HNHacker News
TopNewBestAskShowJobs

jjk166

8,522 karma · joined April 26, 2017

submissionscomments
jjk166··on War game exposed U.S. vulnerability to low-tech warfare
Well in the recent Iran war they kept their major ships far from Iran and didn't try to do any major operations in the area where small Iranian boats like those used in the wargame would be a threat. Indeed none of the vulnerabilities exposed by the Millennium Challenge were actual issues in the recent conflict. Whether that means the lesson was learned or the red team tactics were exploiting issues specific to the wargame is not clear.
jjk166··on War game exposed U.S. vulnerability to low-tech warfare
> "saving up" for massive clouds of long range strike drones every couple weeks, instead of sending out a couple hundred every night.

There are real technical limitations to operating large numbers of drones simultaneously. They are competing with eachother for bandwidth, and the more you squeeze in the weaker they all become to jamming. You need to coordinate larger teams with more resources to operate lots of drones all at once, rather than resupplying smaller teams periodically. Once you have enough drones flying toward a target to survive the defenses, any additional drones are wasted; and for the moment most air defenses are either too expensive or too limited in the number of targets they can simultaneously handle to require large swarms. Smaller numbers of more capable drones can outperform larger numbers of cheaper drones - 1 drone with an 80% chance of neutralizing the target might be cheaper than 2 drones which individually have a 50% chance for a total of 75%; and 2 such drones beat 5 cheap ones. Finally, big attacks are more likely to be detected, either prior to the attack by espionage or while in transit by radar and other sensors - minimizing the time that defenses have to respond or for targets to flee can be much more advantageous than having an extra munition.

All around you want the minimum necessary to get the job done.

jjk166··on War game exposed U.S. vulnerability to low-tech warfare
> My main thesis is that the cost of small autonomous drones have dropped enormously over the past 20 years, and that increasingly drone swarms look like a much more capable and cost effective way of waging war

And how about the cost of drone countermeasures? Ukrainian interceptor drones are already a fraction of the price of a Shahed, and drone jamming systems are evolving at a lightning pace. Every weapon development seems like an unstoppable game changer until an effective defense is widely fielded, but that defense does tend to wind up getting fielded, or the offensive technology is completely superseded, by the time the next major conflict breaks out.

In the current Iran conflict not a single US warship has been damaged by a drone or similar system. Iran is among the leaders in drone technologies and started the war with a substantial stockpile. The US navy had not previously gone up against an adversary in the drone age and all of its ships were designed prior to the start of the Ukraine war. If there was ever a time where a large number of drones could overwhelm a ship with inappropriate defenses, it would have been then. In 5 years there isn't going to be a combat vessel in any first rate navy without dedicated counter-drone defenses and every officer will be well versed in counter-drone tactics. The best drones in the hands of the most experienced units may still be able to defeat these countermeasures, but the idea that you're going to overwhelm an aircraft carrier with 10,000 cheap drones is a fantasy.

jjk166··on Qian Xuesen: The missile genius America lost and China gained (2025)
There are biopic films about people who founded or transformed businesses like Steve Jobs, Roy Kroc, Mark Zuckerberg, the founders of Blackberry, etc. Might not be everyone's cup of tea but I wouldn't describe that genre as boring. Probably the bigger issue is getting people to see a biopic about someone who isn't already a household name.
jjk166··on I found ultra-pure quantum crystals in an abandoned mine in the Atacama desert
https://en.wikipedia.org/wiki/Comparative_advantage
jjk166··on Actually, democracy dies in H.R.
> Why is this so hard to understand?

Because a functioning neocortex prevents it.

Over-hiring is inneficient. If not being able to easily lay people off prevents that issue, it's a good thing.

You're making an logical leap that trying a thing and it failing, or otherwise pivoting requires layoffs despite clear evidence to the contrary.

I can tell you for a fact high performers don't want to see their coworkers laid off, and layoffs destroy team morale more than anything else.

The generalizations make perfect sense because what makes something a layoff makes all the conditions I described true.

I can't imagine any intelligent person arguing in good faith that there is no middleground between freely laying people off on a whim and paying people to do nothing because they can not be fired.

jjk166··on Iran starts Bitcoin-backed ship insurance for Hormuz strait
> full annual income of someone from somalia.

Not if they seize a cargo ship it isn't. Criminals can afford the tools to commit crimes by using those tools to commit crimes.

jjk166··on Actually, democracy dies in H.R.
> They go hand in hand. Companies need to be able to innovate. And yeah that means hiring and firing teams as they see fit.

