Hacker News Reader

Comment by photochemsyn | Hacker News Reader

I like how they don't even try to pretend that this is a route to practical power generation, it's all about research into the fundamental physics of fusion - which is a worthy goal in itself.

> The record shot was a major scientific advance in fusion research, which establishes that fusion ignition in the lab is possible at NIF,” said Omar Hurricane, chief scientist for LLNL’s inertial confinement fusion program. “Achieving the conditions needed for ignition has been a long-standing goal for all inertial confinement fusion research and opens access to a new experimental regime where alpha-particle self-heating outstrips all the cooling mechanisms in the fusion plasma."

View on HN

My understanding is that one big reason so much money was invested in this lab is because the research has direct applications to nuclear weapon design. It's a dual use lab so to speak. The process they are studying is very similar to what's happening in the core of a thermonuclear bomb.

> 1978: This report reviews aspects of the military applications of the inertial confinement fusion (ICF) program at Sandia Laboratories

https://www.osti.gov/biblio/6412035

> Today, research on inertial confinement fusion—the other leading approach—remains largely under the control of US national weapons labs. The military focus has had profound impacts on the development of inertial fusion energy.

https://thebulletin.org/2013/07/nuclear-weapons-the-death-of...

Better they keep a cadre of smart people who can do this stuff than lose the institutional knowledge.

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

Or one could choose to not do any further nuclear weapons research.

I understand that is not the marching orders of the US government but it could be, and it could also be a choice made by individuals on the ground.

I mean, all that game theory that the RAND corporation (and whatever their Soviet counterparts were called) seems to have worked out so far.

I think you’d need some pretty well-qualified people to crunch the numbers before just, winging it with something different. Given the stakes and all.

It worked out only because Stanislav Petrov decided to ignore it all and violate orders.

Which is kind of the point. Everyone involved in the process from the leader at the top making the call, to the individuals manning the silos, has to make a conscious decision to end the world.

No, the point of mutually assured destruction is to ensure that retaliation is credible and even perhaps automated.

You seem to have it backwards. The only reason we are alive is because the designed system that was built failed following a computer error. The point was to respond.

It's observer selection bias. We're preferentially in a universe where global nuclear war didn't happen.

Regarding chalking things down to "personal choice," the pool of scientists is large enough and the pool of permanent staff positions small enough that there will always be scientists willing to make the "personal choice" to do weapons research. That's the thing about "personal choices," someone somewhere, under the right conditions will make the "personal choice" to do something. If you want people to stop making bad "personal choices," change the underlying situation or the politics.

Societies and the world do not run on personal choices. Sometimes personal choices matter and can have reverberations throughout society, but assuming society can function on an accumulation of personal choices is a child's way of viewing the world. If you want to change something you find disagreeable, influence policy makers or somehow figure out how to change the academic landscape in America (which probably means you again need to influence policy makers). Otherwise, you'll keep finding more scientists making the "personal choice" to develop weapons for uncle sam.

I fail to see how the fact that other people might be willing to do this research changes the ethical propositions at play. You are responsible for your research! Especially when you know what it is going to be used for. Refusing to do the research yourself might not stop it, but you refusing loudly, and convincing a couple people around you to not do so, adds up over time!

You are of course correct that the best thing is to institutionally change things. But the institutional changes are usually pushed for due to actual public pressure.

>but you refusing loudly, and convincing a couple people around you to not do so, adds up over time!

No, it doesn't. But the time you convince someone (if you do), there are 5 more people that you now have to convince who weren't on the scene before.

That'd require the likes of China to also want to do that. And there's no hope of that.

China has an explicit No First Use policy. It also has a nuclear arsenal that is more than 10x smaller than the US.

You might say that China has a secret huge arsenal. The US might have comparable things! And even among the assumed amount of 350 warheads, a lot of damage can be done. The US can do a lot of this as well! There is a nuclear triad on both sides.

Perhaps the No First Use policy is a lie. It is not a law of physics. But it is something. It is a posture that makes it harder for a nation to navigate in some liminal space that makes defensive and offensive posturing look exactly the same.

But the main point of course is that the current state of affairs is already amazingly powerful. What more do people want from their nukes? In what universe is 1500 deployed warheads (out of a stockpile of 5000) not enough to accomplish any sort of strategic objective? There have been many conflicts where 0 has been enough

> China has an explicit No First Use policy.

China's "policies" last as long as their dictator cares to keep them. At the first sign of change that policy will disappear as well.

Also it's a clear lie as in a fictional world where China mainland was invaded and China was losing the war, it's rather absurd to think they would just sit on their nuclear weapons and not use them.

