News for hackers. Stuff that matters.
Re:A Stark Nuclear Warning by Gustomaximus on Wednesday June 22, 2016 @09:03PM (#11958225)
From Cuban crisis: "First, Perry writes, the Soviet ships approaching the blockade imposed by the US had submarine escorts that were armed with nuclear torpedoes. Because of the difficulty with communications, Moscow had authorized the submarine commanders to fire without further authorization. When an American destroyer tried to force a submarine to surface, both its captain and the political officer decided to fire a nuclear torpedo at the destroyer. A nuclear confrontation was avoided only because Vasili Arkhipov, the overall commander of the fleet, was also present on the submarine. He countermanded the order to launch, thereby preventing what might have started a nuclear war."

Another one of these 'almost nuclear' stories. There's only so many times humanity can roll the dice and come out clean. Scary stuff.

Re:A Stark Nuclear Warning by krylon on Thursday June 23, 2016 @07:09AM (#11960123)
That is scary.

Even if the submarine had fired a conventional torpedo, that probably would have escalated very quickly.

In "The Fog Of War", McNamara, who was Secretary of Defense at the time tells a story of a meeting between him, Castro, and the man who was commander of the Warsaw treaty troops at the time (I forgot his name, though, sorry). The US apparently knew there were missiles on Cuba, but thought there were no warheads, yet. At this meeting (~1992-ish, I think), the Russian tells McNamara there were indeed enough warheads on Cuba at the time to destroy all major cities on the east coast of the US. So McNamara asks Castro if he knew of that and if he would have advised the Soviets to use the nuclear weapons on Cuba to retaliate against a US attack. Castro, apparently knew of the warheads and had urged the Soviets to use them in case of an attack, knowing full well that Cuba would have been destroyed completely. (At least that is how I remember McNamara telling the story, and assuming, of course, McNamara remembers the incident correctly and is telling the story accurately.)

Re:A Stark Nuclear Warning by ams6110 on Wednesday June 22, 2016 @09:41PM (#11958353)
A nuclear torpedo to take out a destroyer? Seems like the proverbial swatting a fly with a sledgehammer? And what of the submarine? Torpedos are fairly close-range weapons, at least in those days. Would it have survived the blast?
Re:A Stark Nuclear Warning by mikeash on Wednesday June 22, 2016 @10:26PM (#11958485)
A lot of weapons were nuclear (air to air missiles and anti-ICBM weapons are other examples) not because they necessarily needed the explosive power, but to make up for inaccuracy. Guidance in the 50s and 60s isn't what it's like today. When your weapon is likely to miss by several hundred feet, you can resign yourself to missing a lot of shots, or you can stick a nuclear warhead on it and take out the target anyway.
Re:A Stark Nuclear Warning by hackuser on Thursday June 23, 2016 @01:19AM (#11958996)
> A lot of weapons were nuclear ... to make up for inaccuracy

In fact, I've read that that's the real reason for even the large strategic nuclear weapons: The accuracy of missiles launched from thousands of miles away is (or was) dreadful, but if you used strategic nukes, you still could take out the targets: The Kremlin, the naval base, etc.

I've even read that modern 'smart' weapons (the article discussed Perry's work on them and their accuracy) could make nuclear weapons obsolete.

Re:A Stark Nuclear Warning by mikeash on Thursday June 23, 2016 @09:41AM (#11960743)
That was definitely a large part of it. Another part was the desire to maximize what you got out of a particular shot: if you have one bomb per missile, then you want each missile to pack maximum punch. Improving accuracy and the development of MIRVs both contributed to less powerful bombs becoming more desirable.

I'm not sure about precision conventional weapons making nuclear weapons obsolete. I can see it for some cases. Destroying a missile silo can probably be done with a large normal bomb delivered with sufficient accuracy. Destroying an airfield would require a lot of bombs, impractical if you're delivering them on ICBMs. And destroying a well built underground command post probably requires nukes.

