What is transhumanism?

[mindstalk]

Over on rpg.net there's a Traveller thread going on which has included a bunch of transhumanist debate, of sorts. Some people saying the tech assumptions of Traveller are ridiculously retro and out of date, and others sniping at the transhumanism trend, full of fantasy and wish-fulfillment, no more realistic than Traveller and already dying to be replaced by the Mundane SF trend. Which strikes me as nutty.

Traveller: completely jury-rigged FTL system, human-equivalent aliens, and a side dish of psionics.

Transhumanism: well, what is it really? First I'd like to dump the Singularity as a concept non-conducive to communication; Vinge himself meant a bunch of different things by it, sharing a core of "superhuman intelligence will make things weird", and as used the word's only gotten more fractured, with a lot of people using it as "spontaneous nano-upload of everybody".

A better touchstone is GURPS Transhuman Space, I think, a setting which involves increasingly extended lifespans (through a suite of medical interventions I think, not a pill of nano-machines), various genetically engineered humans and animals, AIs, and 'ghosts' -- uploads of human brains. Plus, lots of SPAAACE. Apart from anything explicitly about intelligence boosting (which might break a game setting) I think that's nearly a clean sweep of transhumanist tropes. The suite of concepts could be abstracted as "superb control over life, mind, and matter, with an emphasis on life and mind."

The endstate could be idealized as "eternally youthful supergeniuses who can shape their environment at will, sculpt life, not have to work much, and go anywhere they want." Put that way, it sounds rather wish-fulillmenty, Big Rock Candy Mountain for techies.

Thing is, it's not pulled out of the aether. If you step back and posit superb control over life and mind, eternally youthful supergeniuses seems like a reasonable expectation. The brain is supposed to be a material process, largely information processing and action coordination, which can be understood, modified, improved, and duplicated -- which gives you AI, geniuses, and uploads. People don't die because Atropos cut their thread, they die because of entropy piling up in their cells and organs, because specific organs or processes fail. If the damage can be reversed, or the organ replaced by a new one, aging can be reversed or sidestepped.

To avoid this I think you have to postulate either that we don't get the understanding of life and mind, which seems counter to current trends, or that we do but using the knowledge is unavoidably difficult and expensive. Which is possible, but it's certainly not obvious that it'll be the case, any more than pocket-sized camera-phones which can reach around the world or palmtop computers or sequencing the human genome in surprising time were too complicated to pull off.

As for wish-fulfillment -- well, Iain Banks replied "yes, it is, that's exactly what technology is for." But one might want to look at our society, one where most children don't die, where people can go into their thirties without having someone they know die, where the vast majority of people aren't involved in food production, where life expectancy for all goes up to 80, where we can summon excellent music or video with the touch of a button, have cheaply illuminated nights without smoke, and have comfortable homes even in hot humid areas. That's over the top wish fulfillment fantasy to most of the human race, not even getting into the cars which move at multiple times a horse's speed and don't have to be fed every day, or the laundry machines, or the birth control pills...

One trope I didn't mention was nanotech. I've heard Transhuman Space is restrained in using it. At any rate, I view it as a means, not a goal. The goal is superb control over life etc.; Drexlerian nanotech is one envisioned implementation, with extra appeal for autarchic fantasies by libertarians not content with ordering products off the web from robot factories, paid for by their social democratic matter and energy allowance. An abused magic wand, yes, but not actually necessary for the grand vision!

Along the way, I had a couple of posts about Bujold and transhumanism. And a brief note on how we're living transhumanism -- or what would be the build up to it if it's going to happen.

Comments

[mindstalk.livejournal.com]

To be fair, as I think more, the "fix the brain" part of the eternal youth program might well be a stumper. Even if rejuvenated neurons can be induced to grow, that might just end up overwriting your original memories, resulting in a shiny new brain -- and person. And computer tech really guttering out could crimp the AI front, or limit them to big, cooled computers... OTOH, comps and bots should end up at least a fair bit more complex than they are now, understanding the mind should result in *something* -- I like to talk about the Cognitive Revolution these days, rather than the Singularity -- and one could always grow bigger/smarter brains genetically.

[mindstalk.livejournal.com]

And of course a side-meme of transhumanism is not just that big changes are possible, but that'll happen Soon. In our lifetimes. Or by 2020! This is obviously less certain or likely... OTOH I got provoked to write all this by Traveller discussion, and that game is set three millennia in the future.

[xuenay.livejournal.com]

Even if rejuvenated neurons can be induced to grow, that might just end up overwriting your original memories, resulting in a shiny new brain -- and person.

Sounds pretty doubtful. For one, new neurons already grow within our brains. Any sane process by which new neurons would be introduced would be a gradual one, with them slowly integrating themselves into the currently existing system and adapting their function to fit in. Some memories might be lost, sure, but we forget things all the time, so that's not much of an issue. Memories aren't coded into individual neurons, memories are coded by the way neurons form networks - so introducing new neurons can't really "overwrite the original ones".

(Was randomly poking around to see what recent journal posts Technorati displayed when searching for "transhumanism", for those curious of how I got here.)

[mindstalk.livejournal.com]

I was thinking that to deal with age, we might need a rather higher rate of neural addition, and of new types, and that might be rather different in behavior. Neurons crowding out old ones, making new connections different from the old ones. I know memories aren't thought to be in the neurons (well, they kind of are -- the synaptic weights) but I was trying to see how one could be justifiably pessimistic, given that I seem like a bubbly optimist anywhere outside the extropians list.

