• Google
    This Blog Web

October 2011

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          

RSS Feed

Bookmark and Share

Email Feed



  • Powered by FeedBlitz

« Refining Definitions | Main | Debating the Nanocosm »

February 21, 2004

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Crawford Kilian

Very interesting...I'm writing a nanotech novel too, though with a bit more positive spin. Hope it gets kinder attention than Marlow's!

By the way, have you reviewed William Atkinson's book Nanocosm here?

Brett Bellmore

Technically, you're right about an infestation which begins as a point infection, and spreads only by growth. A rogue nanotech device which spread like pollen, though... It could keep up exponential growth for quite a long time, and would be a real pain to eradicate, even if you did catch it early. You'd essentially have to have a widespread "immune system" already in place, which you'd only have to broadcast target recognition instructions to, and even then you'd probably only keep the threat in check, not totally eradicate it.

Chris Phoenix, CRN

I made a post over on sci.nanotech which Gordon hasn't responded to yet. I made the following points:

1) Gordon misread Freitas' paper. The numbers quoted by Gordon were best case, not worst case. Worst case is very bad.

2) Marlow's book does claim impossible abilities for his nanobots. But if he had written it with only plausible nanobot function... it wouldn't have changed the story much!

Chris

Chris Phoenix, CRN

"By the way, have you reviewed William Atkinson's book Nanocosm here?"

Not here, but I reviewed it for Nanotech-Now.com. That review led to a lengthy email discussion between me and Atkinson. Many good points were raised by both sides. In the end, he did make some concessions about the possibilities of the technology. http://www.nanotech-now.com/Atkinson-Phoenix-Nanotech-Debate.htm points to the original review and the discussion, with a summary.

Chris


John Robert Marlow

To respond to Dr. Pusch:

Among the many sourced quotes contained on my website (www.johnrobertmarlow.com) is the following: *A single nanoweapon or nanoaccident could "reduce the biosphere to dust in a matter of days" [8f], if not literally destroy the planet. [9f]* [Notes refer to website links, not this post.]

The source of the "matter of days" quote is Eric Drexler--who makes the same analogy chosen by Dr. Pusch (bacteria), but comes to a far different conclusion ("Tough, omnivorous "bacteria" could out-compete real bacteria: they could spread like blowing pollen, replicate swiftly, and reduce the biosphere to dust in a matter of days"--source: Engines of Creation, chapter eleven http://www.foresight.org/EOC/EOC_Chapter_11.html).

Thus--far from "paying too much attention to the overblown and scientifically inaccurate claims of doomsday science fiction writers (many of whom are scientifically illiterate luddites)," as Dr. Pusch suggests--I was in fact relying on what Eric Drexler had said. Which is, as it turns out, also what Robert Freitas says in the very paper cited by Dr. Pusch in support of his own contention that such an event would take days or weeks (in the case of a city) or two years (in the case of the biosphere) to transpire.

While it is true (as stated more than once in the NANO novel) that the maximal rate of nanoswarm expansion cannot be sustained beyond the first instant (else the earth and, indeed, the entire universe would be consumed in short order)--maximal expansion throughout the swarm is not required in order to destroy a city in minutes (which is what Dr. Pusch is apparently referring to here when he states: "the response time required to deal with a rogue nanotech infestation will _NOT_ be on the order of "minutes" as you claim, but on the order of days or weeks -- i.e., comparable to the response time required to deal with an epidemic outbreak of a typical infectious disease").

In his paper Some Limits to Global Ecophagy by Biovorous Nanoreplicators, with Public Policy Recommendations: http://www.foresight.org/NanoRev/Ecophagy.html), Freitas holds that the limiting factor which determines the maximum dispersal velocity of small nanites ("biovorous nanoreplicators") is wind speed-which he cites as being (global mean wind speed) ~10 meters per second. This yields a swarm expansion rate of 1 kilometer per 100 seconds. A quick look at a map will reveal that San Francisco (a rather large city, and site of the climactic scene in the NANO novel) would, at that rate, be doomed in minutes and not--as stated by Dr. Pusch--hours or days. Furthermore, wind speeds in coastal cities can be considerably higher than average, making the figure just cited conservative. It should also be noted that Freitas' figure applies only to nanites which have NOT been specifically designed to maximize airborne dispersal rates ("isolated replibots lacking significant aeromotive capabilities").

Next, Dr. Pusch asserts that "Dr. Robert Frietas has shown that the maximum possible "biomass to nanobot" conversion rate of self-replicating nanobots cannot possibly be greatly larger than that of already-existing self-replicating organic lifeforms such as bacteria, so that far from "destroy[ing] the planet in a matter of days," the timescale for complete "ecophagy" (conversion of the entire biosphere into nanobots) is in fact quite long -- on the order of _TWO YEARS_, not two days."

Actually, Freitas' paper says nothing of the kind. On the contrary, the paper specifically states (just below Table 4) that (as CRN co-founder Chris Phoenix pointed out in an earlier response to Dr. Pusch's post) "a moderately efficient goo, not trying to avoid detection, could in theory convert the biosphere in 2.44 days."

