Jobs that robots can't do

[arkirk]

You seem to be blind to the fact that this is a wasteful age...the age of planned obsolescence.

That is not a technological problem. It's an economic one.

A lot of my working life was involved with emerging technologies in waste treatment. I am aware of a pulsating pattern of wasteful deployment of some technologies that generate huge amounts of wastes that can only be considered toxic.

And? Nobody's said that technology is all good all the time. Just that on the whole, they're a net positive. Besides, if you were still working in waste treatment, you'd know that that toxic waste output is decreasing; not increasing; with the deployment of most of the latest technologies.

The problem is the social organization that fosters technologies that jam things like coca cola into people with little thought for their well being and health.

Social organization and technologies don't jam coca cola into people. People do that themselves.

That is my definition of technological fundamentalism.

Your definition appears to be a strawman.

There is a necessary linkage between a deteriorating environment and a social or economic theory that seeks short term economic gains at every turn and never deals with its own wastefulness.

And the development of technology is not intrinsically linked to any of this.

You sound like someone who'se rabidly opposed to warfare; and therefore decries the use of fabrics because they're used in military uniforms.

Besides, even with seeking short-term economic gain, some companies and countries ARE dealing with their own wastefulness. Especially since doing so yields even more profit. For example, many of the companies in the massive port of Rotterdam produce large amounts of CO2 pollution. That's pure waste, right? Wrong. Significant volumes of that pollution are collected, filtered, and then pumped a bunch of miles over to largescale greenhouse farming where the added CO2 results in as much as a 30% crop growth yield. That's the profit driven economic system curtailing waste, right there. And they've been doing this for decades now.

It isn't perfect, but decrying the modern age as one that does nothing to deal with wastefulness is ridiculously ignorant. Do not project your decades out-of-synch American experience onto the rest of us, please.

Your argumentative system seems to be locked in hyperdrive...and aimed at me. The problem is you are driving carelessly and making propositions that you truly cannot have any knowledge of. Almost all forms of pollution are on the increase worldwide. I guess that doesn't seem to register on your ultra modern equipment. The technology of a society is always co-opted and guided by those who wield the most power in society. Today, that is the bankers, military industrial complex, the marketers, and major corporations. Our society makes and pretty much uses what these people want us to use. Their reasons for wanting anything at all are usually related to keeping their position in the economy. Much of our technology is devoted to serving those interests. I am amused to your resistance to news from the past and your contempt knowledge of our history. You really ought to cool it a little...chew some hormone chewing gum ...calm down. Your lack of regard for me personally is so insulting and condescending, I see no reason to continue this discussion. You are not trying to inform me and so many of your remarks are so ludicrous I will waste no more time on this.

 

[bilby]

That is not a technological problem. It's an economic one.

A lot of my working life was involved with emerging technologies in waste treatment. I am aware of a pulsating pattern of wasteful deployment of some technologies that generate huge amounts of wastes that can only be considered toxic.

And? Nobody's said that technology is all good all the time. Just that on the whole, they're a net positive. Besides, if you were still working in waste treatment, you'd know that that toxic waste output is decreasing; not increasing; with the deployment of most of the latest technologies.

The problem is the social organization that fosters technologies that jam things like coca cola into people with little thought for their well being and health.

Social organization and technologies don't jam coca cola into people. People do that themselves.

That is my definition of technological fundamentalism.

Your definition appears to be a strawman.

There is a necessary linkage between a deteriorating environment and a social or economic theory that seeks short term economic gains at every turn and never deals with its own wastefulness.

And the development of technology is not intrinsically linked to any of this.

You sound like someone who'se rabidly opposed to warfare; and therefore decries the use of fabrics because they're used in military uniforms.

Besides, even with seeking short-term economic gain, some companies and countries ARE dealing with their own wastefulness. Especially since doing so yields even more profit. For example, many of the companies in the massive port of Rotterdam produce large amounts of CO2 pollution. That's pure waste, right? Wrong. Significant volumes of that pollution are collected, filtered, and then pumped a bunch of miles over to largescale greenhouse farming where the added CO2 results in as much as a 30% crop growth yield. That's the profit driven economic system curtailing waste, right there. And they've been doing this for decades now.

It isn't perfect, but decrying the modern age as one that does nothing to deal with wastefulness is ridiculously ignorant. Do not project your decades out-of-synch American experience onto the rest of us, please.

Your argumentative system seems to be locked in hyperdrive...and aimed at me. The problem is you are driving carelessly and making propositions that you truly cannot have any knowledge of. Almost all forms of pollution are on the increase worldwide. I guess that doesn't seem to register on your ultra modern equipment. The technology of a society is always co-opted and guided by those who wield the most power in society. Today, that is the bankers, military industrial complex, the marketers, and major corporations. Our society makes and pretty much uses what these people want us to use. Their reasons for wanting anything at all are usually related to keeping their position in the economy. Much of our technology is devoted to serving those interests.

