Oldie but goodie. A piece of fiction describing how a computer system can do the job of human managers at fast food restaurants (scarily plausible), how this leads to a dystopia (slowly getting implausible), and how to avoid this scenario and reach utopia (give me a break).

New to LessWrong?

New Comment
38 comments, sorted by Click to highlight new comments since: Today at 1:22 PM

Spoilers...


Was anyone else jarred by the way Burt spoke so cynically about how he and the narrator would never leave their bleak Terrafoam prison because their lives in the 2030's were historically analogous to people living on a dollar a day in 2005, when people rich enough to help others out of third world squalor didn't do anything because they they were distracted by how amazing it was to swim in a pool! Burt argued that they had no more realistic prospects for improvement than people in the slums of Calcutta because no one who owned capital would take the time out to help them.

...and then after the deus ex kicks in at the end to save just them, neither Burt nor the narrator care in the slightest to help anyone they'd just spent months or years living with in the Terrafoam housing projects. Missing any old friends? Nope, not in the slightest! And in their new world no one seems to clearly understand the mechanisms of the new prosperity nor do they seek to use that prosperity to help those with whom they should notionally be able to empathize.

I fear that the author actually believes this stereotyped and non-reflective content is not humorous as propaganda but is somehow actually inspiring and visionary. This hypothesis would suggest that the author is kind of like Ayn Rand, except less skilled.

On the other hand, the author might just be playing out this game with a straight face, like Jonathan Swift where the idea is to aim high and hope some people get it rather aiming low in the fear that someone might not. However the message I get when I interpret the story this way involves breathtaking levels of misanthropy and hopelessness... which is somewhat inconsistent with high artistic aspirations.

Did I miss something obvious?

I will have you know that I got out of my swimming pool and did my part to support the local economy by getting a $5 massage.

Make poverty history!!!

I think it's a genre convention of utopian fiction -- take an observer from the mundane world (which may be a crapsack, and plant them in the midst of the wonders of Utopia. For me, given the strong resemblance of the Australia Project to the Culture, it's impossible to imagine that they don't have their equivalent of Contact (and Special Circumstances), but that the narrator never was introduced to them. I lean towards the Author Failure explanation, though I don't think it's actually possible to be less skilled than Ayn Rand.

I don't think it's actually possible to be less skilled than Ayn Rand.

It's odd; I've heard these kinds of swipes a lot, and yet I really like Rand's fiction. It raises an interesting question: do tastes in literature vary so much such that whether someone is a great writer or unskilled (!) is wholly in the eye of the beholder? Or is there an anti-halo effect going on, where people think, "Rand's philosophy is wrong, therefore she must be bad at writing, too"?

Eh, her main literary flaws are the Author Filibuster and the use of Strawman Political villains and Mary Sue heroes. The definitive takedown was by Whittaker Chambers in the National Review in 1957.

Of course, other writers surely have written worse books than Atlas Shrugged, and not been so universally slagged, so there may be an anti-halo effect going on. That doesn't change the fact that Atlas Shrugged is terribly written.

That doesn't change the fact that Atlas Shrugged is terribly written.

Can we reduce terribly written into testable empirical statements? Without necessarily calling them flaws, I agree about the use of author filibusters, straw villains, and Sues. I also agree that the philosophy is wrong, and that many people hated the book, including Whittaker Chambers.

All this having been said, I expect you will agree that writing a thousand-page novel that sells millions of copies is a rare feat that requires no small amount of writing skill. I find it hard to believe that millions of people would buy a book for no reason whatsoever. If the claim is merely that Rand fans have bad taste and questionable morals, then we do not really disagree in the rationalist's sense; I can only shrug and say, "De gustibus non est disputandum." I merely wish to emphasize that even catering to people with bad taste and questionable morals requires what we would call writing skill; not everyone can do it.

I can only shrug and say, "De gustibus non est disputandum"

Beware this piece of cached wisdom.

I merely wish to emphasize that even catering to people with bad taste and questionable morals requires what we would call writing skill; not everyone can do it

True, but there may be a low-hanging-fruit effect: it could be that Rand's literary skill is much less of a factor in her success than the fact that she was in the right place at the right time to meet the thitherto-unmet needs of a certain audience.

definitive takedown

This is the review that says (page 3):

From almost any page of Atlas Shrugged, a voice can be heard, from painful necessity, commanding: "To a gas chamber -- go!"