Technorati sees me? Wow. Fast pickup, too. Hi.

[dsgood.livejournal.com]

A properly-working bureaucracy is a form of AI.

[mindstalk.livejournal.com]

Heh. I suggested that in a class paper. A rather low level one, though.

[pompe.livejournal.com]

There are several problems. One is that lifespans aren't increasing quite as impressively as they should. I don't really know how much of those extra years nowadays, with say the increased lifespans the last 25 years, are active years or simply more effective pre-death care - that is, better skill at slowing dying instead of longer living.

If you look at it in the long perspective it is also dubious. A hundred years ago a Swedish male lived for 55 years, on average. (When the Statistics Bureau started working in 1751, it was 34) Today, 79. 24 years. But the _remaining_ years at 65 have gone up from 12.6 to 17.6. Just five years (but admittedly three of them past 1980). For women it is in fact better, but that peak came earlier. The increase of their lifespans-past-65 rose more sharply in the mid 20th century. This suggest to me that while we live longer and lifespans increase effective immortality isn't predictable as such, it looks more like we push average lifespans upward over 250 years but don't extend the top nearly as much.

Second, I'd like to stress the difference between understanding a process and recreating it for a complex purpose. We've known how solar fusion works for over 60 years and we can use the knowledge to blow up cities, but we're mediocre at using it to produce energy for a more complex purpose. Similarly, we're rather good at genetics, can do cut&paste operations but we aren't nearly as good with the complex purposes. What we suggest when we argue that superb control over mind and matter is a reasonable expectation is extrapolate too linearly in an area of diminishing returns, in my opinion. It's the inevitable S-curves of life.

[mindstalk.livejournal.com]

All good points.
I believe I've read that the elderly are healthier and more active than before, implying healthy life expectancy is extending as well as total, rather than just stretching the lingering period.

Yeah, that's Olshansky's argument. Of course, that's linear extrapolation of existing statistical trends. :) A big jump would have to come from a big jump in technology -- calorie restriction in a pill, or spare parts. That's actually where I pin my hopes, often joking that I hope to live to see "modern medicine", defined as fully functional spare parts for the body, up to and possibly including the brain. I see a big difference between medicine with full replacement capability and our medicine which is rather limited in capabilities: kill that bacterial infection, kill that cancer, prime the immune system against that virus, plug in dialysis for a kidney. Of course, our bodies weren't evolved to enable proper maintenance, so it'll be tricky.

Fusion's a bit tricky. We think we understand how it works in stars, and if we're right we could make a new star. Pretty simple if you have enough H2! Maybe even manipulate a star by doping it. What we're trying to do is to do fusion differently than its done in a star, using electromagnetic confinement despite its leaky nature, rather than massive confinement.

That might be a telling analogy for AI vs. brains. On the other hand, planes vs. birds might turn out to be a better analogy. Or pumps vs. hearts.

Genetics... I'd have said we're beginning, not rather good. Still at the bottom of the S-curve, not the top. We haven't been doing it that long (or we have, but with clumsy tools) and it benefits directly from Moore's Law. It's not that many decades ago that we found out what DNA was. We're finding out how Hox genes work and such. We don't have giant databases correlating human genes and traits. We're nowhere near saturation...

[pompe.livejournal.com]

But we always try to do things differently. Applied genetics don't use natural selection or the generational spans the world does. (Not to mention nanotech).

And what we need to do to correlate genetic databases versus human traits as far more complicated than simply sequencing DNA. You need something to correlate with. Some of it we can do with mice and similar cheap lab animals, but some we need to correlate with people. Which either means long generations or correlation with large databases which does not just list people but also their traits. That's doable but hugely time-consuming and packed with legal issues - imagine trying to correlate and prove in a _medically acceptable_ way which genes give bonus Intelligence. I'd not like to be running a medicorp being sued for supplying faulty Creativity genes. And then you need to account for the influence of the environment and how to activate the genes...

Realistically speaking, if people are queasy about National ID's I can imagine practical problems with genetic correlation databases too. And here comes another problem. The bodies which conceivably could (and already have) assemble such databases and correlation records, over reasonable time, with the legal issues covered, and with the long investment returns involved, are not corps but nations. But most nations may feel a bit queasy about that, and most of those who won't simply do not have neither the records nor the research capacity.

[xuenay.livejournal.com]

China might do it.

[pompe.livejournal.com]

Yes, but it'd require a major investment in bioscience and testing, just keeping the scientists from going Amro would be hard - and you probably want a more levelled (and less toxin-enriched) society to correlate genetic differences. Plus you have a defiency of siblings to correlate with. But admittedly China is the best bet.

This sort of project would have worked in Norden 1957, but I'm suspicious if it'd work in 2007, barring some sort of political/ideological upheaval.

[xuenay.livejournal.com]

As for lifespans, SENS (http://www.sens.org/) seems promising, though of course they could run into any number of unexpected roadblocks. (De Grey made quite an inspiring speech (http://video.google.com/videoplay?docid=8554766938711591377) about it a while back, though obviously he's bound to be the most optimistic one.)

Still, it's true that if that approach - or something like it - meets failure, lifespans probably won't extend too radically. Only as much that you can do - beating major diseases will help increase the average longevity, but old-age -caused detoriation of bodies is the worst killer in the end.