Worse, sometimes cities really might be targeted deliberately, at least as a deterrent. A war limited to military targets might be seen as acceptable, while having cities destroyed could make it unacceptable, so there's incentive to threaten retaliation against cities in the event of an attack. No amount of precision lets conventional weapons substitute for nukes there. (Note that MIRV and better accuracy does still push for smaller weapons, as many smaller weapons detonated over an area is more efficient at destroying things than a single large weapon.)

Re:A Stark Nuclear Warning by kijiki on Thursday June 23, 2016 @12:57PM (#11962500)
A good book discussing the drivers (and some of the technology involved) of increasing ICBM/SLBM accuracy is "Inventing Accuracy: A Historical Sociology of Nuclear Missile Guidance".

https://mitpress.mit.edu/books/inventing-accuracy

Re:A Stark Nuclear Warning by saiya-jin on Thursday June 23, 2016 @04:46AM (#11959628)
if the torpedo misses, how would the warhead get triggered if not from impact? would there be a timer instead and be primed X seconds after launch?
Re:A Stark Nuclear Warning by ptaipale on Thursday June 23, 2016 @06:57AM (#11960076)
There's a lot of human ingenuity applied in deciding when and how to blow things up. Munition can be timed, or use some sort of proximity sensing, etc.

Even artillery grenades, hurled out in great numbers, have such technology (and have had since WWII); for a nuclear torpedo, even a radar-operated fuze or at least a magnetic sensor would make sense.

Re:A Stark Nuclear Warning by vonmoltke on Thursday June 23, 2016 @10:13AM (#11960961)
While there isn't much public information about the Type 53-58 (the Soviet torpedo in question), it was probably a straight-runner with a range of 11nm - 13nm. It was also intended for area effect, not direct impact. Considering these parameters, it was probably fused to trigger the warhead at the end of its run, i.e. when the engine ran out of fuel. It may have had a timer or "run to detonation" feature instead, but as I said public information is scant.
Re:A Stark Nuclear Warning by undersuit on Thursday June 23, 2016 @11:43AM (#11961862)
>11nm - 13nm

Nautical Miles? I guess it's ok that an Imperial symbol clobbers a metric symbol when they are trillion magnitudes different.

Re:A Stark Nuclear Warning by mikeash on Thursday June 23, 2016 @11:49AM (#11961922)
Yes, nm is a common abbreviation for nautical miles in cases where they're used. And thankfully it is unusual to encounter a situation where you can't tell which one was intended.
Re:A Stark Nuclear Warning by pm90 on Wednesday June 22, 2016 @09:45PM (#11958360)
I may be wrong here, but a surface detonation is unlikely to damage a submarine that is deep underwater. That is the reason why a fleet of nuclear submarines are the best nuclear deterrent: hard to take out since they can be anywhere undersea, mobile and ready to counterattack at a moments notice.
Re:A Stark Nuclear Warning by graycat on Thursday June 23, 2016 @06:42AM (#11960002)
> hard to take out since they can be anywhere undersea, mobile and ready to counterattack at a moments notice.

When I was working in that stuff, the usual point about the SSBNs (the US long range missile firing submarines) was that they were darned difficult to find. Also, it was standard to assume that they moved in some variation on Brownian motion so that, even if did get some data on their location, it was not very useful for extrapolation.

So, a major point about the design and operation of the SSBNs was that they be darned difficult to find -- via passive sonar, magnetic anomaly detection, active sonar, or anything else. Also, a big deployment of search resources that might be more able to find an SSBN would be regarded as provocative.

Then there was the point: If the Soviets did find a US SSBN, now what? If they track it for very long, then that is provocative. If they don't track it, then they will soon lose it again. If they sink it, then there are all the other US SSBNs ready to fire back.

So, the Soviets would have to find AND sink ALL the US SSBNs at the same time. Finding all of the SSBNs at the same time without being provocative was regarded as essentially impossible.

So, maybe somehow there could be a nuke war but limited to sea. Then could pick off each SSBN as it was found one at a time. Then how long would the SSBNs last? Once I was asked to answer that! So, I derived some math, wrote some software, submitted my results, and the next day my wife and I went for a vacation in Shenandoah! Later my work was sold to a leading US intel group. I could tell you which one, but then I'd have to ...!

