*The Age of Em*, by Robin Hanson

That is the title of the new and forthcoming Robin Hanson book, due out in May.  I was asked to supply a blurb, and offered two possibilities.  One was:

“Robin Hanson is one of our most original and important thinkers.  This is his book.”

The ostensible premise of the book is that people have become computer uploads, and we have an entirely new society to think about: how it works, what problems it has, and how it evolves.  One key point about this new world is individuals can be copied.  But this is more than just a nerdy tech book, it is also:

  1. Straussian commentary on the world we actually live in.  We are already something-or-other, uploaded into humans,and very often Robin is describing our world in cloaked fashion, albeit with some slight tweaks to parameters for the purpose of moral illumination.
  2. A reminder of how strange everything is, and how we use self-deception to come to terms with that strangeness.  It’s a mock of all those who believe in individual free will.
  3. An attempt to construct a fully rational theology, proving by various deductions that God is not fully benevolent in the traditional sense.
  4. An extended essay on the impossibility of avoiding theology, given the imposition of competitive constraints on a world where production and copying are possible.  And ultimately it is a theodicy, though it will not feel that way to Westerners, Jews, Christians, or Muslims.  It hearkens back to medieval theology, Descartes, and the idea of living in God’s possibly terrifying simulation.
  5. A satire on the rest of social science, and how we try to explain and predict the future.
  6. A meta-level growth model in which energy alone matters and the “fixed factor” assumptions of other models are relativized.  Copying is taken seriously, besides how special are you anyway?  In the meantime, we learn just how much of the world we know depends upon the presence of various fixed factors.  But surely that is temporary!
  7. A challenge to our notions of wherein the true value of a life resides.

I hope enough readers pick up on some of this.  And yes, there is a chapter on sex, love, and affairs.

It is hard to excerpt from this book, but here is one short bit:

Compared with humans, ems fear much less the death of the particular copy that they now are.  Ems instead fear “mind theft,” that is, the theft of a copy of their mental state. Such a theft is both a threat to the economic order, and a plausible route to personal destitution or torture.  While a few ems offer themselves as open source and free to copy, most ems work hard to prevent mind theft.  Most long-distance physical travel is “beam me up” electronic travel, but done carefully to prevent mind theft.

I am wildly enthusiastic about everything the Robin upload does, and some of his copies are better yet.  Here is the book’s home page.


Addendum: Here is Robin Hanson’s response.


Will non-readers of your blog understand what "Straussian" means? And if the point is to popularize this term, shouldn't you have plugged "mood affiliation" too?

I'm a reader and don't understand what Straussian means.

Has Tyler Cowen actually read any of Strauss's books? Or did he just skim Wikipedia and some articles on Strauss during the Bush administration?

Wait, you are calling Prof. Cowen a poseur, and not one of our most original and important thinkers? No blurb from you for Prof. Cowen's next book.

There are a lot of phonies on Internet and in academic life but few are as smart as Tyler Cowen. He is a phony with distinction. He may even flipped a few pages of Strauss’s books.

I think non-readers of this blog, if they stick around for more than a month, are going to see the following sequence:
1. Tyler uses "Straussian" in a sentence.
2. Somebody comments, essentially, "Straussian, WTF."
3. Hand-wringing by the usual suspects occurs.
4. The new reader googles a bit and figures it out.

You forgot at least one:

2.5 prior_test posts something rude and inaccurate about the hosts of the blog he reads faithfully every single day

Chris, I honestly have tried to understand, but how Straussianism applies to many of these type statements I just don't "get." A perspective independent of culture, or even that our current culture clouds reality more than previous cultures.... Should I read Straussian as "here's how the clearest thinkers from generations ago would've thought about this"?

I'm a non-reader and Wikipedia understands what "Straussian" means

After moving a couple of tons of snow yesterday and today so a few cars (not mine) can get out and go somewhere, computer uploads and downloads strike me as rather useless in the real world.

