The age of em

I really don’t know what to make of this book. It was on my bedside table for more than 3 weeks. This is very, very unusual. It is not that I couldn’t read it, it was actually quite interesting but I could manage only a few pages at a time, and that’s because I needed time to think about and reflect on what I had just read. It sounded so weird and so unbelievable that I wasn’t sure if this was a serious work from an accomplished academic, as Dr Robin Hanson, or a science fiction dystopia. Perhaps, it’s both.

Ok, let’s start. Nowadays, our economy doubles roughly every 15 years, from every 1000 years during the farming period. If this trend were to continue, we should expect, according to the statistics models, that sometime during the next century, our economy to go to doubling every one month, or so. This will last for a couple of years maybe, and then something entirely different will happen that would change everything. What would cause that disruptive change? The arrival or artificial intelligence, that is robots, smart enough to substitute wholesale for human workers.

There are several stories how this will become possible. One scenario is that we will keep writing and accumulating better software. At this rate, we need about three to four centuries for a full AI. Another scenario is that humans would be able to modify their biology so as to achieve a superhuman intelligence. Robin Hanson, who studied physics and was a software engineer before he become an economics professor at George Mason University, believes there is an entirely different way, and that is porting software from where it already exists, and that is the human brain.

To do that we need three things. First, fast computers, much faster than those today, second, scanning machines that will be able to produce a most detailed scan of the human brain with all its particular cell features and connections, and finally, computer models that would be able to process signals for each brain cell. If we have all these, then we can have robots which will be whole brain emulations or “ems,” for short. Robin Hanson expects that the first ems will appear within roughly a century or so.

Emulations are not a new idea, there have been in science fiction for many decades. Arthur Clarke investigated this idea in his novel The City and the Stars, in 1956 and more recently in 2003, in his novella The Cookie Monster, Vernor Vinge explores the idea of conscious computer simulations.

In The Age of Em, Robin Hanson focus on robots and how the world would look like at the next big era after ours. He explores what happens in this strange world that it is dominated by trillions of ems that live and work in liquid-cooled skyscrapers, in very dense and very hot cities where they can quickly interact with other ems. He talks about the physics the economy, the organisation and their attitudes to law, politics, love, sex, and a lot of different other things.

Ems are smart, efficient, conscientious and workaholics. They basically work all the time but despite their hard work, they earn just enough to survive. Because they are so many and because it is easy – although not inexpensive – to make a lot more, the value of the work goes down. It is the supply- demand concept of economics.

Ems congregate in related “clans” and make use of Decision Markers to make key group decisions. Psychologically are very human. They are not ordinary humans, but they have all the psychological tendencies that humans have. They are after all emulation of human brains that are put on a computer. Therefore, they develop relationships, they have friends, lovers, work connections. Some ems have bodies, others do not.

Emulations do not have to feel pain, or hunger. They don’t have to face death as they can make billions of copies of themselves. Em ethics are different from ours. Their society is less democratic and gender-balanced, more divided into district classes, and its leaders are more accessible and trusted.

But what about humans? If they survive, they will retire somewhere outside the dense cities and will live on pensions doing whatever they want to do. It is not very clear, humans are not the purpose of Robin Hanson.

The Age of Em is an efficiency, undemocratic utopia or dystopia, depending on your perspective. The em scenario is not entirely persuasive to me but the book provides a baseline for future study in the important topic of artificial intelligence.

It's looking increasingly likely that eventually we'll be clever enough to create artificial intelligence with at least human-level capability. Two likely ways we might do this are 1) to increase the sophistication and coordination of our intelligent algorithms, or, 2) to learn how to simulate the human brain in a computer in such a way that the simulation has equivalent capabilities to the original brain. Hanson puts his bets on the second option happening first, and has written this book to explore what this era will be like as emulated brains ("ems") proliferate and organize into their own cultures.

The book is largely a set of densely-packed, fabulously multidisciplinary extrapolations from existing human historical trends, cultural varieties, and individual eccentricities that attempts to anticipate what this emerging em culture will look like before it is quickly swallowed up by whatever it invents to replace itself.

It is a difficult book to read. Hanson uses discipline-specific jargon and terms of art without pausing to explain them precisely, which often made me feel like I was floating on a cloud of gauzy sociological generalities. It's also often very difficult to figure out why Hanson (or "we" as he sometimes likes to call himself) is making a particular prediction, how confident we should feel about it and why, what the disregarded options were and why they were disregarded, and how the various predictions and their variances might be expected to interact with each other. Often he seems to just be confidently extrapolating from a tiny handful of data-points, not because there's good reason to believe that mere extrapolation has predictive power, but because if you have faith in extrapolation you can predict confidently and predicting confidently is fun. Without much to go on to evaluate his particular guesses, you just have to credit Hanson as being a very smart polymath who is obsessed about this particular topic, and take it or leave it.

This would be an invaluable resource for someone intending to write a book of cyberpunk fiction about this possible era.