This with energy dissipation at 10MJ/kg. Freitas also points out that both Eric Drexler and Ralph Merkle have conjectured that efficiencies on the order of 0.1 MJ/kg may be possible in some circumstances, and refers to this as "a possibility that should be investigated in the present context in a future theoretical study." Such efficiencies would make 2.44 days seem an eternity. Further, this is the time required to convert the entire biosphere--and as Chris has pointed out to me, "if we reach the 2% goo stage globally, or even the 0.02% stage, we're probably already doomed."

Freitas' paper largely concerns itself with describing what is-in comparison with what is possible-slow-motion goo which expands at a limited rate in order to postpone swarm detection. The scenario presented in NANO, Superswarm, and the interview cited by Dr. Pusch, on the other hand, deals with the maximum possible rate of swarm expansion.

Or does it? For, as Chris also pointed out to me, swarming omnivorous nanites which convert the environment to goo (and copies of themselves) may not be the most dangerous or the most swiftly-spreading nanohazard. That dubious honor may belong to something which simply destroys things and moves on without massive self-replication.

In any event, the scenario employed in NANO also involves a seed AI which is involved in the design, production, and dispersal of nanites (this is the "wildcard" to which Chris refers in his earlier post). Needless to say, a true AI will rapidly determine and implement maximal efficiencies which humans may be incapable of determining, much less implementing, in any meaningful timeframe.

I'd like to conclude by saying that Dr. Freitas' paper appeared after the initial draft of NANO was completed (it takes a LONG time to get a book published); nonetheless, even if the specific technical details of the rogue swarms present in novel, superswarm paper, and interview are in some way incorrect-it seems clear that the same (and perhaps worse) dangers are well within (to borrow a phrase) the limits of the possible. (Chris Phoenix: "I don't think it would change the book much if he had written it to conform to reasonable physical limitations. Less chemistry, more micro-scale robotics... and you'd get much the same impact.")

When it comes to novels--tech-thrillers in particular--the practice which makes for the most exciting read and the broadest appeal is to push things to the limits of the possible. That doesn't mean what is possible today; it does mean engaging in some degree of reasonable extrapolation.

Scientific papers, while important, are read by few; novels are read by many. My aim in writing NANO was and remains to create a greater public awareness of nanotechnology and its staggering potential to benefit Mankind--and to destroy us. I think everyone here can probably agree that increased awareness of the pros and cons is a good thing, and it is my profound hope that this novel will accomplish that.

Readers whose interest is sparked by the book are directed by it to my website (www.johnrobertmarlow.com)--which links to a growing number of nanotechnology resources (including sci.nanotech and the Center for Responsible Nanotechnology), which readers can explore to learn more, form their own opinions--and act on them.

That-and not which specific details of a future technology will yield the result Freitas, Drexler, Merkle, Chris Phoenix and others all seem to agree is possible-is, in my opinion, what matters most in the context of a work of plausible fiction.

---John Robert Marlow

www.johnrobertmarlow.com

Brett Bellmore

Indeed, the key question is not, "How fast would grey goo FINISH destroying the ecosystem?", it's, "How fast would grey goo grow to be too powerful a threat for the defenses in place to cope?" Because once that has happened, the destruction of the ecosystem is a forgone conclusion, however long it takes.

That key time is a function of how fast the goo can spread and grow, (If it grows rapidly, and spreads slowly, if worst came to worst you could nuke it.) how much of it there is before it's detected, the response time of your defenses, and the power of your defenses.

I have concluded that, in order for a defense against grey goo to be feasible, the defense has to amount to a not insignificant fraction of the biomass of the planet, and be omni-present. Before the goo appears, any enviroment which provides the resources it needs to grow must ALREADY be permeated by the defense system, so that the threat can be detected in a very short time, and face very nearly instant response.

An accidental grey goo, which I agree is fairly unlikely, could probably be coped with by a lesser system. A malicious deliberate grey goo is another matter; It would probably be designed for covert growth to some critical size, followed by worldwide dispersal, and a fast growth phase. It would be a VERY tough opponent.

Which is why I have also concluded that it's the near inevitable fate of our ecosystem, aside from specially protected pockets, to suffer "ecophagy". Nanotech is going to all but completely replace DNA/protein life eventually. The only real question is whether it will be a controlled process which leaves a great deal of natural diversity in place, or the grey stuff.

In other words, if you can't beat 'em, join 'em. The ultimate defense against grey goo is to become nanotechnolgy ourselves.

Janessa Ravenwood

But Brett, that's means we'll all have to become government property - at least if CRN were to get it's way. Or should I say, the property of their international agency that owns all nanotechnology in real-time.

I personally aim to have an internal, integrated nanofac in my body. Hence, one of the many reasons CRN's proposals scare me is I really don't want to be owned by some regulatory agency.

The comments to this entry are closed.