Astrochicken.

Astro. Chicken. Seriously.

(I figured I should take a break from all the Times bitching. How I came across this: while writing the previous post, I thought of other people I'd like to see on the Op-Ed page, and the Dyson clan came to mind, and...)

Futurist fun in Wikipedia: elevators and fountains

Savvy science fiction fans and futurists should be familiar with space elevators. And, sure, those are a pretty cool idea, but the other day I read the Wikipedia article on this subject and came across the truly mindblowing idea of a space fountain: a floating platform that's kept aloft not by orbital velocity, but by a continuous stream of mass accelerated upward from the planetary surface. The station at the fountain's summit stays aloft by absorbing kinetic energy by deflecting the mass back down to the base station.

Imagine keeping a dinner plate aloft by firing a tennis-ball gun at it. Every time the plate starts to descend, the gun shoots another tennis ball. When the tennis ball hits the plate, the plate absorbs some of the tennis ball's energy, and it gets to stay in the air for another tiny fraction of a second. Now imagine that the dinner plate is a space station floating a couple of hundred kilometers in the air, and the gun is firing billions of tiny tennis balls at incredibly high speed every second. And instead of bouncing off the plate, the plate has electromagnets on it that bend the stream of microscopic tennis balls in a U, sending them shooting back down to the tennis ball gun at equally high speed.

OK, so the analogy breaks down. It's still a cool idea.

Incredibly enough, judging by the Wikipedia articles, it even seems that (apart from the fact that it requires incredibly massive, massive, gi-normous energy input) a space fountain appears to be easier to build than a conventional space elevator. For Earth-orbiting elevators, a conventional space elevator would require material with tensile strength exceeding diamond.[0] Space fountains can be built from ordinary materials.

Now we just need a whole lot of energy. Nuclear fusion, maybe?

Speaking of energy, the space fountain article's "near-term applications" section is almost as fascinating as the idea itself:

A closed loop projectile system could be used for energy storage, similar to a very large flywheel, providing load leveling for terrestrial power grids. If the closed loop was long enough it could even be used for power transmission.

A very small-scale fountain tower could be used for constructing tall antenna masts rapidly, perhaps for news events and military operations. A larger and more permanent fountain tower could be ten or twenty kilometers tall, allowing one facility to provide radio and television broadcasts to enormous areas such as the steppes of Asia. Fountain towers might also prove to be an economical alternative to communication satellites for point-to-point television and FM radio communication between the various islands of some of the smaller nations in the Pacific Ocean. An elevator and observation platform could also be added as a tourist attraction.

Arched fountain structures similar to the launch loop could also have useful small-scale applicaions, notably the construction of bridges. Projectile-supported fountain bridges could be made arbitrarily long, without the need for support pillars anywhere along their span.

A couple of other links, closely or marginally related:

• Atlantic article: "Stairways to Heaven".
• Giant water fountain found in space

[0] Interesting fact: according to the Wikipedia space elevator article, diamond actually has a tensile strength only marginally higher than quartz; and neither is even remotely strong enough for a space elevator. Carbon nanotubes seem to be the best current hope.

Humans: Prepare to meet your strange new peers

(More bookmarks cleaning.) Hans Moravec asks: When will computer hardware match the human brain? (Interested readers should definitely read the commentary as well.) The interesting thing about Moravec's analysis is that it is immune to a couple of the standard objections:

• Standard Objection 1: "AI researchers have been promising imminent breakthroughs for the past 50 years. This is just another wildly optimistic prediction." Moravec's analysis notes that, although AI research has been conducted for 50 years, the amount of funding available to AI researchers has decreased dramatically, with the result that AI researchers have had nearly a constant amount of hardware power at their disposal until fairly recently. He also argues that AI strongly tracks advances in hardware --- the AI software you write for a 100-MIPS computer is not just a bigger, faster version of the software you write for a 1-MIPS computer, it's designed fundamentally differently. Multi-million-MIPS computers will enable vastly new approaches that enable quantum leaps in AI power.

• Standard Objection 2: "Neurons are far more complex than is popularly believed. Your estimates of brain complexity are far too conservative." This is irrelevant to Moravec's analysis, which relies on an analysis of a functional unit of the nervous system, not raw neuron counts:

  More computer power is needed to reach human performance, but how much? Human and animal brain sizes imply an answer, if we can relate nerve volume to computation. Structurally and functionally, one of the best understood neural assemblies is the retina of the vertebrate eye. Happily, similar operations have been developed for robot vision, handing us a rough conversion factor.