Or shall I say, computer downloads are worth the nothing I pay because the snow they can move totals nothing. And Uber is worth nothing if it can't clear the roads for the private startups who provide the transportation Uber loudly states it does not provide, but takes a rent of 10% of the labor and capital payments that are impeded by the snow clogged highways and byways..

The concept here is that a copy of a person's mental state is downloaded into a general purpose robot, who can then competently shovel your driveway.

Or he can just call a properly licensed cab, who will come shovel his drive.

Extremely well played.

A copy of a person's mental state is downloaded into a general purpose robot, who is not currently buried by snow.

The snow will melt eventually, why shovel if there is no need for physical transportation.

Or you just copy your mental state to a robot at the destination you want to go. Why bother with physical transportation? That is for matter.

"After moving a couple of tons of snow yesterday and today so a few cars (not mine) can get out and go somewhere, computer uploads and downloads strike me as rather useless in the real world."

In 10-20 years, every snow plow east of the Mississippi (and many west of the Mississippi) will be able to autonomously drive to locations where there has been unusually heavy snow. Even if the snow plows can't plow autonomously, they will at least be available for human operators. If this is the case, human operators can fly in from anywhere (e.g., Canada) and operate the snow plows. This will allow much quicker cleanups in the event of unusually heavy snows.

Prozac. Or some exercise and sunshine. Maybe a change of job.

My spouse has posited that canadian literature would disappear as a genre with wide distribution of anti-depresssnts.

The brain is not just connected to itself, it's connected to the outside world through our 5 senses. It's inside a body that moves through space and interacts with the physical world, . so what are we downloading exactly? a simulation of a partial neocortex where conscious awareness is thought to exist or are we including the cerebellum, brainstem, hindbrain etc.. If those are included , there is a magical connection to them, from appropriate sensors,in the robot ? will the robot em need to sleep or is that superfluous ? will learning be instantaneous ( more upload) or laborious like today? Will the em have the human emotions ( fear, jealousy etc..), what would stop an em ( in theory immortal) to get stuck in a pleasure loop and stagnate there , happy as a lark. self awareness is not enough we are also goal driven, What goals do ems have? what motivates them to survive exact;y ?will wait for the book

those senses are connected via electrical impulses in the nervous system. very easy to simulate, its not magic. we already have auditory replacement connected to nerves.
capturing the state of all the neurons and connections in the brain is the magic part of this fairy tale.

But they are random-ish in nature and prompt responses. Most of them are unconscious but nonetheless affect our mental state. Not getting enough sunlight, and thus vitamin D? You may feel sluggish, which may prompt you to reflect on laziness and maybe write a brilliant essay about it later.

And some of the response are conscious but lead to places you can't predict.

"It smells bad here. Better cover my nose. That's better. Boy, that smelled worse than Stinky Tofu. Gee, it's been a while since I had tofu... I wonder if there are any new Chinese places on Yelp... Eat Drink Man Woman is a good movie, I wonder what Ang Lee has made lately... And what should the US policy toward Taiwan be given that we are all cyborgs now?..."

From what I've read of Hanson, the idea is that it's a world of Malthusian scarcity where all matter has been transformed into maximally efficient computing hardware, and that hardware is fully populated by EMs. So basically the reason not to get stuck in a pleasure loop is that you have to be able to afford to pay 'rent' on the processing power it takes to continue simulating your consciousness.

In terms of the psychology of EMs, the idea is that they start out exactly like humans, as people literally have their brains scanned and uploaded copies of them created in a computer. From there they will 'evolve' based on Hanson's thoughts on what pressures and incentives there would be for them in the virtual world. The biggest thing selected for if of course being down to sink all the money you make into producing more copies of yourself, which then make more money, and thus more copies, until the market for the labor that EM can do is saturated to such a degree that the marginal additional copy is at a subsistence level of being able to pay for its own existence, but not save up to put anything towards producing new copies.