Ah, the math wasn't either artificial intelligence or machine learning!

Re:A Stark Nuclear Warning by kastnerkyle on Thursday June 30, 2016 @03:30AM (#12007087)
OT: Can you send me a link to your 1999 Elsevier article? The old thread is locked due to age but the topic is very relevant to me...

As to on topic discussion - wouldn't you only need to take out the comms? Then the fellow subs wouldn't know anything was amiss.

Taking down several redundant, military grade links doesn't sound easy, but finding every sub in the ocean seems much harder.

Maybe there is a watchdog / normal coded communication/expected updates, so it would have to be more of a MITM thing. Even in the worst case, it would leave a limited window of time to act. Not to mention the submarine would likely be forced to try and hail "ground control" which would make DF + geo much easier...

Re:A Stark Nuclear Warning by srean on Friday June 24, 2016 @04:41AM (#11967903)
You do seem to have a distaste for the term machine learning.

It is applied math, anybody who claims otherwise could not be more wrong. It is applied probability theory and computer science with lot more focus on algorithms than is expected of a statistician, and yes without a fetish for asymptotic normality. Usually, but not always, with a stronger focus on prediction accuracy rather than on accuracy of parameter estimates. In that sense its closer to (non-parametric) decision theory than to run of the mill stats.

Re:A Stark Nuclear Warning by graycat on Friday June 24, 2016 @10:10AM (#11969855)
Yup. I got all that from Leo Breiman, not his earlier work as an "academic probabilist" but his later work in classification and regression trees (CART), random forests, etc.

Yup, if look carefully, in many common cases can do well on prediction accuracy even when parameter accuracy is not so good. E.g., can even do well on prediction accuracy when parameter accuracy is zilch, e.g., some, but not all, cases of over fitting.

You are right about the applied math: Machine learning has a lot that is good and new, however the good is mostly not new and the new, not very good. E.g., I watched a Prof Ng lecture on the gradient descent unconstrained minimization involved in their fitting, and it seemed that either their minimization problem is awfully easy or Prof Ng didn't know much about the applied math of such minimization going way back.

So, machine learning stole a lot from some now quite classic applied math, assigned some new names, e.g., learning and another anthropomorphic name they dreamed up for iteration step size, and claimed something new. That's called academic theft, plagiarism, or some such, right?

Then there's the hype. So, just from machine learning, the pop culture press is extrapolating to the Terminator movies. Absurd.

Then, bigger, there's the implicit assumption, if only from just playing with the words, that the main path to smarter machines is machine learning. Well, if the effort is limited to Breiman's piecewise linear curve fitting, that path is absurdly limited. Much better for making machines smarter, i.e., do more, work better, is just to start with lots of now classic pure/applied math, get out a clean sheet of paper, and derive what need for the problem at hand. As much as I like Breiman, both his early work and his last work, there're oceans more good work in applied math that is nothing like Breiman's last work, e.g., with high irony, his earlier work!

Uh, here's another case where it looks like a machine has learned -- stochastic optimal control. In some cases, it can look darned smart. Another place -- game theory, to be more specific, the von Neumann-Morgenstern work. Then the Nash generalization. But such things are far from the current machine learning.

Re:A Stark Nuclear Warning by srean on Friday June 24, 2016 @05:38PM (#11973698)
As I have mentioned several times to you start with vapnik. That's where modern ml begins. Brieman is great but he got a few things wrong, for example he thought boosting and bagging are fundamentally the same, that is they are just variance minimization methods. In fact boosting us what gives random forest its power. BTW less well known trick, use random rotation not just random selection of features or datapoints. Their's lot more to ml than decision trees. One of its fundamental framework is convergence of empirical processes, which is essentially glivenko cantelli on steroids. Vapnik and Chervonenkis provided the first modern breakthrough result but a lot has happened since then.

I would say screw the hype just read what it is. I think your disdain for the name 'machine learning' is getting in your way.

