Air traffic controllers pushed to the brink | Hacker News Reader

Air traffic controllers pushed to the brink

nytimes.com

229 pointsren_engineer335 comments

non-paywall link: https://archive.is/ld9Bm

https://web.archive.org/web/20231203014618/https://www.nytim...

Oh, thank you. I see that full links do not require javascript and captchas with cars and traffic lights. I though that web-archives becomes plainly unusable, but it seems that it is only for short links.

I wonder, is there any way to expand short links without captchas.

archive.org and archive.is are different services

I think this is probably something with your DNS or moreso with archive.is's policies https://news.ycombinator.com/item?id=19828702

There's an increase in incidents because ATCs are overworked and understaffed. The knee-jerk reaction is to try to automate the work away. A few problems with that kind of thinking:

- Automation already exists - e.g., the collision detection warning mentioned in the article. I don't know much about air traffic control but I imagine there's some element of defense in depth. ATCs dozing off or running on fumes won't cause collisions right away (these incidents are mostly close calls) but it sounds like it's only a matter of time before all the other defenses fail and a major accident happens.

- There's always going to be work that's hard to automate. Plenty of software teams have on-call rotations to cover production services. You'd think that if any field would have their operations be fully automated, it's software, but nope. We need to stop thinking that the physical world is easier to deal with - it's often much harder.

- Automation costs money to develop and deploy. Apparently one of the early outcomes of union action back in the 70s was faster automation rollout. [1] For an executive, why spend money on automation if you could just squeeze more out of your workers and compromise on safety instead?

[1] https://en.m.wikipedia.org/wiki/Professional_Air_Traffic_Con...

> Plenty of software teams have on-call rotations to cover production services.

There are plenty of valid criticisms of the cloud but this is what you are really paying for. The value proposition of the cloud is infrastructure-oncall-as-a-service.

You usually still need someone oncall who determines whether it’s a cloud problem or your bug.

RAAS (Runway Awareness and Advisory System), the equivalent system installed on aircraft, also needs to become standard equipment. It is for some airlines, but it is by no means mandatory.

Automation in ATC exists a little yes, as you said for collision detection, ground proximity warning and runway infringement. That's mostly it.

We have some cool toys to make life easier like CPDLC to communicate via data-link instead of voice, multi layer radar screens with tons of functions, ATIS generation etc but those are not automation, just QOL improvement and if anything goes wrong we go back to old school ATC by voice and maths.

There is a lot of automation in the data part of ATC, so flight plans processing, radar correlation, meteo data etc etc. But this is just normal IT stuff in 2023.

Some cool project are getting there, like virtual towers with augmented reality by using cameras instead of looking out a window. My company did recruit an expert in big data and ML to develop new tools and automate stuff, but always data related, nothing to do with controlling and aircraft on a radar screen.

The lack of automation is causing ATC work to be stressful. People orchestrating complex procedures over a shared vhf radio channel is inherently hard and error prone. Which is why it involves people endlessly confirming and double checking known bits of information.

Automation could simplify that a lot and remove a lot of the uncertainty. Making communications easier might save some lives. A lot of accidents are caused by human pilots and typically that involves them making mistakes in stressful situations. Dealing with ATC is very stressful for pilots. Task saturation is something they all fight throughout their career. IFR pilots have to train and fly a lot to stay current; the ability to keep up with the radio communication and fly procedures safely is quickly lost when they don't. That's how hard it is. A pilot that is not current is not a safe pilot. That's why the rules are so strict for this.

If you look at this space from a distance, it's kind of obvious that more automation is going to be inevitable. Airspaces are busy now and they are going to get a lot more busy in the next decades. At least part of the traffic is going to be autonomous. Those planes will likely be communicating with ATC via networked protocols that have yet to be proposed and standardized. There is no reason why those protocols could not work with other planes. And it would free up ATC staff and pilots to do more urgent things. The way ATC is currently done is not scalable or future proof.

> The lack of automation is causing ATC work to be stressful. … > And it would free up ATC staff and pilots to do more urgent things.

There’s a problem that can happen with this, though, wherein the “easy” routine work is automated out and the number of humans involved is scaled down, but the work those people do then winds up being 100% the urgent/stressful stuff.

I’ve heard that ER doctors have been experiencing this in recent years–not quite with automation so much as just offloading, as less serious cases are picked up by nurse practitioners/urgent care centers, staffing has been adjusted and the actual physicians wind up with a steady stream of stressful high-acuity cases back to back to back.

I agree we might need a minimum staff level of, say, 1 air traffic controller per airport, but unless I've been grievously misinformed about the safety of flying, urgent/stressful emergencies should be very rare.

I suspect the main thing limiting ATC automation is the fact ATC is all done by voice using incredibly shitty radios, and equipment upgrades take like 50 years to roll out.

> I’ve heard that ER doctors have been experiencing this in recent years–not quite with automation so much as just offloading, as less serious cases are picked up by nurse practitioners/urgent care centers, staffing has been adjusted and the actual physicians wind up with a steady stream of stressful high-acuity cases back to back to back.

Yet at the same time, ERs (near me at least) are still inundated with so many patients that average wait times are 4-5hrs.

Maybe you are in an oasis of reason where this doesn't happen. But most ERs I have heard of that are overcrowded is because people go to the emergency room when there is no actual emergency.

