LA-602 vs. RHIC Review

LA-602: Ig­ni­tion of the At­mo­sphere with Nu­clear Bombs, a re­search re­port from the Man­hat­tan Pro­ject, is to the best of my knowl­edge the first tech­ni­cal anal­y­sis ever con­ducted of an un­cer­tain dan­ger of a hu­man-caused ex­tinc­tion catas­tro­phe.

Pre­vi­ously, Tel­ler and Konopin­ski had been as­signed the task of dis­prov­ing a crazy sug­ges­tion by En­rico Fermi that a fis­sion chain re­ac­tion could ig­nite a ther­monu­clear re­ac­tion in deu­terium—what we now know as an H-Bomb. Tel­ler and Konopin­ski found that, con­trary to their ini­tial skep­ti­cism, the hy­dro­gen bomb ap­peared pos­si­ble.

Good for their ra­tio­nal­ity! Even though they started with the wrong con­clu­sion on their bot­tom line, they were suc­cess­fully forced away from it by ar­gu­ments that could only sup­port one an­swer.

Still, in ret­ro­spect, I think that the ad­vice the fu­ture would give to the past, would be: Start by sit­ting down and say­ing, “We don’t know if a hy­dro­gen bomb is pos­si­ble”. Then list out the ev­i­dence and ar­gu­ments; then at the end weigh it.

So the hy­dro­gen bomb was pos­si­ble. Tel­ler then sug­gested that a hy­dro­gen bomb might ig­nite a self-sus­tain­ing ther­monu­clear re­ac­tion in the ni­tro­gen of Earth’s at­mo­sphere. This also ap­peared ex­tremely un­likely at a first glance, but Tel­ler and Konopin­ski and Marvin in­ves­ti­gated, and wrote LA-602...

As I un­der­stand LA-602, the au­thors went through the math and con­cluded that there were sev­eral strong rea­sons to be­lieve that ni­tro­gen fu­sion could not be self-sus­tain­ing in the at­mo­sphere: it would take huge en­er­gies to start the re­ac­tion at all; the re­ac­tion would lose ra­di­a­tion from its sur­face too fast to sus­tain the fu­sion tem­per­a­ture; and even if the fu­sion re­ac­tion did grow, the Comp­ton effect would in­crease ra­di­a­tion losses with vol­ume(?).

And we’re still here; so the math, what­ever it ac­tu­ally says, seems to have been right.

Note that the Man­hat­tan sci­en­tists didn’t always get their math right. The Cas­tle Bravo nu­clear test on March 1, 1954 pro­duced 15 mega­tons in­stead of the ex­pected 4-8 mega­tons due to an un­con­sid­ered ad­di­tional nu­clear re­ac­tion that took place in lithium-7. The re­sult­ing fal­lout con­tam­i­nated fish­ing boats out­side the de­clared dan­ger zone; at least one per­son seems to have died.

But the LA-602 calcu­la­tions were done with very con­ser­va­tive as­sump­tions, and came out with plenty of safety mar­gin. AFAICT (I am not a physi­cist) a Cas­tle Bravo type over­sight could not re­al­is­ti­cally have made the at­mo­sphere ig­nite any­way, and if it did, it’d have gone right out, etc.

The last time I know of when a ba­sic phys­i­cal calcu­la­tion with that much safety mar­gin, and mul­ti­ple an­gles of ar­gu­ment, turned out to be wrong any­way, was when Lord Kelvin showed from mul­ti­ple an­gles of rea­son­ing that the Earth could not pos­si­bly be so much as a hun­dred mil­lion years old.

LA-602 con­cludes:

“There re­mains the dis­tinct pos­si­bil­ity that some other less sim­ple mode of burn­ing may main­tain it­self in the at­mo­sphere… the com­plex­ity of the ar­gu­ment and the ab­sence of satis­fac­tory ex­per­i­men­tal foun­da­tions makes fur­ther work on the sub­ject highly de­sir­able.”

Decades af­ter LA-602, an­other pa­per would be writ­ten to an­a­lyze an un­cer­tain dan­ger of hu­man-cre­ated ex­is­ten­tial risk: The Re­view of Spec­u­la­tive “Disaster Sce­nar­ios” at RHIC.

The RHIC Re­view was writ­ten in re­sponse to sug­ges­tions that the Rel­a­tivis­tic Heavy Ion Col­lider might cre­ate micro black holes or strangelets.

