Bryan Caplan  

The Age of Em: Reply to Robin

Reply to Jerry O'Driscoll... 3 Questions for Robin...

Here's my reply to Robin's reply to my criticisms of The Age of Em.  I'm the main text; he's in first-level blockquotes; I'm in second-level blockquotes; he's in third-level blockquotes.

1. Robin only pretends to dodge philosophy of mind. .. He tacitly accepts an extreme version of "Ems are just as human as you or me" - and builds the whole book on this assumption. The tell-tale sign: The Age of Em says vanishingly little about the lives of biological humans during the Age of Em! .. he seems so wedded to this philosophical (not social scientific) position that he can't even feign agnosticism. What would feigning look like? Split the book evenly between discussion of the lives of biological humans during the Age of Em and ems during the Age of Em.

These are complaints about language and emphasis. On language, since I'm constantly applying human and social sciences to ems it would have been quite awkward to use any other than our usual language for describing people. It is hard to imagine a readable book where words like "people" are constantly replaced with phrases like "machines that act like humans but do remember I'm not making any philosophical claims here." On emphasis, very little happens to biological humans during the em era, so equal emphasis would mean a very short book. That conflicts with my goal of showing just how much I can say about this scenario.

I'm not advocating awkward language.  My point is that Robin barely discusses what is, by normal standards, the most important aspect of the Age of Em: The lives of biological human beings.  My explanation: Despite his claims of agnosticism, Robin thinks biological humans will be unworthy of his interest once trillions of ems exist.  This makes sense if ems are literally conscious.  Otherwise, not.

Robin's explanation - that "very little happens to biological humans during the em era" - is bizarre.  On his own account, biological humans become fabulously wealthy people of leisure almost overnight.  That's a big deal in itself, with far-reaching social and political implications.  How happy will humans be?  How many kids will they have?  How will status games change?  What will happy to partisanship?  Religion?   

3. Robin has a bizarre definition of "marginalized" .. biological humans .. They'll be outnumbered, and perform little "hands-on" work. But they'll be fabulously wealthy and ultimately in charge." .. 5. Robin's conclusions only sound "taboo" because he's using language strangely. .. Robots will "dominate" us no more than rank-and-file workers "dominate" shareholders.

As I mentioned in a previous post, many have reacted to talks I've given by complaining about humans no longer being at the center of action, even when they understand that biological humans could for a while own most of the em world, and thus direct how spare resources are spent. I used words like those people use to acknowledge their concerns.

In Robin's scenario, such concerns are silly.  Does it make sense to feel bad for GM's majority shareholders because they don't personally assemble cars? 

4. Contrary to Robin's suggestion, there's near-zero correlation between income and conservatism. .. "subsistence farmers tend to have values more like those of poor/conservative people today." .. Robin could say he's defining "conservatism" in a technical or apolitical way. But when you're writing for an audience, the author rightly bears the burden of highlighting non-standard usage.

In that context I had just cited studies on strong correlations between the culture and wealth of nations, and I had just explained in quite some detail the kind of "conservatism" I meant there. It is true that I didn't mention explicitly there that the word "conservatism" is used in many different ways. But the book would be a lot more tedious if every time I introduced and used a term I explained the many other ways people have used that term.

My complaint is not just that "conservatism" is used in many ways, and Robin picks one.  My complaint is that "conservatism" is primarily used in one way that isn't Robin's way, leading to confusion.  This is a symptom of the unfortunate diary-like style of The Age of Em - the fact that you have to pre-understand Hansonian thought in great detail to grasp what he's saying. 

7. Robin's efforts to calm readers' fear of the future consistently backfire. Example:

Readers of this book may find near subsistence wages to be a strange and perhaps scary prospect. So it is worth remembering that such wages in effect applied to almost all animals who ever lived, to almost all humans before a few hundred years ago, and for a billion humans still today. Historically, it is by far the usual case.

Imagine a middle-class American's child is destined to earn a subsistence wage. Would it make the parent feel better to hear, "No big deal, your child will face the same fate as almost every animal who ever lived, almost all humans before a few hundred years ago, and a billion humans today"? No, even worse!

Saying something is "worth remembering" just means that it may change how you think about that thing; it is not the same as saying "don't worry." It isn't my job to make readers like the age of em, but it is my job to make sure they keep important considerations in mind.

This looks like motte-and-bailey to me.  Robin routinely tries to paint the Age of Em in a favorable light.  Here's one memorable instance.  When you point out that his arguments are unconvincing, he protests he's merely trying to describe the future accurately, however awful it may be.  A few days latter, he resumes his advocacy.

10. Robin's argument against the Terminator scenario is much weaker than it looks. His words:

A reasonable hope is that ordinary humans become the retirees of this new world. .. ems may be reluctant to expropriate or exterminate ordinary humans if ems rely on the same or closely interconnected legal, financial, and political systems as humans, and if ems retain many direct social ties to ordinary humans.