  ...

  It takes robot vision programs about 100 computer instructions to derive single edge or motion detections from comparable video images. 100 million instructions are needed to do a million detections, and 1,000 MIPS to repeat them ten times per second to match the retina.

  The 1,500 cubic centimeter human brain is about 100,000 times as large as the retina, suggesting that matching overall human behavior will take about 100 million MIPS of computer power.

  In Moravec's analysis, it's irrelevant how computationally complex individual neurons are. The volume of neurons in the eye performs a visual processing task that can be simulated by 1,000 MIPS. Regardless of how these neurons accomplish this task, it appears that 1,000 MIPS is roughly adequate to replace the function of that volume of neural matter. It stands to reason that other functional units of the brain would require similar amounts of computing power per unit of volume.

  Of course, neurons may be more densely packed and interconnected in the brain proper than in the retina, but probably not by more than an order of magnitude or two. Assuming continued exponential growth in computating hardware, a mere order of magnitude only signifies a few years' difference: perhaps it will take fifty-three years instead of fifty. Hardly a basis for pessimism.

I don't buy Moravec's analysis entirely (there are some obvious objections, although no fatally conclusive ones, to Moravec's analysis) but it's an interesting read nonetheless.

"Transhumanism": Friend or foe?

While I was visiting NYC, I had a conversation with PM and JW (and later BM) about transhumanism, a topic that's been occupying a lot of my thoughts lately. Basically, we all agreed that it seems possible to radically alter certain aspects of the human form and psyche. What we disagreed about was whether one should use the label "transhumanism": I thought that it was an accurate word that served the useful purpose of provoking thought, whereas PM and JW thought it was an inaccurate piece of mystification whose only purpose was to obscure power imbalances in existing human societies. To summarize PM's arguments, perhaps, unfairly:

• Humanity is not defined by its physical form, psychological urges (sex drive, etc.), or limitations (mortality, etc.); it is defined by the fact that it creates cultural artifacts. Therefore, even "transhumans", regardless of how bizarre they become, will still be "human", because they will create cultural artifacts.
• The differences between "transhumans" and present-day humans will be no greater than the differences between present-day humans and primitive humans from, say, 10,000 years ago. Therefore, calling these future beings "transhuman" is misleading.
• It is evil to have a society in which some people are "human" while others become transhuman (or, since PM says that the "transhuman" label is irrelevant, "much more powerful humans"). Therefore, talking about transhumanism is wrong.

All three of these points seem fundamentally wrong. To begin with, the first point's definition of "human" is deeply flawed. For example, this definition would call "human" any hypothetical race of space aliens, as long as they had something you could call culture. A spacefaring race of metal-eating bacterial colonies that communicate using protein-bearing spores? A society of sentient asteroids that travel across the galaxy searching for stars whose gamma ray signatures express the [*untranslatable*] of the [*untranslatable*]? Would these aliens be "human" as long as they produced "cultural" artifacts? This contention is absurd.

We can expect that radically altered humans would be no less foreign to us than these hypothetical aliens. No doubt they will share some characteristics with us, but they may share equally many characteristics with a termite mound, or a wheat field, or a hurricane. To call such beings "transhuman" or "posthuman" seems only accurate.

As for the second point, consider a hypothetical baby Alice born 10,000 years ago in a hunter-gatherer culture. Imagine transporting Alice into the present and raising her in a modern society. Now imagine downloading Alice's psyche into a colony of hyperintelligent nanomachines designed for asteroid mining, while simultaneously altering her personality so that she could never feel any emotions. I think you have to be willfully contrary to claim that the latter leap is no larger than the former leap. Hunter-gatherer Alice and modern-society Alice both have two arms, two legs, a face, a single brain; they'll grow up into people who walk around, who eat and drink, who desire sex, who fear death, who may feel love or anger or hate or jealousy. Nano-colony Alice's body is not even composed of biological materials; her thought processes are deeply foreign to every human being who has ever lived. To me, it seems obvious that the leap between nano-Alice and modern-society Alice is at least as great as the leap between Australopithecus and Homo sapiens. Only a tendentious reading of the facts could conclude otherwise. We call Australopithecus proto-human. We should call nano-Alice post-human.