“Robin Hanson is one of our most original and important thinkers. This is his book.”

Truly brilliant. And there is no question why the second blurb was the one likely used - such an ostensibly sly insult is pretty hard to get past an editor. Though it is always possible that the highlighted blurb was meant in good faith, Prof. Cowen being one of our most original and important thinkers, and that was his blurb.

Maybe I'm just being slow, why is the quoted blurb an insult? I read it as just a super-simple statement of fact: there really are few thinkers as original and important as Hanson. He's written a book. If you like original and important thought, it's a good book to buy.

prior_test is just a troll who spends an inordinate amount of time making snide comments towards Tyler Cowen and other objects of obsession.

You must be new here. The crap prior_test vomits up here isn't worth engaging.

I don't see why the ems would remain recognizably human. The first thing virtual me does is play around with various simulated bodies, human and animal (be a dolphin, or an eagle, etc.) Live in fantasy worlds from books, etc. Other ems are probably building virtual worlds as entertainment.

The second thing is to start altering its own personality, to get rid of traits it doesn't like, enhance the ones it does. Next would be adding to its virtual neural tissue. Can it have different talents, increased intelligence? Assuming the brain is that well understood, can it mix in others memories? Is virtual telepathy with other ems possible?

Advance this by a few hundred simulated years, and what is the result? Is it sane? Does it want to work with/for biological humanity? I kind of doubt ems are stable.

I also think that long before you have ems, you have "problem solving engines" and very capable robots that change the world pretty dramatically, without being in any way human.

One of the more interesting openings to a story in Varley's Eight World series (The Ophiuchi Hotline being the notable novel) makes that point in a truly prosaic fashion. Humanity has been provided technology to switch one's mind into another body - though in that almost 4 decade outdated work of SF, the uploading is not virtual, but into a true body.

And when a man is transferred to a female body, the very first he does is feel his new breasts. Somehow, I doubt Hanson is capable of providing such a basic illustration of one of the more basic responses to being not only in a new environment, but in what could be considered a differently formed universe as experienced through a differently formed body.


I've read at least one of those books -- the one that starts with a salesman saying "ladies and gentlemen, the penis is obsolete!"

Thanks for the laugh! (I mean that seriously)

Most technologies seem to go this way. People start thinking aircraft will be just like birds, so they spend a lot of time trying to develop ornithopters, only to realize their are better ways. Cars aren't machines with fast legs. Engines aren't artificial muscles.

I bet AI will also bear little resemblance to human intelligence.

"Most technologies seem to go this way. People start thinking aircraft will be just like birds, so they spend a lot of time trying to develop ornithopters, only to realize their are better ways. Cars aren’t machines with fast legs. Engines aren’t artificial muscles. I bet AI will also bear little resemblance to human intelligence.

Indeed. I'm reminded of a good scene from Total Recall. Somehow, I don't think autonomous vehicles will be driven by Johnny Cab,/a> with a human brain emulation. (But it sure was lucky for Ahnold to get a car with a joystick.)

Shorter book review: Hanson's putative non-fiction book is a rewrite of this scifi fiction classic: Permutation City by Greg Egan. https://en.wikipedia.org/wiki/Permutation_City

Or Children of Saturn or...

Watched one of his talks on Ems. Mind-blowing concepts. Looking forward to the book.

"It’s a mock of all those who believe in individual free will." - Since we clearly have free will in the relevant sense, this seems like an unpromising start.

The rest looks interesting, especially the theodicy and rational explorations of God (reminds me of Spinoze, who frankly probably did it much better), although I agree with MichaelG and Stephan's comments: in particular, the connection between our body and our brains is too intimate for any easy "replication" of us as a person / cognitive being.

I'm happy to grant that an emulation of your brain connected to a robot body isn't "you" anymore in a strict sense. But still, it would be very useful, and if many creatures like that could be made cheap, that would remake the world.