This next bit is comedy gold:

I am amused to your resistance to news from the past and your contempt knowledge of our history. You really ought to cool it a little...chew some hormone chewing gum ...calm down.

When you follow it immediately with:

Your lack of regard for me personally is so insulting and condescending,

It is as if millions of irony meters cried out in pain, and then were silenced.

I see no reason to continue this discussion. You are not trying to inform me and so many of your remarks are so ludicrous I will waste no more time on this.

You certainly shouldn't waste any more time on your belief system that is based on absurd falsehoods like "Almost all forms of pollution are on the increase worldwide".

The EPA certainly don't agree with you. But perhaps you know better than those whacky environmental scientists, through your magical power of being quite old.

Nostalgia is personal. The past seems better because our memories fool us into thinking it was. But scientific study of the facts shows us that our memories are wrong. The past contained some great times on a personal level. But the world as a whole is nevertheless better today than it has ever been, by almost any objective measure you care to choose.

 

[Togo]

This is getting too long, so I'll summarise. As I said before, I don't think we're actually that far apart...

First of all, I'm not saying that robots WILL do it better than we can... I'm just saying that there's no valid reason to think they might never be able to.

Ok, that's a fairly major point, and I'd agree with you. The point I was disputing was that robots would necessarily replace humans, and then we got bogged down on what is likely in the near future.

Secondly, we obviously don't need robots to do things better than we can do them in order to replace humans with them. We don't even need them to do it cheaper than us. After all, if that were true, we'd already have sent men to mars.

Agreed. It looks like we just mean different things by 'replacement'. Obviously robots are useful, and do tasks we might otherwise do ourselves. The same is true of dogs.

You could, but having one universal sex-bot isn't the same as having a unique and individual prostitute.

And why would you ever know the difference between having a universal (or more likely, a sizeable selection of AI personalities to choose from) sexbot or one that is genuinely unique? Do you imagine people will take their sexbots to dinner and discover that the guy at the next table has the exact same one as yours?

Yes, absolutely. That's something that prostitutes do, after all.

This is the point that keeps on coming up - whether we're talking about straight one-for-one replacement in which a robot takes over what a human does by duplicating every quality the human has, in the same way that the human does, but better, or whether we're talking about a robot just being used to do a task that a human does now. The latter seems painfully straightforward, the former is the one that to my mind has issues.

Maybe we're just talking past each other here?

What I'm saying is that any machine that has the same or similar features to a human is going to share the same issues as the human, because those issues arise not from being biological, or being human, but from being complicated.

I have seen no credible argument put forth to convince me that these issues stem from "being complicated". That's just your gut feeling.

Absolutely. Just as it is your gut feeling that these issues are not only solvable, but trivially so. I don't think this thread can support the detail that would needed to produce full arguments on these points, and I'm happy to disagree on something I see as a matter of degree, rather than a fundamental difference.

And, as I've already argued, it ignores the fact that intelligence allows us to give robots with a design *superior* to our own.

I think design can only be superior for a given purpose. I'm extremely sceptical of the idea of a design that's just generally superior

This, I've not heard of? With the same size, weight and performance characteristics? Do you have a cite?

http://io9.com/scientists-just-created-some-of-the-most-powerful-muscl-1526957560

""The energy per cycle that we obtain from these artificial muscles, and their weightlifting abilities, are extraordinary," says Baughman. "They can lift about 100 times heavier weight and generate about 100-times higher power than natural muscle of the same weight and length."

Very impressive! I've seen similar but not this particular application, which looks very cool. It also looks quite slow, working on thermal absorption, which was the issue with previous attempting to get strong artificial muscles. So I've got to ask if it really has the identical performance characteristics I was asking for?

Again, not heard of this from my contacts in neurophysiology and AI research. Citation please?

http://news.illinois.edu/news/14/0508plastic_ScottWhite_JeffreyMoore_NancySottos.html
http://www.sciencedirect.com/science/article/pii/S1875389211004159

etc etc. It's impossible (for now) for humans to regenerate past a certain damage threshold, a limitation that doesn't apply to self-healing materials to the same degree.

This I have seen. These systems can't regenerate past a certain damage threshold either, a much lower threshold than in humans, since they rely on the underlying superstructure for both delivery (large capillaries) and shaping of the new material.

While this is, again, very cool, you're kidding yourself if you think that this is anywhere near as good as human healing.