I strongly disagree with your allegation of academic dishonesty though. Optimization has been a fundamental part of ml from the very start and with a strong overlap in the community. I am sure nemirovskii will ring a bell. In any case how can ml steal from applied math when it is applied math. The lingo differs here and there: estimating parameters become learning parameters etc.

Andrew NGO's course is not for you. Go directly to the books and then conference and journal papers.

Focus varies but info theory, stats, signal processing, approximation theory are pretty much the same thing, so are their open problems.

Btw you could not be more wrong about stochastic optimal control or repeated games being far from current ml. Oh and another major major tool in ml is geometry of function spaces. You are reading the wrong sources, probably the populist ones. Get to the real stuff given what i know of your taste in math i think you will enjoy it.

Re:A Stark Nuclear Warning by graycat on Tuesday June 28, 2016 @10:06AM (#11993868)
Thanks, I'll keep that.

"glivenko cantelli on steroids", good. Sounds like they actually did something.

Yes, I'm torqued by the new learning labels on old bottles of pure/applied math, but that is not in my way.

> The lingo differs here and there: estimating parameters become learning parameters etc.

Rubs my fur the wrong way.

If they have some stuff beyond borrowing from Breiman, okay.

What's "in my way" now is my startup: I've got the math derived and typed into TeX and the 80,000 lines of typing for the code, with the code running, intended for production, and in alpha test, so just for now I no longer have any pressing math problems to solve.

But, in time, I may return to my math and tweak it a little to try to get some variance reduction. Maybe some of the better machine learning literature would help, or maybe I'll just derive it myself again.

Function space geometry is about where much of my core math is.

Thanks.

Re:A Stark Nuclear Warning by srean on Tuesday June 28, 2016 @11:37AM (#11994585)
Heh! indeed, its all geometry :)

Happy to hear back from you. I am actually gladdened that your anomaly detection work is getting some interest on HN lately. Hope something comes out of it. I am now slowly coming to the conclusion that pushing better methods on an existing stack would be really hard. Too much friction, too much politics. Perhaps the way is to create your own better cloud of servers, but that's really big league stuff. Not sure I have the stomach for that.

Curious if you have given any thought to the choice of the metric space where you define your statistics. That might play an important role from what I have seen. There might be an interesting manifold story there.

Big spoilsports are non-stationarities and even bigger are those fat tails. If only everything had a moment generating function.

I see that you have been pointed to Abu-Mostafa he is definitely a good source. Not that Andrew Ng is unaware of the stuff, far from it, he is fighting a different battle: to make parts of ML a commodity.

If you have time then you can browse the proceedings of COLT (conference on learning theory) and ICML.

> or maybe I'll just derive it myself again.

You almost always have to derive it yourself anyway even after you have seen the derivation by somebody else.

Re:A Stark Nuclear Warning by selimthegrim on Monday June 27, 2016 @10:03AM (#11986189)
http://amlbook.com

https://www.youtube.com/watch?v=mbyG85GZ0PI (incidentally for graycat, Yaser Abumostafa is a Muslim Egyptian immigrant from Cairo) He covers the VC dimension in https://www.youtube.com/watch?v=Dc0sr0kdBVI, and leaves the proof to an appendix of the book.

Re:A Stark Nuclear Warning by graycat on Tuesday June 28, 2016 @10:37AM (#11994107)
Thanks. I looked a few minutes at two of his lectures. I'll keep the URL and watch his lectures during dinners.

Lucky he got out of Egypt before they strung him up! Such good people is what US immigration has for some decades now tried to be about. Maybe we will get back to it.

"Muslim"? I don't care if he is Zoroastrian either. Or worships some sun god. I don't care about his religion. I do care if he wants to blow up buildings. Somehow I doubt if he does.

Looking at his videos, my first cut, crude guess is that he is looking at modern generalizations of old discriminate analysis. Yup, that can be important. Maybe it could be important for, say, one of my old interests, anomaly detection, say, as a doable alternative to Neyman-Pearson where often in practice we don't have nearly enough data. Maybe his interest is in medical diagnosis which, IIRC, was some of Breiman's interest.