The problem: As Robin explains, in one human year, ems experience millennia. So even if each generation of ems only has a .5% chance of expropriating humanity, the chance of expropriation per human year is around 40%.

Bryan misreads me as trying to offer more reassurance than I can. I was clear that even if humans survive the year or two that comprises the age of em, I can say little about what might happen after that. "A reasonable hope" is quite different from "don't worry." I would be remiss if I didn't at least point readers in the direction of a reasonable hope, even if I can offer few guarantees.

Never mind "guarantees."  My argument above implies biological humans are likely to be wiped out one year into the Age of Em as Robin describes it.  Is my argument wrong?  If so, why? 

Bryan's last two objections, on economics, are the ones I take most seriously.

8. Robin greatly overstates the quality of life for ems. .. Why wouldn't ems' creators use the threat of `physical hunger, exhaustion, pain, sickness, grime, hard labor, or sudden unexpected death' to motivate the ems? Robin elsewhere talks about `torturing' ems, so why not?" .. Modern systems of slave labor - see Stalinist Russia and Nazi Germany - used pain freely, because the penalty for quitting was death.

I was careful not to claim that ems would not be slaves. I just suggested that we didn't have good reasons to expect most being slaves. Most people in history haven't been slaves, even when competition has been strong. There are still some slaves in the world today, and they aren't known for being spectacularly productive workers due to frequent use of torture and pain. Nor were Stalinist Russia and Nazi Germany slaves known for stellar productivity. We have a larger literature on motivating workers, and threats of pain only seem to be useful for the most routine and physical of labor.

As I said, this large literature focuses on motivating workers who are free to quit.  If workers aren't free to quit, terror is an effective motivator - even for complex tasks.  Again, read the history of the Soviet nuclear program.  Stalin's top scientists worked in the shadow of death.  Since they couldn't flee, they worked like dogs to give him nuclear weapons, and reached their goal rapidly.  Just one example, but a powerful one nonetheless.

I agree Soviet and Nazi slaves' productivity was normally low, but the reason is simple: Their labor camps did not prioritize productivity; their main aim was crushing hated enemies, not maximizing output.

Historic slave systems did often augment negative incentives (torture, death) with positive incentives (better treatment, cash).  But there's a simple economic story: When slave-owners have imperfect information about slaves' productivity, high quotas lead to lots of counter-productive punishments.  Threatening to execute everyone who falls below the 90th-percentile of output, for example, requires slave-owners to kill 90% of their slaves.  Information about ems' productivity, however, should be much more accurate, especially since most descend from a small number of exceptional humans.  These are ideal conditions for heavy use of negative incentives.

When you seek workers to be creative, think carefully, take the initiative, or persuade and inspire others, you mostly seek other motivations. Our best analogy for ems should be the few hundred most productive people in the world today, and threats of pain are not remotely what motivate them. Who thinks torture would make them more productive overall in the long run?

If they're not free to quit, I think exactly that. 

9. Robin's arguments for his single craziest claim - global GDP will double every "month, week, day, or even faster" - are astoundingly weak. .. In the real world, however, there are literally hundreds of bottlenecks that radically retard this kind of growth. Politically, something as simple as zoning could do the trick. .. the most favorable political environments on earth still have plenty of regulatory hurdles .. we should expect bottlenecks for key natural resources, location, and so on. .. This alleged "concrete clue" is nothing compared to the bona fide "concrete clue" that almost all fantastic claims are false. And the idea that the global economy will start doubling on a monthly basis is fantastically fantastic. This has to be the least Bayesian part of the book: We start with a claim with a near-zero prior probability, make a couple of flimsy arguments, and somehow end up with a high posterior probability.

One could have similarly argued that fundamental growth bottlenecks must prevent the previous observed huge jumps in growth rates, such as from foraging to farming, or farming to industry. And plausibly related obstacles did prevent those eras from starting as soon as they might have. But eventually obstacles were overcome. No doubt our current economy tolerates many delays that would have to be cut to enable much faster growth, and the em economy won't start as early as it might because of regulatory and other delays. My book is mainly about what happens once those obstacles are overcome. Does Bryan really think such obstacles could never be overcome?

To make global GDP double every month, you don't have to overcome some bottlenecks.  You have to overcome an accelerating series of bottlenecks.  The bigger and faster the changes you seek, the more obstacles you meet. 

Even when doing so might quickly allow a city or nation to dominate the world? His "near-zero prior" seems to come not from any fundamental analysis but, from his strong reliance on intuition;

Whatever you call it, I'm exercising common-sense skepticism.  When someone predicts huge changes, I scoff unless they have overwhelming evidence in their favor.  So should we all.

I suspect he would have similarly assigned a very low prior to manned flight in 1850, or to space flight in 1900.

Technically, there was manned flight in 1850.  More to the point, there's a world of difference between predicting specific technological advances, and claiming they'll quickly and constructively revolutionize society.  I can believe there's a 1% chance ems will emerge in a century.  That's not crazy.  But it is crazy to think the emergence of ems will lead global GDP to start doubling on a monthly basis.  For that, a conditional probability of one-in-a-million is generous.