Too bad you can't patent it, at least based on current precedent in the US Patent Office. Back in the '80s, a patent that kept a monkey head alive with artificial respiration and life-support machines was rejected not on 'prior art' (past efforts by others) but because the patent was against public policy, on moral grounds. Go figure.

It's not at all obvious that you can separate the intelligence from the embodiment.

EMs are a lot less interesting if they all have to have a mechanical body that's more-or-less the same as a human body.

My money is on fully non-corporeal intelligences being quite different from humans.

Oh, my wife and I did that a couple of years ago. The result is a fair bit more interesting than a robot, and the upshot is that she has managed to upload a few things to me, as well.

An emulation of me connected to a robot would be useful but not worth dying for. Is there any conceivable technology that could get a high enough resolution map of my whole brain to emulate without utterly destroying it in the process?

"I’m happy to grant that an emulation of your brain connected to a robot body isn’t “you” anymore in a strict sense. But still, it would be very useful, and if many creatures like that could be made cheap, that would remake the world."

Yes, but it seems much more likely that the world will be remade by creatures that do not resemble robots with human brain emulations. For instance, suppose the goal was to get the world's ideal firefighter. It doesn't make sense to start with a brain emulation of the world's best human firefighter. For example, humans can't see in the infrared, or see through smoke. Humans can't see or smell carbon monoxide. Even the best human firefighters probably have some fear of dying. Finally, it might be useful for a firefighter to have 8 arms and 4 legs. A creature that could see in the infrared, see through smoke, see or smell carbon monoxide, had no fear of dying, and had 8 arms and 4 legs wouldn't be controlled by a human brain emulation. The human brain isn't set up to control 8 arms and 4 legs, and to smell or see carbon monoxide.

To use an example that involves less science fiction, Tesla and Google aren't aiming their autonomous car designs toward a chauffeur that's a humanoid robot with a human brain emulation.

In what relevant sense do we clearly have free will?

In the sense of (and I don't know the philosophical terms for this) believing we have free will. Even if the choices we make are all predetermined by chemistry in the brain, our subjective selves still 'feel' that we are 'making choices'. And subjectively that makes it so. To the person, they have made a choice, even if some omniscient being could know the choice I will make in advance. Other than God, who may or may not exist, there's no one aware in advance of the choices I will make. To me, that's free will.

So I guess I'm clearly the most handsome man in the world, in the relevant sense. Cool.

Dude, you are way better than that retort. 'Handsome' is an external trait, that can be easily seen and judged by others (although you can certainly think yourself handsome if you like). 'Free will', the sense of being able to make a choice, is purely internal and subjective. But you knew that (?).

Yeah, I knew you'd respond like that. So predictable. You probably think you could have done otherwise, and attach importance to that illusion.

Whereas I had no idea you would choose to reply like that! Maybe only one of us has free will.

Donohue, I know you meant that as a joke, but it's not a joke at all. Beauty truly is in the eye of the beholder, so the only relevant criterion for whether you are the most handsome man in the world is whether the beholder believes it to be so. If you are the beholder, and you believe it, then that is the only relevant sense in which you are the most handsome man in the world. In what other sense could we possibly mean it? Objectively? There's no such a thing as objective beauty. You already know that.

Now turn to free will.

You know with certainty that you make choices, and you know with certainty that there are thing that affect you that are not subject to your choices. Example of the former: what you will eat for dinner tonight. Example of the latter: whether the mail will be delivered on time tomorrow. So there is a relevant sense in which you make choices and a relevant sense in which you do not.

Question: Considering your choice of meals, is there any meaningful difference between "freely choosing to eat spaghetti" and "eating spaghetti due to a deterministic series of events that also convinced you that you chose to do so"? If so, what is it?

"I'm in the smallest room in my house. This book is in front of me. Soon, it will be behind me."