The US also wouldn't use them in a first strike, but drawdowns have to happen on both sides.

> You might say that China has a secret huge arsenal. The US might have comparable things!

The US cannot. That's the entire point of international inspections. China would have to allow US inspections of their arsenal like is done to the US.

Policy is political and will change if the situation changes. If China was the only one in the world with nukes, of course they'd use them.

China's warheads are also generally higher yield - they're intended for counter-city ops, not counter-force.

The idea with the currently deployed warheads is that even after losses to a hostile first strike, they can still respond and cause unacceptable losses in the enemy. How many nukes does it take to cause 10% fatalities in the USSR? How many in China? And what percentage of those nukes are destroyed before launch, and how many make it though enemy defences to detonate where they should?

And of course, nukes decay.

The point is that we need to maintain that stockpile and knowledge. If the US abandons its nuclear capability, there is nothing to hold others to their no first strike policy.

Definitely. I was likely being too glib in my response here, and wasn't trying to say to throw all the nukes into the trash.

I think that it is possible to maintain an existing set of weapons without trying to continue pushing the envelope towards even more destructive weapons, and that 5000 is a lot/enough.

We stopped trying to get more destructive a long time ago. Modern weapons are smaller than they were at their peak.

If anything I'd image modern research is more interested in things like precision, collateral damage reduction, etc.

And make the world a much less safer place? Why would anyone in their right mind do such a thing?

With existing nuclear weapons the US could easily wreck large population centers very easily. They also completely destroy a nations infrastructure. Blowing up population centers of course would be a case of indiscriminate bombing and kill tens of thousands of civilians.

In WW2, even before dropping bombs on Hiroshima and Nagasaki, the US committed to a large bombing campaign that completely destroyed the country. The damage was so vast and total that the Japanese government did not even register Hiroshima to be unusual! Around that time other cities were also being razed to the ground, it just took more bombs to do so. During that war, in Europe and in Japan there were large bombing campaigns that indiscriminately killed many civilians. I find it hard to justify this (no, "they were doing the same" does not actually feel like a legitimate reason, especially when the narrative nowadays is that it was a battle between good and evil)

Low-yield tactical nukes seem to be playing with the idea of being able to have this option to massacre people more easily and play around with the idea of using nukes with no consequence. Higher-yield nukes seem to serve the purpose of more effectively wiping chunks of a city from a map. What magical weapon are we looking for that will make the world safer, and that accomplishes strategic objectives that are not already covered by both the existing nuclear arsenal and the conventional weapons of the US? Remember, normal bombs actually do things too!

Meanwhile, actual, legitimate No First Use policies would be much more effective at making the world a safer place. Stronger treaties to reduce nuclear arsenals, giving militaries less space to just push other countries around.

> I find it hard to justify this

What would you have done instead to reduce their civilian casualties, under the constraints that the Allies still win the war and that Allied casualties don't increase?

[0] is a critique of Gladwell's book about those bombing campaigns. I do not believe it will have all the answers for you, but there is one component of truth: The US was looking for an unconditional surrender for Japan. That was a choice.

Unconditional surrender in Japan lead to the US occupational government for 9 years (along with over 20 years occupation of Okinawa, which had to be pried out of the hands of the US military, who wanted to indefinitely keep it as a colony).

Now, I'm not going to cry over Nazis or the Japanese military not being able to get a conditional surrender. But Japan is especially stark: except for a handful of high profile executions, not only did most decision makers not face punishment, many of them ended up back in the government and as heads of state. All of this more or less at the behest of the US.

What is my point? The US chose to reach for an unconditional surrender, for some pretty realpolitik objectives, prolonging the suffering of the civilians on the ground. There is decent documentation that part of this is so that the US could determine surrender conditions rather than the USSR. There is an alternate universe of a negotiated surrender that shortened the Pacific front.

What does that universe look like? I do not know. It probably would not have lead to the US having military bases all over the country. But conditional surrenders are the norm in wars. Air campaigns allow for militaries to just decide "no, we will just continue to attack forever, as we suffer few casualties anyway".

[0]: https://thebaffler.com/latest/narrative-napalm-kulwin

The US basically did what no losing power ever think would happen. They do unconditionally surrender but then the US enforced pretty damn easy terms on them.

Given that they let the emperor continue, they might as well have just allowed surrender with the promise the emperor would remain. But of course that choice was not made at that point.

An what I think people need to consider is that all of WW2 basically happened because the allies did not force unconditional surrender on Germany in WW1. A mistake that many wanted to avoid. Of course famously Wilson debated this question with Henry Cabot Lodge and that disagreement is a huge part in the US rejecting Versailles. Wilson justified himself by basically formulating the two Germans theory, ie there is the evil militaristic Prussian Germany and the good German people. Henry Cabot Lodge didn't buy that.