On a related note, here is the economist Nick Rowe's recent excellent post about what will happen if most human labor can be automated fully and cheaply at some point in the future:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/01/robots-slaves-horses-and-malthus.html

It's by far the best and clearest analysis of the issue I've seen so far. In case anyone is interested, I left a few comments in the discussion there, which I'd say resulted in some additional insight.

My comment there:

Assume robots are the same as humans.

A situation which is very unlikely in the first place - and extremely unlikely to last for very long if it somehow magically happened. Machines already vastly exceed human capabilities in many areas.

This essay stops before things get interesting. What will politicians do if they have a big mountain of unemployed human voters to feed?

What will politicians do if they have a big mountain of unemployed human voters to feed?

In the given Roboslave world, the government would run enough roboslave farms to produce food to feed them. And possibly, enough to feed everyone.

Wasn't that the point of slaves and machines? That they work so we don't have to?

(and at what point in the last tens of thousands of years did feeding yourself stop being your problem and start being your representative politician's problem?)

The usual problem with the unemployed humans is where the money to support them comes from. The usual answer is taxation of the robot companies. For that to work very well, there had better not be too many tax havens - and there had better not be too much of a "race to the botttom" between governments to host (and tax) the companies. These requirements seem moderately taxing.

Feeding unemployed humans is the government's problem in my country - and in many other countries with a welfare state. The more unemployed humans there are, the more likely they are to vote for a welfare state.

The usual problem with the unemployed humans is where the money to support them comes from

Yes, but that's only a problem in that you have to tax rich people to get their money to distribute it. My first reply was therefore "the government will tax the rich to feed the poor, what else could they do?".

But after a bit of thought, I realised that if the government owns self replicating slave robots and land, then it can use the slaves to create food without needing to tax anyone. The robots can't be taxed because their earnings go to their owners, but in this case they aren't earning anything because they give the food away. Their efforts don't take any human input so nobody needs a salary, and they can do what people can do so they can run farms and distribute food.

So, without comment on the morality of slaving human-equivalent robots (whether the robots care isn't discussed in the link), feeding unemployed people is a non-problem - the self running Roboslave farms are free food fountains.

They are borderline cornucopia machines limited to whatever humans can make and the right resources being available - in this case sunlight, land and seeds.

Yes, a benevelont socialist government could decide to feed the humans.

One problem then would economic competition between the governments with the human resource drain and the ones without.

However, to get to that point in the first place, some major government changes would be needed - in many capitalist countries.

My biggest problem with this story is that Jacob, the narrator, has no assets to sustain himself after he loses his job. He doesn't have funds in his bank account, he doesn't own anything that isn't rented, and when the robot asks him if he has any means of support unknown to the system, he replies "no".

Well, why the fuck not? He has been working as an administrator for 20 years, on top of his previous work as a teacher and as a fast food employee. What in God's name has he been doing with all his money? The cost of living can't possibly have increased when the whole point of Manna and the robots is that they are more efficient than human beings, hence why they replace them. Even land should get cheaper once people start being shipped off to the Terrafoam projects, which we know from Burt's case was already happening 10 years prior to the date Jacob gets fired. It should be cheaper to live in the 2050 the story presents that it is right now.

I know people often make financially stupid decisions because humans are not automatically strategic and because a lot of their choices are motivated in large part by status rather than by solely fiscal considerations, but this is too much to be plausible. Jacob has known for at least 10 year that people who can't sustain themselves end up on welfare and get shipped to Terrafoam, yet he hasn't been keeping a savings account, a retirement plan, an investment portfolio, or even a property of his own? What did he do, blow through his entire paycheck every month by living like a millionaire and renting stuff he couldn't afford to buy until the day he inevitably lost his job and got corralled into the lowest rung of society? Is that the fashionable thing to do in 2050? And very fashionable it must have been indeed, since the story implies that everyone from the poor to the middle class, except for a relatively small number of "rich", have all ended up on the projects.

It's a good contrast between two types of post-scarcity economies: one where a capitalist status quo still applies (the dystopia), and one where material goods end up like open-source software does today (the utopia).

I think that's the big takeaway - you shouldn't eliminate labor scarcity without deploying some kind of new economic distribution model.