Robin may protest he's simply applying standard growth theory to a novel situation.  A better description is that he's mechanically applying standard growth theory to an unprecedented situation the model was never designed to handle.

But as I said above, I expect many others agree with his intuition, and I thank him for saying explicitly what others only think.

Not just "many others."  I predict at least 95% of empirical growth economists would agree with me.  Indeed, I'd be surprised if Robin could find any empirical growth economist with no prior affiliation with futurism or science fiction who'd view his conditional growth prediction as plausible.

Robin has helped me more than anyone else to internalize Bayesian thinking.  I'm flabbergasted, then, at how un-Bayesian his growth predictions are.  If "Technology X will cause global GDP to double" doesn't deserve an extraordinarily low prior probability, what does?  And what evidence has Robin produced to justify raising that prior above the microscopic level?

COMMENTS (6 to date)
michael pettengill writes:
To make global GDP double every month, you don't have to overcome some bottlenecks. You have to overcome an accelerating series of bottlenecks.

You need to double labor costs every month.

You can pay the robots, but the robots must spend what they earn. Or else tax the robots at 100% and pay welfare to the humans so they buy gdp.

Producing even $1 of gdp per year with no one buying it for $1 results at some point in zero gdp. What is the point of just filling a warehouse.

Bill Woolsey writes:

I think I understand the long run Malthusian equilibrium. The fundamental cause is the biological drive to reproduce. Sexual desire leads to rapid population growth and a growing labor supply that outstrips any increase in labor demand. Real wages drop to the point where parents can only afford raise enough children to replace themselves.

The population is then stable at these low wages. The return on capital is just sufficient to replace the existing capital stock. The landowning class, however, lead lives of plenty in a crowded world. (That can be modified too so that each landowner earns just a little. Or the only people who can survive are those who work as much as possible and own a tiny bit of land.)

This theory does not seem to apply in a world with birth control. Humans still have a desire for families, but not necessarily such large ones that population and labor supply grow so fast and to drive down real wages to subsistence.

While the cost of producing more EM's is zero, what is the motivation, the benefit, of greatly increasing their supply? They don't appear to be produced as a side effect of some pleasurable activity by existing EMs.

If we think of EMs as simply "free" software to run hardware, it would be irrational to produce hardware just to be able to utilize more free software. You produce the hardware for some purpose, and then create as much software as needed to operate it. That the software may be copied at appreciably no cost just means that is not a constraint.

Somehow, it is these rank and file EMs consciousness that results in the irrational result of using appreciable all the resources of society to construct hardware to house more EMs.

It seems to me that there is a massive market failure if the unintended consequence of EMs doing what they choose is a society where all resources are aimed at maximizing the number of EMs.

If all people want to have as many children as possible, that would result in an economy that supports the maximum population. I don't see that as irrational, exactly, though it does seem that in the long run we end up with lots of people each of which lives in a small family and very poor.

But why do EMs want there to be as many EMs as possible?

The the biological people are in charge, why would we let resources be used just to increase the total quantity of EMs?

One way to understand Marx's critique of capitalism is that more and more capital goods are be used to produce more and more capital goods. This is what the EM economy appears to be. I don't think a market economy works like Marx thought. What is it about the EM economy that leads to it to be that way?

Mark Bahner writes:

"While the cost of producing more EM's is zero, what is the motivation, the benefit, of greatly increasing their supply?"

How many people in the world are working on lowering the cost of photovoltaics per kWh produced? How many people in the world are working on lowering the cost of batteries per kWh stored?

Now, suppose you raised the number of people working on those problems by 1 thousand, 1 million, or 1 billion. Don't you think that would hugely speed up progress on those problems?

And what about the number of people working on fusion? Curing cancer? Artificial organs? Nanobots for medicine? Mining asteroids? What if the number of people working on all those things was increased by a factor of 1 million or 1 billion? Wouldn't that produce pretty spectacular progress in all those areas?

John T Kennedy writes:

" have to pre-understand Hansonian thought in great detail to grasp what he's saying."

That is a big problem. I often find it very difficult to unpack a FB post from Hanson is ultimately intended to convey something pretty simple.

I appreciate that I don't have to understand the thought of Caplan, Huemer, or Scott Alexander in great detail to understand what they're saying at any given point.

[Broken url removed. Please check your links while previewing.--Econlib Ed.]

Peter Gerdes writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Mark Bahner writes:
If "Technology X will cause global GDP to double" doesn't deserve an extraordinarily low prior probability, what does? And what evidence has Robin produced to justify raising that prior above the microscopic level?

It depends what "technology X" is.

If "technology X" is "a better mousetrap" it does deserve an extraordinarily low prior probability of doubling global GDP.

But what if it's "a computer with power comparable to a human brain for a few thousand dollars"?

Comments for this entry have been closed
Return to top