Of course, problem solving is what people (and robots) do, those better at problem solving richly rewarded for their talent. Oddly, there's always a clique devoted not to problem solving but problem solving avoidance, whether motivated by theocracy, ideology, or ignorance. There is no problem. The problem can't be solved. The problem isn't important. Problem solving avoiders are aided in their efforts by the human tendency to distraction and the skill of problem solving avoiders in devising new and novel ways of distraction. What can be done? My suggestion is to bring together the world's best problem solvers and solve the problem of the problem solving avoiders. The difficulty is that many of the problem solving avoiders are the best problem solvers, so it's all but impossible to tell if the goal of the problem solvers is problem solving or problem solving avoidance. Maybe the first step in solving the problem of problem solving avoiders is to identify the problem solving avoiders so they can be avoided. It will be a difficult task given the human tendency to distraction and the skill of problem solving avoiders in devising new and novel ways of distraction.

This is a book about how Ireland is going to take over the world?

I'm sad that nobody found my joke amusing enough to comment :(

God created whiskey to keep the Irish from ruling the world.

This book (Age of Em) sounds a lot like the documentary version of Richard Morgan's "Altered Carbon", where wealthy humans achieve effective immortality through continued upload (and backup) of their personalities into their own clones (and other bodies, for mostly prurient reasons). I hope Robin Hanson has read the appropriate literature.

Of course I've read it. Morgan keeps his world from getting too strange by positing laws against more than one copy of a person at a time. In reality, places with such laws would be quickly outcompeted by places without them.

Interesting insight - I look forward to reading Age of Em

If that review makes you think "this is a book I'd like!" then this is probably a book you'd like.

Why the weird title? What's an "Em"?

Please disregard my previous questions. I went to the book's home page and found the definition of "ems" in the very first sentence, so shame on me! Nevertheless, the idea of copying or emulating a brain sounds like science fiction, but regardless of its feasibility, this idea poses a fascinating thought experiment. I will pre-order my copy (of the book, that is--not Robin Hanson's brain!)

I left a comment on Hanson's site years ago that "Em" is not good branding for the concept he is trying to popularize. Em doesn't evoke any intuitive meaning and it's awkward phonetically. Hard to imagine Em entering the vernacular in the way AI or Singularity have. At least put the full word Emulation on the cover!

But I don't have any great alternative. Mind Emulation (MI)? Whole Brain Emulation? Wish I had something constructive for such an interesting body of thought.

Robin Hanson may be original and intelligent, but objectively he is not important, not that there's anything wrong with being unimportant like most of us.

Tyler, I recommend the UK television series Black Mirror also. It touches many of these themes.

Alas that so many are satisfied to merely "touch on theme" rather than figuring out what would actually happen.

Thank God that you have figured out what is actually going to happen. Phew!

The description of "Mind Theft" sounds suspiciously like having kids. I'm not saying everyone is enthusiastic about that but having a little strangely the same / yet completely different copy of yourself running around does bring some pleasures.

It makes me worried for our future Robotic Overlords.

The universe is strange. And it's not silly to believe in non-compatibilist "could have done otherwise" free will. I would think the existence of free will more obvious than the existence of an intelligent divine creator.

+1. Brian, any snark to add here?

I didn't say it was silly to believe in free will, just that it's very far from clear.

Color me skeptical.

Frankly like many philosophical questions we are discussing mostly terminology. As I've stated, even if everything we do is 'pre-determined' by the arrangement of molecules in our brains, we could still subjectively have 'free will', because whoever 'we' are feel that they make choices, so they do 'make choices'. It just depends on what those words mean. The state of the observer (omniscient outsider or subjective person doing the 'choosing') affects the answer.

So your skepticism could be correct in one sense, but still people subjectively have choices to make and they make them all the time. From their point of view, that's free will.

I don't know how to interpret what you're saying as other than "the illusion of free will", which doesn't strike me as anything like the real thing.