Sean McMeekin in his new book makes the point that FDR announced the unconditional surrender doctrine basically to placate Stalin and did it basically without fully coordinating that with the British. Attempting to offer the German military some sort of deal that would have made them remove the Nazis prevent Eastern Europe falling to Communism would have been a better policy. That is of course very controversial.

And if I have learned one thing, its that unconditional surrender is always a incredibly hot issue with people passionately arguing both sides.

> the Japanese government did not even register Hiroshima to be unusual!

You need to read more non-revisionist history books.

From [0]:

> In the three weeks prior to Hiroshima, 26 cities were attacked by the U.S. Army Air Force. Of these, eight — or almost a third — were as completely or more completely destroyed than Hiroshima (in terms of the percentage of the city destroyed). The fact that Japan had 68 cities destroyed in the summer of 1945 poses a serious challenge for people who want to make the bombing of Hiroshima the cause of Japan’s surrender. The question is: If they surrendered because a city was destroyed, why didn’t they surrender when those other 66 cities were destroyed?

> On August 2, you would have arrived at the office to reports that four more cities have been attacked. And the reports would have included the information that Toyama (roughly the size of Chattanooga, Tennessee in 1945), had been 99.5 percent destroyed. Virtually the entire city had been leveled. Four days later and four more cities have been attacked. On August 6, only one city, Hiroshima, was attacked but reports say that the damage was great and a new type bomb was used. How much would this one new attack have stood out against the background of city destruction that had been going on for weeks?

Of course the perspective of people who actually were on the ground is totally different. But if we are talking about strategic objectives, then how the leadership sees thing is in fact relevant!

[0] https://web.archive.org/web/20160221160530/https://foreignpo...

It is single use masquerading as dual use.

The purpose is probably personnel related: employees on this do not need a security clearance, so cost less.

This is false, almost all of those employees have a Q clearance. You can search the job listings for keywords like ‘Wci’ ‘high energy density’, etc to confirm.

Part of the purpose is definitely personnel related though. Part of the US nuclear deterrence is the projection of having a large, highly skilled nuclear weapon related workforce.

Those experiments and supercomputer modelling is what allowed US to get sub-10kt nukes without actual testing. Credible promise of responding with those small nukes directly against Russian regime is what stopped Putin's threat of using nuclear weapons in Ukraine.

Wrt. inertial confinement fusion productization I think the delay is intentional (just look at Sandia z-machine results from 20+ years ago and all the ways of tempering and redirecting progress since then there) as such schemes allow for fusion weapons without fission primary which will completely break the non-proliferation regime.

The US had nukes substantially under 10kt long before these experiments or the existence of supercomputers…

https://en.wikipedia.org/wiki/Davy_Crockett_(nuclear_device)

W-54 isn't here anymore. So instead US have tuned down W-76 into 5-8kt. I.e. getting new capabilities without testing (there are recent tuned down, though not that low, versions of B61 too). And that is open information. One can expect that classified would be at least a step ahead, i.e. something like 1kt. Coupled with high precision delivery and earth-penetration designs (that US has been using across the range - from conventional to B61) that makes for extremely effective deterrence as it allows to take out a dictator like Putin deep in his underground bunker if he crosses the line (Putin becoming the first and primary target himself is the cornerstone of the current deterrence architecture), and other strategic keypoints without initiating full scale war.

> Putin becoming the first and primary target himself is the cornerstone of the current deterrence architecture

Do you have any reference for this?

I'm familiar with US Nuclear bunker buster bombs. But I've never seen any writing that makes the claim the US deterrence policy is targeting Putin.

Indeed [1] makes the claim that the US in incapable of successfully attacking Russian command and control bunkers and seems from a reasonably credible writer.

[1] https://www.realcleardefense.com/articles/2020/12/01/putins_...

> Credible promise of responding with those small nukes directly against Russian regime is what stopped Putin's threat of using nuclear weapons in Ukraine.

Link to credible reports where the US said they would respond with nukes? AFAIK, this never happened and I paid close attention

I have no inside knowledge whatsoever but we can all rest assured that each side is in a near constant back and forth of implicit unstated "communication" about capabilities and doctrine.

Merely publishing a paper on a certain subtopic in the fusion space can easily be interpreted as an implied threat or threat response.

Of course the US does have a stated doctrine of using nukes only in response to nukes used against it or its allies. It is enormously doubtful that the US would trigger an end-of-days scenario in response to Russia using tactical/low-yield nukes against a non-US-ally like Ukraine, but the uncertainty is for sure purposefully cultivated.