A B.Sc. the­sis by Shameer Shah of MIT, Per­cep­tion of Risk: Disaster Sce­nar­ios at Brookhaven, chron­i­cles the story be­hind the RHIC Re­view:

The RHIC flap be­gan when Walter Wag­ner wrote to Scien­tific Amer­i­can, spec­u­lat­ing that the Brookhaven col­lider might cre­ate a “mini black hole”. A re­ply let­ter by Frank Wilczek of the In­sti­tute for Ad­vanced Stud­ies la­beled the mini-black-hole sce­nario as im­pos­si­ble, but also in­tro­duced a new pos­si­bil­ity, nega­tively charged strangelets, which would con­vert nor­mal mat­ter into more strange mat­ter. Wilczek con­sid­ered this pos­si­bil­ity slightly more plau­si­ble.

Then the me­dia picked up the story.

Shameer Shah in­ter­viewed (on Nov 22, 2002) Robert Jaffe, Direc­tor of MIT’s Cen­ter for The­o­ret­i­cal Physics, a pi­o­neer in the the­ory of strange mat­ter, and pri­mary au­thor of the RHIC Re­view.

Ac­cord­ing to Jaffe, even be­fore the in­ves­tiga­tive com­mitte was con­vened, “No sci­en­tist who un­der­stood the physics thought that this ex­per­i­ment posed the slight­est threat to any­body.” Then why have the com­mit­tee in the first place? “It was an at­tempt to take se­ri­ously the fears of sci­ence that they don’t un­der­stand.” Wilczek was asked to serve on the com­mit­tee “to pay the wages of his sin, since he’s the one that started all this with his let­ter.”

Between LA-602 and the RHIC Re­view there is quite a differ­ence of pre­sen­ta­tion.

I mean, just look at the names:

LA-602: Ig­ni­tion of the At­mo­sphere with Nu­clear Bombs
Re­view of Spec­u­la­tive “Disaster Sce­nar­ios” at RHIC

See a differ­ence?

LA-602 be­gan life as a clas­sified re­port, writ­ten by sci­en­tists for sci­en­tists. You’re as­sumed to be fa­mil­iar with the mean­ing of terms like Bremsstrahlung, which I had to look up. LA-602 does not be­gin by as­sert­ing any con­clu­sions; the re­port walks through the calcu­la­tions—at sev­eral points clearly la­bel­ing the­o­ret­i­cal ex­trap­o­la­tions and un­ex­plored pos­si­bil­ities as such—and fi­nally con­cludes that ra­di­a­tion losses make self-sus­tain­ing ni­tro­gen fu­sion im­pos­si­ble-ac­cord­ing-to-the-math, even un­der the most con­ser­va­tive as­sump­tions.

The RHIC Re­view pre­sents a non­tech­ni­cal sum­mary of its con­clu­sions in six pages at the start, rel­e­gat­ing the math and physics to eigh­teen pages of ap­pen­dices.

LA-602 con­cluded, “There re­mains the dis­tinct pos­si­bil­ity that some other less sim­ple mode of burn­ing may main­tain it­self in the at­mo­sphere...”

The RHIC Re­view con­cludes: “Our con­clu­sion is that the can­di­date mechanisms for catas­trophic sce­nar­ios at RHIC are firmly ex­cluded by ex­ist­ing em­piri­cal ev­i­dence, com­pel­ling the­o­ret­i­cal ar­gu­ments, or both. Ac­cord­ingly, we see no rea­son to de­lay the com­mis­sion­ing of RHIC on their ac­count.”

It is not ob­vi­ous to my in­ex­pert eyes that the as­sump­tions in the RHIC Re­view are any more firm than those in LA-602 - they both seem very firm—but the two pa­pers arise from rather differ­ent causes.

To put it bluntly, LA-602 was writ­ten by peo­ple cu­ri­ously in­ves­ti­gat­ing whether a hy­dro­gen bomb could ig­nite the at­mo­sphere, and the RHIC Re­view is a work of pub­lic re­la­tions.

Now, it does seem—so far as I can tell—that it’s pretty damned un­likely for a par­ti­cle ac­cel­er­a­tor far less pow­er­ful than ran­dom cos­mic rays to de­stroy Earth and/​or the Uni­verse.

But I don’t feel any more cer­tain of that af­ter read­ing the RHIC Re­view than be­fore I read it. I am not a physi­cist; but if I was a physi­cist, and I read a predi­gested pa­per like the RHIC Re­view in­stead of do­ing the whole anal­y­sis my­self from scratch, I would be fun­da­men­tally trust­ing the ra­tio­nal­ity of the pa­per’s au­thors. Even if I checked the math, I would still be trust­ing that the equa­tions I saw were the right equa­tions to check. I would be trust­ing that some­one sat down, and looked for un­pleas­ant con­trary ar­gu­ments with an open mind, and re­ally hon­estly didn’t find any­thing.

When I con­trast LA-602 to the RHIC Re­view, well...

