The differences between the signalling values of universities is not flattened in any way, so the quality of the signal is preserved, indeed if the university happens to have a one of the few top lecturers or departments that make these lectures it probably gains a big prestige boost. This might not look like it will eliminate jobs, yet it automates a significant use of a academicians use of time.

And 10 years after that, what happens if we get a Googleversity that offers a quality computer science online education? Google could easily afford to hire part-time professors or professionals or even mere supervisors to have students in key countries take the tests in person. Google could also set the standards arbitrarily high. Having a degree from Googleversity might not signal conformity but it would signal intelligence, knowledge and conscientiousness. For some jobs this will be enough. At the very least Google itself would probably seriously consider hiring any applicants from the top 1% or so. If this proves successful, other companies might follow.

Comment author: NancyLebovitz 12 January 2012 09:59:02AM 5 points [-]

My impression is that requirements for credentials are at least as much an effort to identify conscientiousness and compliance as they are an effort to identify knowledge and intelligence-- which is not to say that it's a reliable method of identifying either.

Comment author: Konkvistador 12 January 2012 10:56:07AM *  7 points [-]

My impression is that the credentials are far more about conscientiousness and compliance than knowledge or intelligence.

Comment author: Barry_Cotter 13 January 2012 03:27:45AM 2 points [-]

However as cognitive science and genetics advances we may yet see say the legalisation of the use of IQ tests or other tools for employment screening. This would eliminate much of the services (and waste) associated with education. If a place like Singapore did this, it just might be a big enough economic boost that other nations might follow. (emphasis from quoter)

Given that the USA is the only nation I have reason to believe has banned (for all practical purposes) using IQ tests for employment screening there's something else going on as well. It's not like the Korean couldn't decide to hire school leavers based on their exam results, given that they already do that at one remove by hiring from universities that are ranked by prestige/average student points.

Comment author: Larks 12 January 2012 10:47:28AM *  1 point [-]

Same here. Let us put it on prediction book and assign odds once we agree. We could even make more than one prediction.

Ok, so how about something like

  • A = the proportion of the population of [region] getting an undergraduate degree in [year] will be higher than it was in 2010
  • B = the proportion of the population of [region] getting an postgraduate degree in [year] will be higher than it was in 2010
  • C = the proportion of the population of [region] employed in education in [year] will be higher than it was in 2010
  • D = the proportion of the population of [region] employed in higher education in [year] will be higher than it was in 2010

And we'd prediction book:

  • A iff B
  • A iff C
  • A iff D
  • B iff C
  • B iff D

Edit: formatting

Comment author: Barry_Cotter 13 January 2012 03:30:43AM 0 points [-]

Consider distinguishing between research and taught postgrads, and postgrads that build on the undergrad and ones that are postgrad purely or primarily for signalling purposes, like JDs, MBAs and MDs.

Comment author: Larks 13 January 2012 08:53:27AM 0 points [-]

Can we easily get statistics on those?

Comment author: Miller 17 February 2013 07:23:10AM 0 points [-]

I'm surprised posts like this are not more commonly discussed around here.