But, first cut, it looks like, again, the criterion will be, does the model fit the data well? That is, we have little or nothing to recommend the model except that it fits the data well. But, then, in the case of his lectures, it looks like maybe he is making progress to also knowing that the model will predict well. I'm looking forward to how he does that.

In contrast, if that is important, in my work in anomaly detection, discussed here on HN often enough, I found false alarm rate from some derivations in applied probability with no model fitting at all. Okay, I don't care if the cat is black or white as long as it catches mice.

From a glance, it looks like he is addressing what is meant by learning -- terrific! Not just throwing words around! Then he seems to be addressing when such learning is feasible, etc. Sounds good; I've wondered some about something like that.

But my interest now in what he is doing is a bit limited since the core math in my startup seems to be quite different.

Thanks.

Re:A Stark Nuclear Warning by srean on Tuesday June 28, 2016 @01:18PM (#11995556)
> it looks like maybe he is making progress to also knowing that the model will predict well. I'm looking forward to how he does that.

Yes that's exactly it. By way of Vapnik and Chervonenkis' result (essentially an uniform law of large numbers) one upper bounds the expected accuracy (over the unknown distribution) of a classifier in terms of the training error and another quantity that depends on the class of hypothesis that one is using. One can give bounds even when one is using an infinite class, for example all linear functions in the feature space, or some Hilbert space of functions.

This was one of _the_ major early break through result. Its often quoted in the context of ML but it really is a result in probability theory. Since it bounds the most pessimistic situation possible, they are quite bad (although achievable).

It also brought about a paradigm change in the mindset. Since the optimal classifier is just the thresholded conditional density, early approaches had focused mostly on estimating this conditional density. But that's an impossible task. V&C showed even if you do not have enough data to learn the density, you may have more than enough for good prediction accuracy. Don't learn the conditional density, just learn the discriminating function directly by optimizing its expected loss.

People have moved to different tools to bound expected prediction accuracy. You get a lot more reasonable bounds, say with the PAC-Bayesian theorem.

Key thing is that these are distribution independent, non-asymptotic and also dimensionality independent.

Re:A Stark Nuclear Warning by tr1ck5t3r on Thursday June 23, 2016 @08:51AM (#11960481)
This was around https://en.wikipedia.org/wiki/SOSUS and better stuff is around today.

Nothing like a big bang to scare the population into projectfear.

Re:A Stark Nuclear Warning by vonmoltke on Thursday June 23, 2016 @10:16AM (#11960990)
> If they track it for very long, then that is provocative.

Both sides spent the entire Cold War shadowing each other constantly. There was nothing provocative about it.

Re:A Stark Nuclear Warning by graycat on Thursday June 23, 2016 @10:59AM (#11961422)
I question if the Soviets were able to shadow a US SSBN. I shouldn't give more info.
Re:A Stark Nuclear Warning by elorant on Thursday June 23, 2016 @05:09AM (#11959698)
Russians tested a 4.8 Kiloton T-5 nuclear torpedo in 1957. They used three decommissioned submarines as targets. The weapon was detonated at 20 meters below surface and sank two of the three submarines at a distance of 6.5 miles. As previously mentioned nuclear torpedoes of the era relied on blast damage rather than precision. The shock wave could easily break the hull of any vessel, be it underwater or on the surface.
Re:A Stark Nuclear Warning by narag on Wednesday June 22, 2016 @11:37PM (#11958692)
Can a submarine that is deep underwater launch torpedoes? I didn't think so.

OTOH, a nuke doesn't need to yield several megatons. First devices were about 20 kilotons and it's possible to create smaller ones. I'm sure I've seen some video of 1 kiloton bomb. Also, intuitively, it doesn't seem that it would destroy a sub a few miles away looking at Baker test:

https://www.youtube.com/watch?v=5nWFx-zmI0k

Around 1:25 there's a brief fragment showing the explosion underwater with the usual expansion-contraction cycle.

