Comprehensive List of Existential Risks, Part 2 Tuesday, Dec 5 2006
risks 3:30 pm
Again, I’m trying to create a comprehensive list of existential risks.
Very low-probability:
1. killer natural virus.
2. alien invasion
3. asteroid impact
4. simulation shuts down
5. gamma ray burst
6. supervolcano eruption
7. black hole impact
Wouldn’t kill everyone, but still worth preventing:
1. nuclear holocaust
2. runaway climate change
3. repressive global dictatorship
The important ones:
1. superintelligence – not just AI – but superhumans too
2. deliberate misuse of nanotech (arms race, nanoweapons)
3. accidental misuse of nanotech
4. killer artificial virus
5. antimatter holocaust?
6. particle accelerator disaster
Ways to counteract:
1. friendly superintelligence
2. nanofactory restrictions
3. universal sousveillance
4. ocean habitat
5. subterranean habitat
6. antarctic habitat
7. space habitat
Here’s a “grid of risk” I’m trying to build, to get a better classification scheme going.
Superweapons Biotechnology Nanotechnology Cognitive technology
Nuclear weapons |Hybrid/designed diseases |Arms races |Neural implants
Solar weapons |Full-fledged hyperdisease |’Grey goo’ |Wireheading
Kinetic weapons |Artificial life (nanobiotech) |Thermal limit |Superintelligence
This grid of risk doesn’t look terribly incomplete, but it deliberately ignores antimatter weapons, because I don’t know enough about them, particle accelerator accidents, which I don’t feel qualified to speculate too much on, or weapons being used to trigger a supervolcano eruption or gigantic landslide (La Palma). Let me know if you feel anything is missing, or misclassified, or not a risk at all.
“antimatter weapons, because I don’t know enough about them,”
Amat weapons are like nukes, except that they heat an entire area with radiation instead of creating a single point of destruction. And you have to make antimatter which takes a hell of a lot of energy input.
“particle accelerator accidents, which I don’t feel qualified to speculate too much on,”
Cosmic rays regularly bombard us with a million times higher-energy particles than our largest accelerators (see http://en.wikipedia.org/wiki/Oh-My-God_particle), and we haven’t been destroyed in ten billion years. So until we go above 10^20 eV/proton, rank this in the “very unlikely” category.
Thanks, Tom, for being handy with this detailed an array of specific info.
‘Antimatter weapons’ as normally conceived just aren’t a significant threat. They don’t do anything that fusion weapons can’t do, and they’re massively more expensive to make given for any plausible antimatter production strategy. For anything below a few megatons an antimatter bomb isn’t even likely to be lighter, as the extra mass of the containment mechanisms will likely exceed the savings in fuel mass.
What /is/ a significant proliferation threat is the ability to start a fusion chain reaction without the use of a fission trigger. Right now we’re close to being able to do this, but it takes a huge building full of equipment (i.e. the national ignition facility). Nanotechnology precision manufacturing and advanced materials may allow the production of laser or particle accelerator based systems capable of triggering a fusion bomb with desktop equipment. Refined uranium or plutonium is difficult to get hold of and building a stage fission-fusion weapon is a really tough physics and engineering problem (much harder than the making a simple fission device). With a non-nuclear trigger, pretty much all you need is a block of lithium deuteride (trivial to make or acquire) to make a fusion bomb of potentially unlimited size. That said, antimatter could still have a role, as antimatter can in principle act as a fusion trigger and making or acquiring enough antimatter to do this is a much more practical proposition than creating enough to have a useful direct yield.
“Right now we’re close to being able to do this,”
No, we’re not. ICF is making progress, but it doesn’t work by “chain reaction”; all of the pellet is compressed and heated by the laser system, not by other parts of the pellet that underwent fusion. Bigger pellets require a bigger laser system; bomb-sized ones would require a system around three orders of magnitude more powerful.
Lasers work by having all of their molecules release their energy at the same time in a chain reaction; if you want more energy, you need more molecules. If you want high power levels, you need to have all of the molecules ready to release their energy very quickly; but then some will release energy spontaneously, and then you have to have an enormous power system to continuously pump energy into the laser. No cheatin’ the laws of thermodynamics.
“block of lithium deuteride (trivial to make or acquire)”
Good luck trying to get one. I’ve been prowling on the Internet for months for potassium deuteride, a closely related chemical, and there’s simply no market for it. Chemical suppliers don’t stock lithium deuteride, and it requires industrial equipment to manufacture. Even lithium hydride is hard; Cole-Parmer has discontinued it (http://www.coleparmer.com/catalog/product_view.asp?sku=8820420)
Laser fusion for power applications doesn’t use a multi-stage reaction because there’s no desire to have a multi-megaton explosion going off in the reaction chamber. However from what I’ve read such reactions are quite possible without needing the elaborate radiation implosion system current bombs use; essentially you start fusion in a small initiator pellet (using heating and implosion) housed in a small cavity in the centre of the main fusion fuel block. The expanding fireball compresses and superheats the inner wall of the cavity, which fuses, expanding the fireball further and causing more fusion, and so on until the whole assembly is a fusing plasma. A precision ignition system does not need to compress and ignite the entire fusion fuel mass the way that a fission-fusion bomb does, though getting the size of the trigger pellet (and hence ignition energy requirement) down to an absolute minimum will certainly be tough engineering problem. Injecting somewhere around a milligram of antimatter into the centre of the lithium deuteride block should also work.
I don’t think manufacturing or stealing lithium deuteride will be tricky for any group with the ability to create or steal a laser fusion trigger or a milligram of antimatter.
I copied this to http://future.wikia.com/wiki/Existential_risk&action if you don’t mind.
Hmmm…not cool…any systematic thinking, yet, as to what we can do to minimize the threat of non-fission-triggered run-away fusion explosions?