I'm not sure what you're talking about here. Obviously 3-4 hour workdays by now was an absurd prediction, but there's absolutely no doubt that automation is starting to approach the kind of levels where radical changes are needed to the way we divide wealth.

Oh, we're well passed that level, I think. And we've had some changes in the way wealth is divided. We need a lot more though.

One idea I've been thinking about is whether wealth should not be distributed per individual but rather per project. Wealth accruing not to people but to tasks. For an automated economy that makes more sense. That would involve absolutely breaking the link between wealth and social status though...

 

[dystopian]

Your argumentative system seems to be locked in hyperdrive...and aimed at me.

People often think it's personal when others disagree with their cherished ideas.

The problem is you are driving carelessly and making propositions that you truly cannot have any knowledge of.

Name one such proposition I've made.

Almost all forms of pollution are on the increase worldwide.

This simply isn't true. First of all, looking purely at the absolute global numbers doesn't really support your general argument very much; since any such increase will primarily be driven by the industrial rise of developing nations. You need to look at technologically sophisticated nations and their output instead if we want it to be relevant to the discussion. However, even looking at the global numbers it simply isn't necessarily true. CO2 emissions have levelled off, and are in fact lower now they were 10, 15 years ago. How about oceanic pollution? Well, the overall picture isn't good, but oceanic oil pollution has been going *down*. Those are two of the biggest and most important pollutants, and they have been *decreasing*.

Locally (meaning, looking at continent sized-divisions), we find that almost all forms of pollution are decreasing in the developed world; which seems to run completely counter to your "boo technology!" tirade.

I guess that doesn't seem to register on your ultra modern equipment. The technology of a society is always co-opted and guided by those who wield the most power in society. Today, that is the bankers, military industrial complex, the marketers, and major corporations.

This is true... but also a woefully incomplete picture. Technology is ALSO co-opted and guided by those who DON'T wield the most power in society. For every banker there is a hacker. For every soldier, there's a gamer whose hobbies drive technology just as much as guns do (if not more so). For every marketer, there's an artist whose needs push technology. For every corporation, there's a non-profit.

Is technology sometimes guided by those in power? Yes. But only very rarely (most of the time they find themselves in a sudden state of panic when they discover technology has passed them by and they need to stay relevant somehow) But they are not the ones that *control* technology. They can nudge things a certain direction... but it is the masses that dictate the road taken.

Our society makes and pretty much uses what these people want us to use.

Bullshit. Our society makes and uses what our *society* wants us to use. For every technology and product successfully pushed by the elite you so fear, there's those that die a quiet death. Most of the time, the people in power don't know shit about technology and only try to control it once its already too late for them to do so.

The narrative you're trying to sell comes close to treading on "illuminati!" territory.

Their reasons for wanting anything at all are usually related to keeping their position in the economy. Much of our technology is devoted to serving those interests. I am amused to your resistance to news from the past and your contempt knowledge of our history.

What the hell are you even talking about? When did I ever show resistance to news from the past? So long as said news is still relevant today, I have no resistance to it.

 

[dystopian]

Ok, that's a fairly major point, and I'd agree with you. The point I was disputing was that robots would necessarily replace humans, and then we got bogged down on what is likely in the near future.

Keep in mind that 'near future' was never defined. Some people think it's 15 years. I would consider the next 50 years to constitute the near future.

Agreed. It looks like we just mean different things by 'replacement'. Obviously robots are useful, and do tasks we might otherwise do ourselves. The same is true of dogs.

It isn't really a matter of replacing, ultimately.

1) We *will* replace ourselves with any of the jobs that are too dangerous for humans (already happening)
2) We *will* replace ourselves with any jobs where the robot/ai is far cheaper or far superior (already happening)
3) This will likely result in ever increasing levels of non full-time employment, which will necessitate economic/political change that lessens our reliance on paid work in order to live.
4) In a society where people don't rely on paid work, robots/ai will start to take even more jobs as the dull and dirty jobs that people used to take solely for the money will now not be attractive enough to attract a sufficient human workforce.
5) Jobs that carry the level of responsibility that people are often uncomfortable handing over to machines *will* be delegated to robots/ai if those systems are superior at the job (already happening).
6) In a society sufficiently automated through these processes, even leadership roles might eventually be handed over in this manner.

Conclusion: speaking from a deployment angle, it is *plausible* that humans can be entirely replaced in any job necessary for the functioning of society. Whether this *will* happen, is another question. It may be that instead, robots/ai will replace us in all areas to some extent, but not entirely. It might even be that we get a Butlerian Jihad.