Re:A Stark Nuclear Warning by vonmoltke on Thursday June 23, 2016 @10:14AM (#11960966)
> Can a submarine that is deep underwater launch torpedoes? I didn't think so.

Where did you get this idea from? Submarines can launch torpedoes at whatever depth the torpedo itself can operate at.

Re:A Stark Nuclear Warning by narag on Thursday June 23, 2016 @12:22PM (#11962216)
Where did you get this idea from?

Movies. On second thought, there was some submarine against submarine battle in Red October... most other submarine wartime movies were from WWII, always shooting at boats and at periscope level to aim.

Oh, and speaking about wars, relevant nick :-)

Re:A Stark Nuclear Warning by ceejayoz on Thursday June 23, 2016 @09:46AM (#11960766)
> Can a submarine that is deep underwater launch torpedoes? I didn't think so.

The maximum depth for a https://en.wikipedia.org/wiki/Mark_48_torpedo is estimated at 800 meters.

Re:A Stark Nuclear Warning by reactor4 on Thursday June 23, 2016 @10:34AM (#11961148)
Ranger Able was 1kt.
Re:A Stark Nuclear Warning by graycat on Thursday June 23, 2016 @06:23AM (#11959915)
> And what of the submarine?

Very good point. When I used to work in that stuff, it was assumed that once know where a sub was, it was easy to kill it -- just pop off a nuke and let the resulting pressure wave in the water crush the hull of the sub.

Sounds like the Soviet nuke torpedo was a very low yield nuke -- if the sub would survive, then the explosion would not have been big enough to have been a trigger of full nuke war like, say, a 10 megaton nuke over a major city.

Re:A Stark Nuclear Warning by nxzero on Wednesday June 22, 2016 @10:01PM (#11958410)
Torpedo on the submarine was a T-5 and if the submarine was within 6-10 miles of the blast it would have ripped the submarine apart.
Re:A Stark Nuclear Warning by privong on Wednesday June 22, 2016 @10:05PM (#11958422)
> Torpedo on the submarine was a T-5 and if the submarine was within 6-10 miles of the blast it would have ripped the submarine apart.

Do you have any additional info on this? A quick wikipedia search turned up a little bit of info[0], but suggests the targets destroyed were 6.5 miles from the submarine that launched the torpedo and presumably that the launching submarine was not destroyed:

On 10 October 1957, in another test at Novaya Zemlya, S-144, a Whiskey class submarine, launched a T-5. The test weapon, code named Korall, detonated with a force of 4.8 kilotonnes twenty meters under the surface of the bay sending a huge plume of highly radioactive water high into the air. Three decommissioned submarines were used as targets at a distance of 6.5 miles. Both S-20 and S-34 sank completely, and S-19 was critically damaged.

[0] https://en.wikipedia.org/wiki/Nuclear_torpedo#T-5

Re:A Stark Nuclear Warning by InclinedPlane on Thursday June 23, 2016 @03:06AM (#11959342)
I'm not sure how you came to those conclusions. A live fire test of a nuclear torpedo prototype seems ... more than a little unlikely. What you have here is the description of a test just like any of the American tests: a bare warhead is detonated under certain conditions, with perhaps targets nearby to determine the effects. Meaning that the targets were 6.5 miles from the warhead. And no "launching submarine" was involved, since only the warhead was tested not the end to end weapons system (which is an extraordinarily unusual occurrence with nuclear weapons systems and has only been done a handful of times in history).
Re:A Stark Nuclear Warning by privong on Thursday June 23, 2016 @08:28PM (#11965350)
> A live fire test of a nuclear torpedo prototype seems ... more than a little unlikely. What you have here is the description of a test just like any of the American tests: a bare warhead is detonated under certain conditions, with perhaps targets nearby to determine the effects. Meaning that the targets were 6.5 miles from the warhead. And no "launching submarine" was involved,

I was just going off what I quoted from wikipedia, which does specifically say that a submarine launched a T-5. Beyond that I can't speak to the accuracy of the wikipedia article.