Don’t get me wrong: I don’t feel the small­est par­ti­cle of real fear about par­ti­cle ac­cel­er­a­tors. The ba­sic cos­mic-ray ar­gu­ment seems pretty con­vinc­ing. Na­ture seems to have ar­ranged for the calcu­la­tions in this case to have some pretty large er­ror mar­gins. I, my­self, am not go­ing to worry about risks we can ac­tu­ally calcu­late to be tiny, when there are in­calcu­la­ble large-look­ing ex­is­ten­tial risks to soak up my con­cern.

But there is some­thing else that I do worry about: The pri­mary stake on the table with things like RHIC, is that it is go­ing to get sci­en­tists into the habit of treat­ing ex­is­ten­tial risk as a pub­lic re­la­tions is­sue, where ig­no­rant techno­phobes say the risk ex­ists, and the job of sci­en­tists is to in­ter­face with the pub­lic and ex­plain to them that it does not.

Every­one knew, be­fore the RHIC re­port was writ­ten, what an­swer it was sup­posed to pro­duce. That is a very grave mat­ter. Anal­y­sis is what you get when physi­cists sit down to­gether and say, “Let us be cu­ri­ous,” and walk through all the ar­gu­ments they can think of, record­ing them as they go, and fi­nally weigh them up and reach a con­clu­sion. If this does not hap­pen, no anal­y­sis has taken place.

The gen­eral rule of thumb I some­times use, is that—be­cause the ex­pected util­ity of thought arises from the util­ity of what is be­ing rea­soned about—a sin­gle er­ror in an­a­lyz­ing an ex­is­ten­tial risk, even if it “doesn’t seem like it ought to change the con­clu­sion”, is worth at least one hu­man life.

The RHIC Re­view is not writ­ten to the stan­dard of care that would be ap­pro­pri­ate if, af­ter the RHIC Re­view was writ­ten, some higher au­thor­ity went through the pa­per; and if a sin­gle ar­gu­ment in it was wrong, any­where, whether or not it changed the con­clu­sion, a hostage got shot. That’s how to think about an­a­lyz­ing ex­is­ten­tial risks. That way, for each and ev­ery el­e­ment of the anal­y­sis, you can find it in your­self to be a lit­tle un­cer­tain about that el­e­ment, even if it doesn’t seem “like it could pos­si­bly change the con­clu­sion”; un­cer­tainty in­vokes cu­ri­os­ity.

The RHIC Re­view was pro­duced by au­thors who were already sure that the RHIC couldn’t de­stroy the Earth, the prob­lem-at-hand was ex­plain­ing this to the pub­lic. If the au­thors de­cided just by eye­bal­ling the prob­lem that the RHIC couldn’t de­stroy the Earth, then the only ac­tual anal­y­sis that took place was con­ducted in 5 sec­onds. Yes, it’s a lop­sided is­sue, but it seems that as a gen­eral mat­ter of policy, any ex­is­ten­tial risk at all de­serves a longer and truly cu­ri­ous anal­y­sis than that.

Though I don’t re­ally blame the RHIC Re­view’s au­thors. No one ever told them that there was such a thing as ex­is­ten­tial risk, or that it raised the stan­dards of anal­y­sis be­yond what was usual in a sci­en­tific pa­per, or that ra­tio­nal anal­y­sis re­quires plac­ing your­self into a state of gen­uine un­cer­tainty about each in­di­vi­d­ual el­e­ment’s ex­act value...

And the much greater rea­son I don’t blame them, is that be­tween the 1940s and to­day, so­ciety has de­vel­oped a “Gotcha!” at­ti­tude to­ward risk.

You can’t ad­mit a sin­gle par­ti­cle of un­cer­tain dan­ger if you want your sci­ence’s fund­ing to sur­vive. Th­ese days you are not al­lowed to end by say­ing, “There re­mains the dis­tinct pos­si­bil­ity...” Be­cause there is no de­bate you can have about trade­offs be­tween sci­en­tific progress and risk. If you get to the point where you’re hav­ing a de­bate about trade­offs, you’ve lost the de­bate. That’s how the world stands, nowa­days.

So no one can do se­ri­ous anal­y­sis of ex­is­ten­tial risks any­more, be­cause just by ask­ing the ques­tion, you’re threat­en­ing the fund­ing of your whole field.

The num­ber one les­son I take from this whole is­sue is that where hu­man-caused un­cer­tain ex­is­ten­tial dan­gers are con­cerned, the only way to get a real, se­ri­ous, ra­tio­nal, fair, even­handed as­sess­ment of the risks, in our mod­ern en­vi­ron­ment,

Is if the whole pro­ject is clas­sified, the pa­per is writ­ten for sci­en­tists with­out trans­la­tion, and the pub­lic won’t get to see the re­port for an­other fifty years.

This is the les­son of LA-602: Ig­ni­tion of the At­mo­sphere with Nu­clear Bombs and the Re­view of Spec­u­la­tive “Disaster Sce­nar­ios” at RHIC. Read them and weep.