Filed under: Trees
Here’s an interesting jeu d’esprit of extreme conservation (via Gary):
My favourite long-term solution is simply to aim for not just a post-industrial civilization but a post-biological one. We can currently roughly foresee how we could go about it. We would fixate our brains (presumably when near biological death), scan them in detail, reconstruct the functional structure and recreate it as software. The successor version would then go on living in virtual reality, with occasional visits to the physical world using a robot, android or just remote controlled human body.
How efficient could a postbiological civilization be? The current IBM roadrunner does 376 million calculations per watts. If we take my mid-range estimates of computing needs, 10^22 to 10^25 FLOPS, then a single emulation would need 10^13 to 10^16 watts. The total insolation of Earth is about 10^17 watts, so this won’t do – there would be space for just a few minds on the entire planet. But current research on zettaflops computing suggest we can do much better. A DARPA exascale study suggests we can do 10^12 flops per watt, which means “just” a dozen Hoover dams per mind. Quantum dot cellular automata could give 10^19 flops per watt, putting the energy needs at 200-2000 watts.
That is between 2 and 20 times the current wattage of a current human. However, we bio-humans get our energy through the inefficient method of having plants collect sunshine (at about 3%) efficiency, then we either harvest them and eat a small part of them (expending a lot of agricultural energy) or have animals eat them (at a few percent efficiency) and finally we eat the result, again with a few percent efficiency. A brain emulation of this type would just need a few square meters of solar panels (plus night-time energy storage). In terms of area and energy required, these postbiological humans would have far smaller material requirements than we do. They could also run slower to save energy.
How much matter would go into this system? Using [quantum dots], each gate would be on the order of a nanometer. Each floating point operation would require about 20,000 gates. These gates would be re-used every 0.1 millisecond timestep, so a full 10^25 flop emulation would need a volume of 0.02 cubic meters. This does not take the rest of the infrastructure into account. Let’s scale it up to one cubic meter. 6.7 billion people would then require the same volume, or a cube with side 1885 meters if bunched together into the ultimate datacenter. That is unlikely to work if the energy use is on the orders of many watts per person, since cooling would be hard (not to mention the vulnerability of having everybody in the same spot). A more likely solution would be smaller centres distributed close to energy sources: a single hydroelectric dam would supply several million people with energy, a square kilometre of 20% efficiency solar panels would supply 150,000-1.5 million people. A 100×100 kilometre area would be enough to run all of posthumanity. And if the reversible computing works, the energy collection infrastructure could be 10,000 times smaller.
Maybe the most sustainable thing we could do would be to aim at a future ensconced in cold datacenters under the subtropical deserts of Earth. Humanity would largely look like a forest of quiet semiconductor trees. We would indeed have become plants.
It reminds me a little, among other things, of this passage from Eating the Sun:
The simplest, and perhaps the most profound, of the differences between those that eat light and those that eat others … stems from the fact that sunlight is, at the efficiencies photosynthesis is capable of, a rather dilute source of energy. To appreciate how insufficient it would be to animal needs, imagine the Green Man of forest folklore. Let us assume that his greenness is due to chlorophyll through which he feeds himself. Given the surface area of his skin—and the fact that at any given time some of it will inevitably be averted from the sun—such a green man would have about as much energy on which to run his metabolism as someone restricted to a diet grown in a couple of square metres of garden. All he could eat in a day would be what his little plot could grow in a day. A few leaves for breakfast, maybe a morsel of root for supper: berries for Sunday lunch.
On the sunshine equivalent of this meagre diet, our Jack of the Forest has no energy for moving, or for thinking—nerves and muscles use a lot of energy. He lacks the energy to breathe in or out, or to keep his body any warmer than the outside air. He’s not good for much except sitting there repairing the daily entropic wear and tear to his body. Indeed he doesn’t really have enough energy for that; quite a lot of him will rot away. The prognosis for the Green Man is vegetable.
Plants simply don’t have the energy to rush around like animals, pumping blood and flapping wings and flashing nerve impulses hither and thither along their limbs. So they eschew the compactness of muscle and opt for the looseness of leaf. The lines of their lives do not rush back and forth across the landscape—instead they are recorded in their shoots and limbs and twigs. Plants have shape where we have behaviour; their history is recorded in their form. Where the swoop of the sparrowhawk falling on the pigeon is gone in an instant, the tree’s decision to grow its twigs this way or that, depending on the light, is written in wood and lasts the rest of its life—or at least until some rough wind or uncouth animal snaps the relevant limb off.
I must admit, I don’t like the idea of the disembodied, emulated life. Although my own is not the finest specimen, I am all for embodiment, not least for the existential constraints it provides; I like my consciousness to come with an off switch. That caveat aside, I have the feeling that given its premises Anders‘ vision is oddly conservative. The idea that such power would be allocated to emulations of individuals seems a little traditionalist; if this trick can be pulled off, then surely it would be to create worlds where consciousness was far less confined, new universes of thought and distributed experience. And having the thinking trees rooted in the Earth seems peculiar; it is a harsh and vulnerable place for such things. As Freeman Dyson has argued on various occasions, places like the Oort cloud seem much more hospitable for life of this type, and indeed of other types.
And why on earth or off it do this in real time? Surely once the clock rate is semi-arbitrary one would either want to go really fast, so as to cram the most in, or — another Dyson idea — very slow: To watch the milky way turn like a waxing moon, to see the quasars redden like leaves in the fall, to tell stories around campfires of thought after the last stars gutter out. And to find out what spring, if any, comes after.
Image from Indium‘s solar materials science blog, used under a Creative Commons licence
2 Comments so far
Leave a comment