"Post-scarcity economy" is an impossible concept even in principle (assuming a human society, at any rate). Beyond a certain minimal standard of living, which is in fact quite low by today's standards, most of the things people really care about are zero-sum. They will struggle just as mightily and eagerly to wrest those for themselves no matter how cheap, plentiful, and high-quality non-zero-sum stuff gets. Moreover, habitable land is always zero-sum, which further complicates things.

For those with less ambition and/or ability, this has a twofold effect. On the one side, it benefits them because in a decently functioning polity, people's efforts to get ahead in life in the hope of zero-sum gains will result in a growing economy making non-zero-sum stuff increasingly cheap, plentiful, and high-quality. On the other hand, it is bad because some things are zero-sum but nevertheless essential for life, most notably habitable land, and people's struggle to get ahead in zero-sum efforts drives the price of these way up. This makes it necessary to work almost as hard as the most ambitious and prosperous folk to be able to afford the increasing price of lodging and other essential access to zero-sum things.

Thus in the contemporary developed world we have a situation where nobody is in danger of starving or not having warm enough clothes, but homelessness is a very realistic threat for poorer people. (As an even more striking illustration of the same phenomenon, in recent years even cell phones and computers have become affordable to many of the homeless folk.) The Manna story also illustrates the same principle accurately: the state finds feeding and clothing the unemployable masses adequately a trivial expense, but the land to house them is expensive and scarce so they have to live packed together like sardines.

"Post-scarcity economy" is an impossible concept even in principle (assuming a human society, at any rate).

How about a totalitarian government with high technology and fertility management?

It doesn't matter. Whatever the institutions of this totalitarian government might look like, there will still be the usual human zero-sum struggle for status, and whatever goods (material or not) are necessary to gain status will be the object of this struggle, just like they are now. Even under the assumption (entirely impossible for a human society) that material zero-sum goods like land are distributed strictly equitably and the population is kept low enough that the amount per capita is large, status is still scarce by definition.

I don't think "post-scarcity economy" is intended to be to do with status.

The word "economy" suggests material resources to me - space/time/matter stuff.

In practice, all intelligent writers of post-scarcity economies realize that status will still be scarce. Doctorow has his post-scarcity economy run almost solely on status-units (whuffies, or whatever); and I think there is a line in Bank's Culture novels somewhere to the effect that 'the Minds could give humans everything they wanted - except to matter'.

[-][anonymous]13y40

I think its more a way to get those who talk about a post-scarcity economy to realize that people don't really care that much about material resources. Its surprising but people do indeed need to be reminded that social animal's lives are about status.

I continue to find your comments really stimulating.

Spoiler alert!

It is amusing that someone espousing the principle "Nothing is owned" would have his hero use the words "my" and "mine" so many times in the penultimate paragraph. And that the plot twist that saves the hero's life is the fact that his parent had purchased some shares of stock several years before.

All in all, it was a weirdly disjointed mix of physical economic realism and rhetorical economic silliness. "Everything is free" we are told in one paragraph. "Everyone receives credits which they can use to buy things", it is explained in the next paragraph. Presumably our hero will be able to pass on his "share" to his descendents, just as his father did. But what happens to children born in Australia without shares? Do they get shipped to America to be warehoused? Population control issues are never mentioned.

People who own robot capital may well choose to live lives of leisure, as they do in this story. People who do not own robots will have to compete with robots in the labor market. At what kinds of jobs will they have comparative advantage? Well, clearly the owners of robots are unlikely to risk losing their capital by assigning their valuable property to tasks in which the robots face the probability of being blown up. So, it is likely that the fate of the poor will be to serve as soldiers in the various armies of the future - whether revolutionary or counter-revolutionary, imperialist or anti-imperialist. Cannon fodder. Except in peace time when they may be kept docile with bread and circuses. If there is much peace time.

[-][anonymous]13y60

Well, clearly the owners of robots are unlikely to risk losing their capital by assigning their valuable property to tasks in which the robots face the probability of being blown up. So, it is likely that the fate of the poor will be to serve as soldiers in the various armies of the future - whether revolutionary or counter-revolutionary, imperialist or anti-imperialist. Cannon fodder. Except in peace time when they may be kept docile with bread and circuses. If there is much peace time.