Re:A Stark Nuclear Warning by arethuza on Thursday June 23, 2016 @06:38AM (#11959980)
The Soviets were pretty keen on live fire tests - their very first test of a staged thermonuclear device was air dropped. Air drop tests were fairly common, AFAIK the only people who have done a live launch of a nuke on a missile were the Chinese.
Re:A Stark Nuclear Warning by mikeash on Thursday June 23, 2016 @10:23AM (#11961053)
The Soviets tested an R-13 SLBM with a live warhead in 1961. A Polaris SLBM with a live warhead was tested by the US in 1962 as part of Operation Dominic. I don't think either country did a live test of an ICBM though.
Re:A Stark Nuclear Warning by arethuza on Thursday June 23, 2016 @12:21PM (#11962199)
I appear to have mis-remembered - the Chinese test was an MRBM:

https://en.wikipedia.org/wiki/Dongfeng_(missile)#Dongfeng_2

Re:A Stark Nuclear Warning by mikeash on Thursday June 23, 2016 @12:48PM (#11962431)
Even so, testing a live nuclear missile with a 1,250km range must be tremendously exciting.
Re:A Stark Nuclear Warning by arethuza on Thursday June 23, 2016 @06:26PM (#11964752)
That's one word for it!
Re:A Stark Nuclear Warning by kijiki on Thursday June 23, 2016 @01:08PM (#11962620)
https://www.youtube.com/watch?v=01WpGDwSPpg

The US tested a live Polaris in 1962. They were responsible enough to cook it off at 11,000 ft, a real combat launch would likely have been detonated closer to the ground.

Re:A Stark Nuclear Warning by beloch on Wednesday June 22, 2016 @11:34PM (#11958683)
I have no solid evidence to help you, but my intuition says this is believable. Pressure waves can travel a long ways in dense media, and it's the interface between media of different densities, such as water of the ocean and the air inside a submarine, where the most devastating effects are felt. Jelly-fish swimming next to a submarine might not feel a thing even while the submariners inside the submarine are being smashed to bits.
Re:A Stark Nuclear Warning by vonmoltke on Thursday June 23, 2016 @10:22AM (#11961050)
Prior to WWII, that intuition led navies to greatly overestimate the lethal range of their depth charges. Once in combat, they discovered that they were way off, close to an order-of-magnitude in some cases.
Re:A Stark Nuclear Warning by vonmoltke on Thursday June 23, 2016 @10:10AM (#11960934)
Torpedoes haven't generally been "fairly close-range weapons" since WWII (air-launched torpedoes notwithstanding).
Re:A Stark Nuclear Warning by arethuza on Thursday June 23, 2016 @07:02AM (#11960093)
Of course, they might not have expected to survive the blast - a lot of people (e.g. bomber crews) expected their trips to be one way.

Also, from what I recall from One Minute to Midnight the conditions on that sub had been absolutely awful for quite a lot time (in particular, no cooling) so they might not have been thinking particularly rationally.

Re:A Stark Nuclear Warning by paavokoya on Wednesday June 22, 2016 @10:08PM (#11958429)
Having flashbacks of 2001: Space Odyssey.. At first the apes postured with intimidation tactics, then one finally took the leap and started using the weapon. It's quite terrifying to consider just how natural and statistically probable it is for nukes to be used again.
Re:A Stark Nuclear Warning by wnoise on Wednesday June 22, 2016 @11:08PM (#11958611)
Really makes you consider the anthropic principle.
Re:A Stark Nuclear Warning by sehugg on Wednesday June 22, 2016 @11:30PM (#11958671)
Also a (currently unconfirmed) account of a belayed launch order given to Okinawa missileers during that time: http://thebulletin.org/okinawa-missiles-october8826

Whistling past the graveyard indeed.

Re:A Stark Nuclear Warning by IanCal on Thursday June 23, 2016 @04:32AM (#11959589)
> There's only so many times humanity can roll the dice and come out clean

Here's a related video on that thought from Minute Physics:

https://www.youtube.com/watch?v=qRnU0bqsyq0