Think about any job you ever worked at.

When layoffs happened, did the work ever get easier or done to a higher standard because of the layoff, or did the sudden loss of knowledge and manpower make your job harder?

When things were bad enough that layoffs were being talked about, who jumped ship, the low performers who don't really do anything or the high performers who can get a job anywhere?

When innovation happened, did it tend to come from the team that would be fired for screwing up, or the team that could confidently experiment and incorporate lessons from past trials.

The idea that weak labor protections is a key requirement for innovation doesn't stand up to the slightest scrutiny.

jjk166··on Actually, democracy dies in H.R.
HR exists exclusively for the numerous situations where the company's interests and employee well-being are in conflict. In any situation where the interests are aligned, there's nothing for HR to do - left to their own devices people will just do the thing.
jjk166··on Actually, democracy dies in H.R.
It takes a special kind of mind to emphasize the importance of listening while simultaneously disregarding what someone actually says.
jjk166··on NYT and vaping: How to lie by saying only true things (2022)
You don't need to prove the opposite of someone's claim to argue their evidence fails to demonstrate their claim is true.
jjk166··on WHO declares Ebola outbreak a global health emergency
But the risk of dying by having your entire body break down in uncontrollable bleeding is substantially less than 100%, and I for one would like to keep it that way.
jjk166··on WHO declares Ebola outbreak a global health emergency
> Is there some small gene change that we're specifically worried about here

Yes. A single gene change allows for airborne Ebola transmission. This gene change has occurred in the Reston strain, which luckily does not cause symptoms in humans.

https://en.wikipedia.org/wiki/Reston_virus

jjk166··on US is starting to see heavy job losses in roles exposed to AI
Pretty sure that would just reinforce my point...
jjk166··on US is starting to see heavy job losses in roles exposed to AI
Are the jobs going away because of AI being applied or is AI being applied because those jobs are going away? You see lots of life jackets being worn as a ship is sinking; doesn't mean people putting on the life jackets made the ship sink.
jjk166··on US is starting to see heavy job losses in roles exposed to AI
The stock market isn't the economy. Note that stock prices are denominated in dollars. If the dollar price of a stock doubles, but in the same time period the value of the dollar halves, then the stock's value hasn't gone up at all. Compare the price of the Dow Jones to to gold and its value peaked in 1999, and has been nosediving for the past 18 months[0]. The ratio right now is the same as it was in late 2008. Obviously gold isn't a perfect standard candle, but given that inflation has obviously been rampant in recent years, we should obviously treat any economic metric that does not control for inflation with healthy skepticism.

[0] https://www.macrotrends.net/1378/dow-to-gold-ratio-100-year-...

jjk166··on U.S. DOJ demands Apple and Google unmask over 100k users of car-tinkering app
That's the narrative. In reality, you don't need a list of every user to find a handful of witnesses. This is very clearly a power play - the feds can't make their case so they're making distributing the app a major PR headache for the platforms distributing the app. Apple and google will quietly change their user agreements to not permit these sorts of apps on their stores. The company's leadership will likely escape prosecution but their business will be crippled if not destroyed despite not being proved guilty of violating any law.
jjk166··on Heritability of human life span is ~50% when heritability is redefined
> Your height is as biological a thing as anything else, but it's tied to your environment in the same sense as hair color.

And you wouldn't draw a distinction between the person who is short because of poor diet and the person who is short because they lost their legs in a car accident? Both are "environmental factors" which affect the distance between the top of your head and the ground, but that's not what we are referring to by height.

> we've always talking about confounded measures.

No, we haven't. It doesn't matter that confounding factors exist in the data, we can and near exclusively do talk about abstract concepts. We live in a world where there are no perfect circles, but we can talk about things having diameters. We live in a world where people die from unnatural causes, but we can still talk about people having natural lifespans. That removing confounding factors is hard doesn't change the fact we routinely make our best effort to do just that because it is necessary for discussing the abstract concept we all refer to.

jjk166··on Heritability of human life span is ~50% when heritability is redefined
> OK, but check this out: Say I redefine “hair color” to mean “hair color except ignoring epigenetic and embryonic stuff and pretending that no one ever goes gray or dyes their hair et cetera”. Now, hair color is 100% heritable. Amazing, right?

It seems incredibly disingenuous to lump together epigentics and hair dye when talking about heritability of hair color. We all know when we talk about inheriting hair color we're talking about natural hair color.

