Virtual Nuclear Weapons Design and the Blur of Reality | Hacker News Reader

Virtual Nuclear Weapons Design and the Blur of Reality

thereader.mitpress.mit.edu

50 pointschmaynard47 comments

Nuclear weapons and MAD provide the best examples of double-think that I can appreciate, and that I even partially buy into. Reading articles like this, I want to jump to judgement and outrage over the physicists and engineers who, I imagine, spend many more hours considering incremental technical improvements than contemplating the reality of these devices. But if I could press a button to relinquish my country's entire nuclear arsenal, would I? Would I rather live in a more morally pure country in that respect, but one which no longer has the safety benefits of mutually assured destruction? If not, then am I just a hypocrite for criticising those with direct involvement?

The logic of nuclear deterrence depends on having an amoral chain of command. Nuclear retaliation against Russia, China, or the United States is suicide, and serves no beneficial purpose in the moment it is carried out. But there is no way to fool an adversary into believing that we are capable of total retaliation without actually putting the command structures and individuals in place which are somewhat likely to follow through. Every time I seriously ponder nuclear war, these questions enter my mind and I can't make headway. Much of it boils down to utilitarianism versus more idealistic thinking.

You might appreciate this panel discussion featuring Carl Sagan and Henry Kissinger:

https://physicsworld.com/a/the-day-after-35-years-later-carl...

I’ve always been struck by the caliber of this broadcast and the audience engagement and questions. From a different time when public discourse on nuclear weapons was common.

Perhaps part of the rising anxiety in society over the past decades is the suppression of the knowledge that the we live with loaded guns pointed at our heads. Sounds like an almost unhinged thing to say — but then again, the policy is called MAD after all.

One of my first jobs out of college was working on simulation code dealing with weapons effects, incl. nuclear weapons.

I know I personally struggled with the ethics of it, but not all colleagues did. To some it was patriotism, to others it was an interesting problem, to yet some others it was just a job.

I learned a lot listening to the effects guys that were around during the live testing. Makes the 2020's seem tame compared to what was going on during the Cold War.

The best case scenario would be if no nukes worked, but everyone believed they did. Unfortunately getting there is essentially impossible. At the moment the US is responsible for most "stockpile stewardship" research, the fruits of which are ultimately disseminated to other nuclear powers. If we were to stop spending billions forging ahead in that area of research, it might make the world's nukes less viable. Since the US has the most by number, our relative advantage would still remain.

And the worst case scenario is if someone finally figured out they didn’t.

The USA tried to give flawed nuclear bomb plans to Iran in a bid to slow down their nuclear weapons program, but it backfired and Iran was informed that the plan had purposeful flaws. It is debatable whether it wound up accelerating the Iranian nuclear program.

Also, the threat of the nuclear bomb is what finally led to the ceasefire in the Korean War. Both North Korea and the USA came in thinking they were the "victors" and that the other side needed to sue for peace. Truman didn't want to threaten the use of the nuclear bomb, having been the one to drop two on Japan. When Eisenhower assumed Presidency, he threatened the use of nuclear weapons, bullying North Korea in the process. Hence, the regime became obsessed with obtaining one so that it could never be bullied again. It viewed it as the prime ticket to true independence/security.

So the USA shot itself in the foot twice when it comes to "modern nations obsessed with nuclear weapons programs that are not friendly to the USA".

>It viewed it as the prime ticket to true independence/security.

It's true though. Nuclear weapons with delivery mechanisms are about the only thing in the modern world which will make others back off. In every other case you're relying on the good graces of others.

North Korea managed to survive up until 2006 without nukes, so obviously there are other means.

At the same time, there are plenty of military planners that don't view nukes as a definitive reason not to pick a fight considering the large amounts of effort they put into planning a nuclear war.

Not every nation can be "lucky" and have a city of several million people located in a nation that few other nations hate yet still conveniently within artillery range.

From what I've read, it was Gaddafi's death in 2011 (after he gave up his very limited nuclear weapons program) that made the North Korean regime realize it absolutely needed a nuclear deterrent.

The Libya adventure probably served to settle any debate within North Korea, but their program is much older and they have shown serious committent to it much earlier than 2011.

In 1994 it was discovered they had reprocessed plutonium several times in the 80's, and it was suspected they had secret reprocessing facilities, and they almost left the Non-Proliferation of Nuclear Weapons treaty to keep inspectors out of those sites.

In 2006 they tested a nuclear device for the first time, although it likely failed.

https://www.nti.org/learn/countries/north-korea/nuclear/

The North Koreans didn't become obsessed with getting nukes in 1953.

Fred Saberhagen's Empire of the East novels proposed a super weapon ARDNEH which did just that, but being fantasy both sides unleashed their own versions and the effect became permanent.