Finally, as for the third point --- that transhumanity implies immoral differences in power --- we can clearly foresee that radical alterations to humanity will become possible. We can also foresee that some of these alterations will make people much more powerful. We can also foresee that some of the alterations that make people much more powerful will nevertheless not be universally adopted --- do you want to be downloaded into a hyperintelligent nanobotic asteroid mining colony? Some people would say yes, but most present-day humanity would say no. In other words, many people value their humanity, or at least certain aspects of it, and would therefore refuse an alteration that took away those aspects of their humanity. Therefore, one has three choices:

1. Establish a world government that bans all transhuman alterations.
2. Establish a world government that coerces everybody into having the same set of transhuman alterations.
3. Establish a society based on ethical relationships between transhumans and humans.

The first two alternatives seem monstrous to me, and therefore we are left with the third. Refusing to talk about transhumanism doesn't make the third alternative any less real. And calling radically altered people transhuman doesn't obscure the power imbalance; rather, it brings that power imbalance to the fore, and highlights its unique properties. Past power imbalances among humans have mostly derived from the fact that some people wish to deprive others of power. The power imbalance between transhumans and humans is more fundamental: it derives from the fact that some humans will not wish to partake of the power offered by transhuman enhancements. To be sure, there may be transhumans who wish to deprive humans, or other transhumans, of power, but the imbalance between transhumanity as a whole and humanity as a whole is new.

When I said that transhumanist ethics should require that transhumans set aside enough space and resources for humans to live fulfilling human lives, PM derisively compared it to "keeping the savages on the reservation". But I don't see any other alternative.

Should transhumans instead compete with humans for the same resources? Transhumans would win every time. A single transhuman might produce more intellectual output than an entire human population of billions. A transhuman might be able to manipulate human legal systems with the ease of a child playing with Legos. And physics dictates that a transhuman psyche downloaded into a computer with the mass of a golf ball would be able to colonize remote planets faster than any human (who would be encumbered by a mass weighing in the tens of kilograms). Transhumans might have the power to destroy human economies, subvert human legal systems, and establish sovereign governments far beyond the reach of any human. The only forces stopping a transhuman from doing these things would be its own conscience and the will of other transhumans, not any human power.

So the resources left for humans would be only those that transhumans made a conscious effort to set aside for that purpose. (Note that a second tenet of transhumanist ethics is universal enfranchisement --- any human who wants to become transhuman should be given not only the liberty but the resources to do so. So humans would be human by choice.) For savages like us, it will be a matter of choosing between the reservation and transhumanity. There's no other choice.

And actually, I don't think PM had any other choices in mind. I strongly suspect that his reaction against my description of transhumanism stems from denial, not any positive agenda. Like many other anti-transhumanists, I think he dislikes the notion of a society founded on vast power imbalances, and therefore prefers to invent reasons not to think about it. But avoiding the idea won't make it go away.

Note, BTW, that most transhumanist thinkers are fully conscious of the risks involved in the transhumanist project, or at least as fully conscious as anyone can be. In fact, analyzing these risks is (yet another) major aim of transhumanist thought. See, for example, Nick Bostrom's many writings, including his discussion of existential risks to the human race, and his unconventional analysis of the future of human evolution.

Memory-erasing drugs and cognitive liberty

J. McNeil points to an ongoing discussion, originating on Brain Waves, about memory erasing drugs: first post, reply at futurepundit, followup. My own position on "cognitive liberty" is pretty close to Boire's; this is why I differ from experts such as M. A. R. Kleiman on the drug issue. Libertarians are ultimately right about drugs: the right to control one's own cognitive apparatus is a fundamental freedom, without which all other freedoms are meaningless.

Incidentally, skimming the Brain Waves blog reminds me that in B. Sterling's novel Distraction, neurotechnology is the Next Big Thing. Those days may be approaching.

Statites!

Whilst reading about Dyson spheres on Wikipedia, I came across this article on Statites; it's a cool idea, but this paragraph jumped out at me:

The concept of the statite was invented by Robert L. Forward. No statites have been deployed to date, as solar sail technology is still in its infancy. See Forward's US Patent 5,183,225 "Statite: Spacecraft That Utilizes Light Pressure and Method of Use." filed 9 Jan 1989 issued 2 Feb 1993.

Let's see, if this patent was issued in 1993, that means it will run out in 2010. Filing a patent costs considerable time, money, and effort; you don't file a patent just for the hell of it, unless you're rich. SO this dude evidently thinks that he'll be able to reap profits from his patent on statites before 2010. Erm, right.