> his paper built a mathematical model that tries to simulate how long people would live in a hypothetical world in which no one dies from any non-aging related cause, meaning no car accidents, no drug overdoses, no suicides, no murders, and no (non-age-related) infectious disease.

Which is exactly what everyone means by lifespan in this context. No one on earth is trying to figure out how much genetics contributes to the odds of being hit by a bus.

> veryone seems to be interpreting this paper as follows:

>> Aha! We thought the heritability of lifespan was 23-35%. But it turns out that it’s around 50%. Now we know!

Which is the correct interpretation. Proper elimination of confounding factors is good science. The previous estimates were low because they weren't properly measuring what we are all referring to when we talk about lifespan.

jjk166··on US budget watchdog estimates Golden Dome will cost $1.2T
While the administration's estimate is absurd, even $1.2 Trillion seems rather low. For comparison the Sentinel ICBM program, replacing the current Minuteman III missiles, which are just one prong of the nuclear triad, with a missile with similar capabilities is estimated at $140 Billion.
jjk166··on Venom and hot peppers offer a key to killing resistant bacteria
Improving phages is dramatically easier - you can just modify an existing one, either through direct engineering or by evolution, rather than having to find a brand new chemical that conveniently does what you want without serious negative drawbacks. 20 years ago the difficulty may have been comparable but the engineering tools for creating synthetic phages have advanced extremely rapidly.

And particularly if you are introducing a vulnerability instead of directly attacking the bacteria with the phage, there is no evolutionary pressure to become resistant until the phage has already done its job. You can even go further and have it insert genes that confer an advantage in addition to the susceptibility, so that even if some of the bacteria are by chance naturally resistant to the phage they get outcompeted prior to the deployment of the killing agent.

jjk166··on A.I. note takers are making lawyers nervous
Is it reasonable to expect any call going through a computer to be off the record, even without AI? Recordings were always discoverable, the only difference is that a paralegal doesn't need to manually go through every recording to determine if it's relevant.
jjk166··on Venom and hot peppers offer a key to killing resistant bacteria
I imagine the long term solution to antibiotic resistance are engineered bacteriophages, either to directly attack the bacteria or to insert genes that make the bacteria susceptible to specific agents. Beyond not having to continuously develop new antibiotics or restrict treatment, these could also be much more targeted treatments with less harmful effects on a patient's microbiome. Of course scaling up to mass production on the level of say penicillin will take some time, but the same is true for new antibiotics.
jjk166··on Venom and Hot Peppers Offer a Key to Killing Resistant Bacteria
I'm partial to concentrated ethanol.
jjk166··on Guitar tuner that uses phone accelerometer
I doubt the sampling rate is anywhere near what you would need to make out dialog in a sound recording. You might be able to tell who is speaking though if you had a voice profile.
jjk166··on Tesla is recalling its cheaper Cybertruck because the wheels might fall off
Tell me you're not an engineer without saying you're not an engineer.
jjk166··on Tesla is recalling its cheaper Cybertruck because the wheels might fall off
> But they could have included an error factor in the designing process.

They almost certainly did. But that error factor is a guess based on limited testing. You never know your true variability until you're building at scale. Waterfall development doesn't work in the real world any better than it does in software.

jjk166··on Tesla is recalling its cheaper Cybertruck because the wheels might fall off
And if engineers were all knowing that is how it would work. Unfortunately the real world is complex and there are limits to how variability can be constrained.
jjk166··on Tesla is recalling its cheaper Cybertruck because the wheels might fall off
It's worth noting that crack formation is affected by more than just the design - variation in material and manufacturing steps could also contribute. A more robust design can potentially compensate for material or process variability, but those variables were likely not known nor knowable during the design stage. We should not boo companies for acknowledging and correcting issues which may not have been reasonably foreseeable.
jjk166··on Child marriages plunged when girls stayed in school in Nigeria
> Most research indicates that child support programs tend to just support people that already planned to have children.

I mean that's the point. We don't want people who have no interest in being parents to become parents for a paycheck. What we want is for people to have the financial freedom to live as they wish to live. In the US 46% of parents of young children report they have fewer children than they'd like due to financial constraints, and 23% of gen-Z report financial concerns as a primary motivator for not having kids. We don't need to go from 0 to 2.3, we need to go from 1.6 to 2.1. That extra half child is gonna come from tipping the scale for someone who is already on the fence, not paying people to be incubators.

← PreviousPage 7 of 34Next →