It's kinda interesting (although probably more of a simple slippery slope) to consider the intersection of this and the classic Reflections On Trusting Trust[1].

[1]: https://www.cs.cmu.edu/~rdriley/487/papers/Thompson_1984_Ref...

> With explosions taking place virtually, how much harder will it be for weapons scientists to confront the destructive power of their work and its ethical implications?

What exactly is her hoped response, once a physicist realizes by watching a live atomic bomb explosion, that these things actually hurt? They'll quit working there? That didn't occur with the first few generations of bomb physicists. Enough stuck around to produce bomb after bomb, both here and abroad.

If anyone needs to learn this lesson, it's the proliferators -- except they'd just be overjoyed at the fearsome might, not terrified. Tehran and Pyongyang won't be deterred by feeling the dreadful rumble of a live test. (I'd pay to watch one, though.)

There may be dark side effects of virtualization, but this article's premise is weak.

I think it was Dan Carlin who suggested it might be a good idea to (every decade or so) conduct a live nuclear test above ground and publicize it widely. Not to make sure the bomb worked - but to remind everyone of the destructive power of these devices

a good idea, but would that not be significantly harmful for the environment, or is the radiation temporary?

Yes somewhat harmful to the environment but not as harmful to the environment as a nuclear war.

I feel we as a species need to get better at weighing up the costs of our action or inaction, making hard choices rather than everyone just fighting blindly for one half of the equation. /rant

Unfortunately, we've set up all our systems - governance, markets - to work through fighting. Politics, business and negotiations all work the same way: every side fights for their immediate short-term interest. One team wins this battle, other team wins that battle, and it goes on and on, until the cost of fighting outweighs any marginal improvement any side could get - and a "compromise" is reached.

I think it all comes down to the fact that we don't trust each other at scale. Because I can't trust that you'll approach searching for solutions honestly, that you'll optimize globally, I have to fight for my own interest at expense of rationality. For the same reason, you end up fighting for yours.

It's ridiculously wasteful, and I think most of humanity's problems stem from it. I can't think of a way out of it, at scale.

Bombs can be "tuned" to burn nearly all or very little of their fissable material. Radiation from the blast dissipiates nearly immediately, but the bomb would need to be designed to provide little fallout(it's doable is what I'm trying to say).

Amount of fallout depends on more things, not just settings of the bomb. Mainly on where the bomb detonates. If close to ground/water, then it's bad, lots of fallout. If in stratosphere, very little fallout.

If you burn less, you lower the yield. Burning more means you turn more mildly-radioactive plutonium or U235 into intensely radioactive fission products. Strontium-90, for example, accumulates in bones and has a half-life of 28 years.

The general rule of thumb is that the radiation in the fallout from a bomb goes down by a factor of 10 for every factor of 7 the time since the detonation increases. So it's 10 times less radioactive after a week as it was after a day. 100 times after a month and a half, 1000 times less after a year, etc.

Under the linear-no-threshold dose model we'd expect several death from fallout induced cancer per test. But on the other hand we know that the linear-no-threshold model is false. If you give some fruit flies a 50 REM dose of xrays all at once they get cancer. If you give them 50 REMs spread out over the course of a month they don't. But we don't really know what the actual model is very well so that's not a risk I'd be happy to take.

It sounds like a good idea, but I think it'll backfire. You'd think that would make people want less nukes, but ultimately, it would make people want more nukes to protect themselves from others with nukes.

After hiroshima and nagasaki, the soviet union, britain, france, china, etc didn't say, "wow that's destructive, lets have less of that". They said, "wow that's destructive" and "we need some".

I think if we did that openly and showed the world the destructive power of nukes every decade, more and more nations would want to get it and we'd have more nuclear powers.

Exactly I can certainly see how a certain president would want an even bigger one than any weapon displayed last by another nation.

The same Dan Carlin mentioned above also pointed out how for the brief period where the USA had the bomb and the USSR did not we bossed them around by threatening to bomb them.

There's plenty of incentive to want the bomb.

Normalising nuclear testing is a terrible idea! This is simply seen by unpacking the blandly neutral statement "nuclear testing" to the more accurate "practice mass murder by use of large explosion using a complicated device".

I don't see how normalizing or not normalizing it will have any effect whatsoever on its proliferation. No country acquires nuclear weapons to actually use them. Instead, it is (rightly) viewed as a great deterrent/insurance against being invaded and overrun.

That requires an intention to use them first against that invading nation.

Also this use would likely not be against the invading forces themselves because you'd be contaminating your own land so the use of nukes as protection against conquest requires not only the willingness to use them first but to use them on your enemy's homeland, likely military industrial targets which will also hit civilians.

That's immaterial, because your potential adversary isn't going to want to find out what your nuclear rules of engagement really are. Possessing nuclear weapons raises the threshold for putting the invasion on the table quite dramatically. And if there is a war, nuclear weapons limit its scope quite quickly.