Correct, this has not happened and will *not* happen

Do sub-10kt nukes really involve any fusion at all?

I have read of spiking conventional fission warheads with a core of fusible material to boost yield, but maximizing yield does not seem to be the goal of these devices.

One kind of the low yield weapons is to have very precisely placed smaller explosion - those ones is mostly fission core, and the other is to have high neutron flux with minimal blast effects - with that instead of a point you'd want to cover a wide area.

Former LLNL employee here:

You're partially correct on accident. It _is_ single use, but not in the way you think. The NIF facility was built for the express purpose of nuclear weapon design, and any fusion science that comes out of it should be considered a happy accident. I can assure you that very nearly 100% of the people working at the NIF have Q level or higher clearance. The costs are absolutely astronomical.

Maybe this is a dumb question: I thought all nuclear bombs were fission, not fusion. How could fusion be at the center of a fission bomb?

I didn’t even know fusion had a weapons research program

All nuclear bombs involve fission. Some (termed "thermonuclear") involve fusion. A fission first stage is detonated to ignite a (much higher yielding) fusion second stage. Most bombs deployed today are thermonuclear simply because it's the most sensible way to scale up the yield of a weapon.

Nitpick, AFAIK (which is t far) the real crust-busters (e.g. Tsar Bomba) use yet a third stage of fissile material that gets rapidly burnt up at truly absurd neutron densities to get that last 40-90 Mt and as a result are dirty as all hell. The Soviets turned that weapon down from 100Mt designed yield to 50 because there was no way to keep it from just radiating everything.

Isn’t the fissile material the only reason there is residual radiation/fallout from fusion devices?

A lot of materials are fissile in the presence of a fusion reaction.

This makes sense, thank you

Usually if it's called an "atom bomb" or "A-bomb" it's pure fission, and a nuclear bomb, thermonuclear bomb, or hydrogen bomb is fission-fusion-fission. A fission bomb compresses the hydrogen to cause it to fuse, and the extra energy and neutrons from the hydrogen fusion cause a whole lot of additional fission in the uranium tamper. For details look up the "Teller-Ulam" design, wikipedia has some good descriptions.

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

The "hydrogen" in "hydrogen bomb" relates to fusion. In a nutshell, these types of devices use a fission bomb to create the environment (pressure/temperature) that causes lighter atoms to undergo fusion, which significantly boosts the explosive yield compared to a pure fission bomb.

I'm curious what you think the hydrogen in a hydrogen bomb would split into.

Sheesh, tough crowd.

We actually got to fusion bombs pretty quickly, in the 1950s. Presumably almost all practical nuclear weapons since then are fission bombs

They're two stage: fission primary and fusion secondary.

Three stage: the fusion secondary typically uses a uranium pusher, which undergoes fission from the fusion neutrons.

In theory you could do a fusion bomb as a stage, that would make be insane. At least if I understand that correctly.

I thought the fusion produces a lot of neutrons that then cause a ton of more fission.

Nope, the most powerful bombs use both - in effect they use fission to perform fusion. Obviously the problem with this for energy production is that it is not a very controlled reaction.

And then they use the neutrons from fusion to cause more fission.

they have a fission primary stage to ignite the fusion secondary stage

> which is a worthy goal in itself.

I'd like us to focus on practical power generation. It will be the most defining aspect of future of US and largely the world. Everything is tied to energy and if we can make energy cheap enough so that its not worth metering; we'd secure the future from literally any calamity (including CC). Even the shittiest efficiency of carbon capture can be put to use when energy is cheap. 4% efficiency? Cool. Entropy increase from residual heat loss wouldn't make meaningful dent on the world's temperature. It is the carbon that is the problem (greenhouse effect).

We have an almost unlimited source of energy from nuclear + solar. There are always going to be people and ideologies that oppose technological progress and prevent humanity from propelling forward. I belong to the camp where I'd want us to become a Kardeshev Type 1 civilization. Fusion would be a direct contributing factor for it.

> we'd secure the future from literally any calamity (including CC)

What is CC here? Cosmic collapse? Credit cards?

Edit: Ah, of course. Thanks.

Not really. A direct hit by a CME from the Sun would probably seriously damage a fusion plant. Magnets would need to be replaced along with a lot of electrical equipment.

Climate Catastrophe?

Climate Change I assume

Climate change

You know as a matter of fact credit cards are not a wrong answer, really. We only got to the point of 9 billion humans on a tiny planet meant for a few million at best, because the only way out of compound interest debt--synonymous with credit cards--is economic growth--usually more humans.