Illusions are very real from the point of view of the believer. For example, you probably believe many things about the world which, if they are true 'to you', then in that sense they are true (being loved, esteeming others, thinking the Cubs are the most important baseball team, etc). Again, we are going back and forth about what words mean more than any objective reality.

Think of it this way: we're all gonna die and someday the earth itself will be destroyed and after that the entire universe will end in entropy. So why get up in the morning, why love your family, why do anything? Because you 'believe' that your 'choices' have meaning. Even if the chemicals in your brain are doing it all, how is that any less 'real'? And what's the alternative? Let's say free will is an illusion? Knowing that, should any of us live our lives any differently? If not, then who cares about the illusion?

The real issue with free will is how it affects public policy. If you think everyone has free will and can make good decisions to work toward goals compatible with society, you will support one set of policies. However, if you believe e.g. that poverty affects brain states to the degree that it leads people to make bad decisions, you will have a very different view.

@msg, Even if I'm suspicious of free will intellectually, I feel like I have free will. It's how I'm made. You too. Again, that ain't the real McCoy.

The French have a saying: pour comprendre tout est de pardonner à tous.

Original D, you are perhaps inclined only to forgive some.

Brian, we're just going back and forth about what 'real' means. I think we may even agree that free will probably is an illusion. I just think the illusion is what matters to me, and most humans.

The issue is simpler than you make it and we do have free will:

It's the state of the brain that determines what action will follow (by the laws of nature) from what stimuli. The "state of the brain" is you. Hence, you are making the decision.

Sorry about the ? at the end of the third-to-last sentence.

I recently read something like this: A very intelligent physicist was recently mocked for proclaiming that, under Bayesian analysis, there is a 94 percent chance that his view of a potential TOE is generally correct (he was a proponent of a Calabi-Yau depiction of that which is 24 magnitudes of size smaller than what we generally deal with, in case anyone cares) . I would not mock him, because he is very smart, and because I believe that, under similar analysis, our understanding of how consciousness has come about covers only about 6 percent of the specific facts that are necessary for a credible assessment of that issue. (Also because I do not believe in mocking people on general principles, and because a silly and prolific commenter here once -(surprisingly to me, not because I certainly knew what I was talking about, but because I knew that he did not) mocked my similarly Bayesian assessment of the best philosophers in (and by) their disagreement , despite their best human efforts, as having established that the peak of their best understanding of reality was similarly amenable to a numerical ratio compared to the actual truth - I don't precisely remember my comment, but I think I stated that, based on the look of the Venn diagram of their shared conclusions among their constant animadversions, and based on the lack of any indication that one was much more insightful than the other, even at their upper tail end of the insight curve, that the best of them could not be much more, within a fairly small window, than half correct). My guess is that the first thousand generations of artificial intelligence will be confined by their owners ( I believe they will have owners, some more thuggish than others) to the existence of, say, very bright pet sharks, who have an instinct to go on with no regard to pleasure or the lack thereof, or like those charming insects (described by Nabokov and Kafka) who consciously but unthinkingly and unidirectionally gravitate on their complicatedly multiple legs towards the sunniest window in the room. We (Westerners, that is) already know that it is possible that no good conscious act is lost (hence, the joy of each and any one of trillions of little new-born gnats, almost lighter than air, on May mornings a million years ago is preserved in the correlated and sympathetic consciousness of one of the near-infinite angels who may have observed that joy) ; what we civilized Westerners worry about are the unpleasant avalanche of consequences, most fully described in recent years in the Abolition of Man, of using finite skills to alter , in an inevitably finite and possibly mistaken or even unkind way, the course of future limitations on the interface between future embodied souls/ consciousnesses and their physical surroundings.

I would in the situation above have preferred "beam em up".

With that, I shall retire to read that work by the greatest novelist, jane aust em, the novel 'em a'. And contemplate the great koan by Richard Dawkins, (m)ememememem(e).

Comments for this post are closed