Just think of how many Indian and Pakistani lives have been saved because the two countries stepped back when shit was about to get real. Like, three times now. All because of nukes. Unless one of them truly commits to invading and losing millions of lives, their wars are guaranteed to peter out as border skirmishes.

During the cold war, it was much debated whether 'limited nuclear war' was possible - for example, if Soviet tanks entered France and France used only small nuclear shells, only those smaller than a large conventional munition, only against military targets and only within their own borders; would that inevitably trigger an escalation to mutually assured destruction or not?

People who thought it would be tremendously convenient [0] if limited nuclear war was possible generally wanted to remove the taboo between large conventional explosions and small nuclear explosions, and so would support programs like the US "Project Plowshare" [1] and the Soviet "Nuclear Explosions for the National Economy" [2] which used nuclear weapons for mining and civil engineering.

With that said, one might very well argue that we are no longer in the cold war, and that our policies no longer need to be guided by the prospect of Soviet tanks entering France. And whether a once-per-decade reminder explosion would erode or enhance the taboo against their use in war.

[0] The WW2 Red Army had fielded more troops than the entire male population of France so France could not win a conventional war. [1] https://en.wikipedia.org/wiki/Project_Plowshare [2] https://en.wikipedia.org/wiki/Nuclear_Explosions_for_the_Nat...

IMO it would turn into a massive tourist attraction and lose the intended effect

“Do it for the ‘gram”, indeed.

My first intuition is how much confidence do you have that a new device that no one has ever used will behave the way the computer says it will.

Nuclear weapons are deterrents.. The goal is for them never to be used. If they didnt actually work at all that would be great. That would just be a layer of safety in case in some catastrophic turn of events they accidentally got used.

>> That would just be a layer of safety in case in some catastrophic turn of events they accidentally got used.

Well, not really, if one of the superpowers sees ICBMs flying their way, they will retaliate immediately, no one will wait around to see if they actually explode or not.

I think the point was that if the retaliation ICBMs fail in the same way, we as a species get a second chance to resolve the conflict without extinction.

The deterrent is only as strong as peoples' belief in it. If the weapons were all faulty, and mere suggestion of it leaked out to world leaders, we'd be back in pre-MAD days in a second.

I mean, just as an example, literally the entire landing sequence for the Curiosity rover we sent to mars hasn't been tested in reality before the actual mission. NASA could prove that it "should" work but no actual proof could be produced(especially the sky crane part hasn't been tested), and it was a mission that has nearly deplated our entire stock of plutonium, so you'd think there would be some way to make sure this incredibly complex landing sequence will actually you know, work.

The Castle Bravo test is a good example of that. More than double the expected yield, simply because our model of Lithium 7 was inaccurate.

https://en.m.wikipedia.org/wiki/Castle_Bravo

Big errors like this have all been tested and modeled away, but the problem is exceedingly complex, and I am just as certain enough uncertainty remains in the current models to be a genuine concern.

I must be missing something about this.

I mean, we have a bunch of designs that were tested and then mass produced. Many thousands of them, altogether.

They do need regular maintenance, of course. Such as replacing stuff that decayed, and stuff that got damaged by the radiation, or just air oxidation and whatever.

So what's the "designing" for? I find it hard to imagine that they'd be making substantive changes without any actual testing. Because they'd have no way to really know.

And conversely, I find it hard to imagine why we need to improve them. They're already too powerful to actually use on the Earth.

So doesn't the increased uncertainty outweigh some marginal increase in efficiency or power or whatever?

You might want to improve the designs to lengthen the shelf-life, to enable the use of smaller amounts of nuclear and conventional explosives, to increase efficiency and reduce the fallout created and so on. There are many more attributes to nuclear weapons than just how big of a bang they make.

Sure, I get that.

But wouldn't folks feel pretty stupid if they'd tweaked their weapons enough that they no longer worked?

The point of article isn't so much about physical vs. virtual testing, but about blindly trusting inherited code.

Some people freak out when they are unable to examine and understand internal working of used libraries, while others don't get it at all and just trust them. For me it is a mystery.

> “physical intuition is a skill you want to keep" (or build if necessary)

In whatever field of work, THIS is most critical.

Grounding in the hard reality.

Without it, you literally don't know what you are missing

With it, you can easily see things that others miss, simply because you stand at a better point of view.

I find the idea that today we still have to worry about an en masse invasion from a random blob of the same planet and specie population quite an aberration... And more disturbing it all starts usually from single individuals....you want to start a war? Good luck! Good for you! should be the answer of the rest of the population, why it doesn't happen? Lack of education, and the perception of unlawfulness? Armed enforcement? It's beyond me...

Reply on news.ycombinator.com