China saw it that way, in the time of Han Chin, the first Chinese emperor. Wealth comes fundamentally from agriculture because then you can make more servants slash slaves for the emperor, that's literally what he called them, then instead of emperor you have an Emperor, Emperor of China. That's what the original historical sources say! More food more people more servitude more wealth for the man at the top of society.

So credit cards are that. Uniformly crazy interest rates, and shitty scams to jack up the rates just barely before getting taken to court. Or a French Revolution, which they know about and fear. Know the harm they do, the houses they take, the homeless they make, the people they imprison indirectly, the children they starve, they know. What's it's name, FICO score, patio11 talked about them, they are 100% certain you--anybody who reads this--is strictly inferior to them. He says if you talk back to their claim you are an inferior debtor who deserves a low credit score they react like it's a shoe factory dealing with a talking shoe. An object. A servant slash slave.

Owes them money just because. Or because that debt was inherited. They actually have all the machine learning models and all the statistics you could possibly ask for (generally they claim this is fraud detection, but it's price discrimination too) to determine exactly how much--to the thousandth of a percentage--they can fuck with people with their usury--their theft--before people go bananas. Usury means you gotta pay back the debt or be homeless. Tolerate crimes in your gainst with no recourse. Any crime. No recourse. In my case murder. No recourse. Cops won't show up for you.

Debt grows surprisingly fast. Just as surprisingly fast as the equity in the home grows surprisingly slow. People always feel cheated by their mortgage because they did get cheated by unforgivably incorrect math. What does that mean? Ignoring all the intermediate steps, more kids to inherit the debt.

And technically--and I can justify this mathematically and in a court of law--even simultaneously--compound interest is contradictory to the laws of physics. It would not work out mathematically even if they did do the math correctly, which they do not. It could work in an infinitely-dimensional universe. If they did it correctly. But not in a 3-dimensional universe. You can have, at absolute most, cubic growth. Otherwise you end up with shitty debt. Unforgivably incorrect math declaring you are a servant slash slave. A letter demanding you make a choice: servitude or tolerating crime against you.

Shitty debt.

Credit card debt.

9 billion humans.

Climate change.

> I'd like us to focus on practical power generation.

Then what they should invest in is fission. Fuel for fission is much cheaper and CapEx is much less. And we have all the basic tech needed to make it happen. Its the missing will more then anything else.

Omar Hurricane is a top-notch name.

Doctor Hurricane. Never forget that.

I feel like the pendulum has swung too far the other direction these days. It used to be we'll have cold fusion in 20 years which was hopelessly over optimistic. Nowdays its. We are spending 10 years on a myopic proof of concept that has no practical uses and never will.

We're just biding our time until 2050, at which point we will unlock the ability to build industrial scale fusion energy plants.

I would love that. But honestly, I would be surprised if fusion energy will be economically feasible before 2100. If ever.

I'd bet on that. I'd give Commonwealth Fusion Systems alone 40% in the next 20 years.

More on recent fusion developments: https://astralcodexten.substack.com/p/your-book-review-the-f...

I hope you win! I usually get my reality checks from Sabine Hossenfelder, who a while back explained that all these fusion claims are wildly optimistic. You can find her video here: https://youtu.be/LJ4W1g-6JiY

I am no scientist, so it is hard for me to know if team optimistic or team pessimistic is right. But even if it is the latter, I think we should put more money and research on it!

The video isn't nuts, but it's mostly focused on clearing up a common misconception around what "break even" means. It discusses how ITER is not going to be powering the grid in 2025 (strong agree) but I don't see any discussion of the commercialization projects (of which Commonwealth is the strongest) or discussion of timescales anywhere near 2100.

> Sabine Hossenfelder

FYI she's largely been debunked as a fear monger. She swings the pendulum too far in the other direction when reality is somewhere in between.

Sources pls?

According to SimCity, the year is 2050 on the dot.

Just watch out for any stray solar microwave beams in the mean time.

Fusion will never be cheaper then fission.

Unless maybe if somebody figures out aneutronic fusion but we are nowhere close yet.

Would be interesting if solar beats it out. Both honestly are fusion energy.

One of my elderly relatives commented last week that he has been hearing that the new experimental results will promise reliable fusion plants in less than a decade ... Since 1960

(He used to be one of those guys inside an ICBM silo during the Cold war)

The private fusion companies still have ambitious (and most likely unrealistic) goals

So did the Moller Air Car.

Moller at least had a brief prototype that hovered.

There are probably real scientists, engineers, and approaches, but it's probably about fleecing dumb investors at a fundamental level.

I maintain that we would be further along with fusion had we not kneecapped fission into a regulatory abyss.

Reply on news.ycombinator.com