I have no idea why you think that the cost of feeding a human soldier, even if you sent him to the battlefield naked and unarmed to fight other waves of human flesh, would be lower than deploying a drone.

Population reduction tempered by charity efforts by the robot owning humans seems more probable. Also over time the majority of non-robot owning humans might be part of feral populations if there is any natural habitat left.

"Those are the space elevators," she said, "You can ride them if you want. They are just starting to be fully operational.

Huh. I guess 4GC annexed Indonesia at some point then...

There is a serious lack of economic logic in the story. Nothing like "about half" of American workers work in environments routinized enough to use something like manna.

Also getting instructions every minute would be extremely unlikely. The employees would hate it, so you would have to raise pay or you would lose workers at the margins.

Many office employees get advice and guidance from machines with high frequency. However, they usually come through screens - not earpieces. Earpieces are an inferior message delivery system for most kinds of machine-originated messages - through being too invasive, and having poor bandwidth - compared to vision.

Earpieces are an inferior message delivery system for most kinds of message - through being too invasive, and having poor bandwidth - compared to vision.

Indeed. In fact, the main advantage earpieces have over visual systems when it comes to the delivery of artificial information sources is the fact that it frees up the visual system for other purposes. (Like steering, for example.)

Indeed. In fact, the main advantage earpieces have over visual systems when it comes to the delivery of artificial information sources is the fact that it frees up the visual system for other purposes.

If we're imagining the very cheapest employees in a dystopian future burger joint, ability to read English might rule out a lot of foreign/older (vision)/uneducated/dyslexic people. People? I mean, human resources.

I've worked with people who would rapidly become comfortable with minute-by-minute instructions... though they might complain about it a lot, for a variety of essentially signaling reasons.

In general, my rule of thumb about giving detailed instructions to people is that it's a bad idea, not because they hate it, but because they become accustomed to it. If I solve the "what should I do next?" problem for someone enough times, they start expecting me to solve it for them in the future as well. I have a hard enough time solving it for myself.

//Not an economist//

The minimum wage creates a class of people who it isn't worth hiring (their productivity is less than their cost of employment). If you have a device which raises the productivity of these guys, they can enter the workforce at minimum wage.

Additionally, there may be zero marginal product workers - workers whose cost of employment equals the marginal increase in productivity that results from hiring them. This could happen in a contracting job market if the fear of losing employment causes other workers to increase their productivity enough. Then you could fire Jack and see the productivity of John increase enough to match the productivity net of costs that Jack provided. If such workers exist, then they can provide a new source of labour even in the absence of minimum wage laws.

I agree with you that there's a lack of economic logic in the story, though.

Well, as labor becomes less scarce, you don't have to care if your employees hate it or not. Enough people will be desperate enough for some sort of work that you'll be able to find people.

It would only catch on if it could survive the initial phase where it was still a minority. If anyone using it was prepared to quit and try for another job then businesses not using it would out-compete businesses using it. It might not have to be enjoyable, but if it wasn't at least tolerable it wouldn't even get off the ground.

a) In that case, in the story it did get off the ground so it must have been tolerable.

b) Criticizing the frequency of instructions is a pretty weak argument against the story plot since it's not critical to the plot. If 1 minute instructions bothers you a lot, imagine the story says they are 2 or 5 minute instructions, or a wristwatch screen with your next instruction on it.

I thought that was the implication. Initially, Manna was actually an improvement on a human.

Nothing like "about half" of American workers work in environments routinized enough to use something like manna.

Do you have anything to support this? It seems to me that organizational charts are usually pyramid shaped, and the lower parts of the pyramids are where the most employees are, where less decision-making jobs are and where such a system would be more useful - and so a lot more employed people are doing jobs which a less skilled / lower cost person could do with guidance than are doing jobs with tough decisions and responsibility or creativity.

Also, that a lot of people have factory / processing jobs which are right now such fixed routines that a manna-like system would skip over them.

The employees would hate it

The point of the system is that the fast food restaurant is able to pay people less and guide them to doing a better job than was done before. That is, people who couldn't get a fast food job before and were unemployable can earn money. Those fictional unemployable people who can now eat are hardly going to complain that their job isn't as much fun as they want, are they?

Manna is a fun story.

The bit where the machines control the humans reminds me of my Automating body and mind video. I actually cite Marshall Brain in that one, two and a half minutes in.