To touch upon some of the points in greater detail:

Point 3: This is particularly noticeable in my country. We have perhaps the highest percentage of people not working full-time jobs in the world. 26.8% of men and 76.6% of women in the Netherlands work less than 36 hours a week. This is compared to just 8.7% of men and 25.8% of women in the rest of the EU. There's even been political discussion about adopting a 21 hour workweek (though the current government isn't ideologically inclined to agree with such measures), since there just isn't enough work to provide all-round otherwise.

All this is because of a number of reasons (the female discrepancy has less to do with sexism and more with a few historical oddities, but those are a whole different topic). One of the reasons is that part-time jobs here are actually quite well paid and the government has enacted laws guaranteeing the right to relaxed workhours. Another, very important reasons, is an exceptionally high degree of automation in a wide range of industries here; which is sometimes far ahead of the situation in other countries (the port of Rotterdam is a good example, with loading/unloading/transport of cargo already fully automated with almost no human involvement back in the early 90's. Many world ports haven't even begun the process of trying to catch up). In some ways, the Netherlands and its work-load distribution might serve as the prototype society for an increasingly automated future.

Point 5: As an example of this I always like to point to the Maeslantkering; a storm surge barrier that protects about 1,5 million people and which is in fact the largest autonomous robot on the planet (as big as the Eiffel tower, and 4 times as massive). Humans do not control the maeslantkering. It is controlled by an AI system that can not be overruled by humans. If the Maeslantkering AI decides that the barrier closes, it will close. Even if that costs billions of euros in economic damages. If the AI decides otherwise, it won't close... even if humans are starting to panic about the storm raging outside. Why do we leave the decision to an AI when other countries would insist on a human executive? Because the AI is simply far better at determining the risk and choosing the optimal response. The risk of a human making very costly mistakes was deemed too high.

The Maeslantkering shows me that humans are perfectly willing to leave important decisions to machines if those machines are better at it than humans. If we are willing to trust a machine with the lives of 1,5 million people; we would trust a machine with the lives of 17 million people. Or a billion. We would trust it with running the economy. We would trust it with running the government. We would trust it with running anything and everything. We're obviously not that concerned with giving the decision to a machine, so long as the machine is competent enough.

Yes, absolutely. That's something that prostitutes do, after all.

I think the proper word for those people are escorts, not prostitutes.

More seriously; so what? That's only marginally more awkward than showing up to a party wearing the same outfit as someone else.

This is the point that keeps on coming up - whether we're talking about straight one-for-one replacement in which a robot takes over what a human does by duplicating every quality the human has, in the same way that the human does, but better, or whether we're talking about a robot just being used to do a task that a human does now. The latter seems painfully straightforward, the former is the one that to my mind has issues.

If we *can* create a robot that does everything we do, that duplicates our "human-ness"; then we *will* create such a robot. There is no question about this. If a thing is possible, someone will do it. Whether said robot fully replaces us so that we can move about in hoverchairs doing nothing at all, or whether the robot just becomes another person part of our society, that is a different matter.

Absolutely. Just as it is your gut feeling that these issues are not only solvable, but trivially so. I don't think this thread can support the detail that would needed to produce full arguments on these points, and I'm happy to disagree on something I see as a matter of degree, rather than a fundamental difference.

I didn't say that the solutions themselves are trivial; just that assuming sufficiently advanced technology, things that would be hard to do with/on/for humans would be trivial for a robot/ai. That may seem like a circular argument, but it's more a matter of distinguishing between the ultimate limits of the two entities. Humans have already reached their natural limit. We could raise those limits through extensive re-engineering, but so long as we stay purely biological the theoretical upper limit will be much lower than the theoretical upper limit of a purposefully designed mechanical being.

I think design can only be superior for a given purpose. I'm extremely sceptical of the idea of a design that's just generally superior

The human form is generally superior to that of a snail.

And if you can imagine that a design can be superior for a given purpose, then you can imagine that a design could be superior in general.

Very impressive! I've seen similar but not this particular application, which looks very cool. It also looks quite slow, working on thermal absorption, which was the issue with previous attempting to get strong artificial muscles. So I've got to ask if it really has the identical performance characteristics I was asking for?

Pound for pound, inch for inch, it outperforms us. You made no requirement that it should operate at our speeds; although that seems to me to just be an engineering issue and not a fundamental one.

This I have seen. These systems can't regenerate past a certain damage threshold either, a much lower threshold than in humans, since they rely on the underlying superstructure for both delivery (large capillaries) and shaping of the new material.

Except that allows for the total regeneration of damaged areas. A human can't just regenerate an arm, even if you still have the bone structure for it.

While this is, again, very cool, you're kidding yourself if you think that this is anywhere near as good as human healing.

On some level it is. I never claimed that it was as good in the total package (yet); just that it was in some ways more efficient, which it is. To me, it serves as a proof of concept. An early step to a future beyond our wildest expectations.