“or a milligram of antimatter.”
Current antimatter production is measures in nanograms. Milligram quantities, even assuming vastly better efficiency, would be worth a ton of money and therefore well guarded. We may not guard our atom bombs effectively, but we’ll be damned if those thieves take a single cent of our profits!
“The expanding fireball compresses and superheats the inner wall of the cavity, which fuses,”
Most of the energy created by DT fusion is in the form of neutron radiation. Fast neutrons are VERY penetrating; a neutron beam will distribute its energy over large quantities of matter. If the pellet is too small, most of the neutrons will go right through, and if it’s too large, there won’t be enough energy to heat it effectively. So you need a pellet that’s at least one centimeter thick to absorb a substantial amount of neutrons, which means you need a centimeter-thick pellet to detonate it, which is several orders of magnitude larger than modern ones.
“Injecting somewhere around a milligram of antimatter into the centre of the lithium deuteride block should also work.”
That would actually work if the block was small enough relative to the antimatter. However, such a pellet of antimatter would, even using much better accelerators, take more energy to create than could be derived from fusion. Hell, we have enough trouble trying to reach breakeven; theoretical maximum for a complete burn is an EROEI of ~400; I am not aware of any nuclear reaction that produces antimatter with an efficiency of 1 in 400, even assuming all the other equipment is perfectly efficient.
I agree that even using antimatter as a fusion trigger is effectively impossible with near future technology and of dubious utility given any currently conceivable antimatter production technology (it might make sense for some starship drive concepts). I’m just saying that it’s a more plausible application than trying to use antimatter as an explosive directly.
Michael has noted the damage potential of orbital kinetic bombardment systems elsewhere in his blog (I’ve also commented on it before), but I’d note that cheap commercial space travel is effectively a proliferation risk for this class of technology. If you have the technology to put a sizeable space station in orbit, dropping an equivalent mass at a few km/s onto a city you don’t like doesn’t take much extra effort. Further out, the kind of tech Michael wants to use to build lifeboat stations and spacecraft is equally applicable to building lunar mass drivers and nudging asteroids into earth-collision trajectories. A MAD scenario in which near-earth space is filled with kinetic bombardment weapons just waiting for the command to deorbit is actually relatively plausible, compared to risks such as ‘particle accelerator disaster’ and ‘grey goo event’.
[...] We know a lot about the current threats, and yet we’re only marginally well-equipped to deal with them. We’ve had fusion weapons for fifty years without wiping ourselves out, but such weapons are making their way into the hands of less stable dictatorships. We’ve had both Clint Eastwood and Bruce Willis save us from being smashed by Hollywood space rocks, but we don’t have a real plan for detecting and destroying real ones. The flu could kill millions of us, but at least we’re talking about ways to mitigate an outbreak. The 20th century may have been the bloodiest ever, but when you think of the risks we faced, we didn’t do too badly. Enter the 21st century. It seems that the transhumanist community has been addressing the possible threats of emerging technology with more seriousness. This is a good thing. There are serious risks posed by nanotech, biotech, and advanced artificial intelligence. But how to enumerate and prepare for them? Accelerating Future is attempting to list and categorize the threats that accidents, military leaders, mad scientists, supervillians and disgruntled grad students will throw at us in the 21st century: The important ones: [...]
“A MAD scenario in which near-earth space is filled with kinetic bombardment weapons just waiting for the command to deorbit”
KE weapons are usable given cheaper launch technology, but they’re not particularly dangerous. A 100-ton impactor moving at 10 km/s delivers only 1.2 kilotons, some of which will be absorbed in the atmosphere. There won’t be any radioactive fallout, but the mass will ionize the atmosphere around it and so will create plenty of UV light and X-rays.
[...] Accelerating Future » Comprehensive List of Existential Risks, Part 2 1. superintelligence – not just AI – but superhumans too 2. deliberate misuse of nanotech (arms race, nanoweapons) 3. accidental misuse of nanotech 4. killer artificial virus 5. antimatter holocaust? 6. particle accelerator disaster (tags: future weapons transhumanism singularity space) [...]
[...] Existential Risks: The development of nuclear weapons marked a disturbing turning point for the human species: we are increasingly coming into the possession of apocalyptic technologies. Soon to join the list are such problems as a malevolent superintelligence, deliberate or accidental misuse of nanotech, runaway global warming, a killer artificial virus, an antimatter holocaust, or a particle accelerator disaster. Read more here and here. Adding insult to injury is the Doomsday Argument. [...]
I have to mention that you need not antimatter or laser to trigger termonuclear reaction. In fact simple cemical implosion cound create some amount of fussion, but enegicaly useful (10**13 neytrons). Also electric lightning could ignate some amount of fussion. If you connect cilindrical implosion with electric spark on axis and add some magnetic field to keep the cord of plasma… Of course it is not simple… but unfortunatelly possible.
Buy super cheap phentermine and tramadol….
Cheap tramadol. Tramadol cheap cod. Tramadol hci online buy cheap tramadol hci online. Buy online cheap and fast tramadol. Cheap tramadol fedex overnight. Cheap comment message preview tramadol….
Milf….
Milf sites. Milf movies. Milf seeker….
àâòîñòðàõîâàíèå ñòàòèñòèêà…
àâòîñòðàõîâàíèå ñòàòèñòèêà…
I was honored to receive a call from my friend immediately he found out the important points shared on your site. Going through your blog article is a real great experience. Many thanks for considering readers like me, and I hope for you the best of success being a professional discipline.
Really intriguing points you’ve got observed , thanks for posting .
Interested in start up a online business on line denotes revealing your service also providers not only to humans within your town, nevertheless , to numerous future prospects which are cyberspace on many occasions. pays day-to-day