I’m sure there’s a number of reasons for this - after about 5 PM, your options are typically the ER or an urgent care. UC’s usually require payment up front, and for people who either don’t have insurance or who have high deductibles, the cost of care at a hospital (which is way more but can be put off) vs the up front cost of urgent care probably pushes people towards ERs.

There’s likely other factors - lack of medical education/first aid, consulting Dr Google and then having anxiety that maybe that slight pain is something horrible, etc.

A public health system could alleviate this stress, if done well. Clinics that are open later than “business hours”, readily accessible telehealth providers, and better public health education could all come about from a public system. Privatized, all of these can only exist with a profit motive, and the insurance and health system are costly.

Isn't that how it should be? Why should a highly paid doctor ever be involved in a routine case that could easily be handled by a nurse?

As an engineer I only get tickets that are too difficult to be handled by support and require code changes. Why should medicine be any different?

Software doesn’t quite have the same fungibility to allow for this, but imagine your job was only being responsible for urgent, high severity tickets throughout your company.

There was no time to mull over a problem, pair code with a peer or junior, or talk to a product owner because everything was urgent and on fire.

Another reason not to scale down the number of humans is that they're needed when automation breaks.

Planes could very well fly themselves 99%+ of the time, but you want to have pilots when there are in flight issues that aren't handled by automation.

They've been talking about automated ATC for at least 20 years now, in part for getting the benefit of more flights arriving and departing per hour than humans can safely manage. https://en.wikipedia.org/wiki/Next_Generation_Air_Transporta...

They have been working on automated ATC for a lot longer than 20 years. I worked at Eurocontrol 1991-1996 where they had been developing a system called ARC2000 since well before I arrived. As the name suggests it was intended to take over ATC in Europe by the year 2000. I worked on a less ambitious system that helped visualise conflicts for manual resolution but that was not a big priority so I eventually left.

My internship in the mid-1990s was with a professor studying VR systems for ATC for low-visibility operation. At the speed this field changes (or doesn't), I doubt I'll see it in my lifetime.

I'm sure it would be done better if implemented over audio, but the image of a frustrated pilot yelling "REPRESENTATIVE! REPRESENTATIVE!!!" just came to mind.

An uncurrent pilot is not inherently an unsafe pilot.

> Dealing with ATC is very stressful for pilots.

This is not my experience at all, and I suspect is also not at all for air carrier pilots who deal with ATC on every flight.

Sure, the guy in the 170 or J3 who mostly flies off his farm might be stressed by flying into the big city airport, but the vast majority of pilots who interact with ATC regularly don't find it particularly stressful.

In fact, given a choice (and as a VFR pilot I have the choice), 99 times out of 100, I will voluntarily call up and receive flight following from ATC. There's pretty much no down side.

> Those planes will likely be communicating with ATC via networked protocols that have yet to be proposed and standardized.

The protocols are not the issue, it’s the high-bandwidth, low-latency wireless network that’s missing. LEO satcom may be an option in the future, but it isn’t yet for flight and safety critical real time applications.

I just finished building a pilot information system for an airline that uses ACARS to communicate. It’s a backup system for a backup system that supports a redundant electronic flight bag tablet.

Instead of more automation, perhaps it would be good to occasionally introduce phantom traffic which is designed to require intervention on their part (e.g. fabricate two flights and put them on a collision path).

The air traffic controllers would be expected to flag them, or intervene, or do whatever they would realistically do in that scenario.

From that, you can determine who's asleep at the wheel so to speak, and send them home. If necessary, planes may need to be grounded until properly alert staff can be available

Not sure how that would work.

"Citation one-three-niner-delta-uniform, Denver Tower, adjust altitude to seven three hundred, bearing one zero five."

"Citation one-three-niner-delta-uniform, Denver Tower, please respond."

Also FYI it's not like anyone can actually be asleep because pilot is required to establish two-way ATC contact to enter the airspace.

Car accidents are up as well, less so this year than last. Both Air traffic and Car accidents have been up since 2020. Its not unique to Air traffic control.

https://www.forbes.com/sites/tanyamohn/2023/09/28/road-crash...

Its also happening worldwide, more air traffic near misses in the EU alongside road accidents as well. I suspect its world wide.

I don't really see how that's related though

ACT test scores have plummeted too. Combined with what we know about the physical brain changes that can be observed after a single mild infection, it all paints a potentially concerning relationship with COVID.

Could also be the consequences of schooling during a lockdown.

People continued to drive.

A lot of people changed their driving habits significantly, and there was a period of significantly less driving. Many people commute much less than before. At some points, traffic was so much below the usual that several record breaking Cannonball runs were had that are unlikely to be broken unless there's a significant relaxation of speed limits nationwide.

> Could also be the consequences of schooling during a lockdown.

Luckily we have the control-group of Sweden, that spent a large part of the pandemic proclaiming that not closing schools was going to have a fantastic positive effect on children's learning.

The actual result: Sweden's PISA results declined much more steeply than countries who used distance-learning during the pandemic.

For yet another round, East Asian countries sit at the top of the PISA rankings. These are also the countries that had some of the strictest responses to CoVID worldwide.

Those things don't necessarily have any relation, but clearly many countries with strict CoVID responses are still doing well in education.

Or maybe ACT test scores have plummeted because these students have experienced 1-2 years of remote teaching, which might have been less effective than in-person school?

Per the comment thread, this doesn't have perfect explanatory power: drivers and ATC have also recorded statistically significant declines over the same time period.

In other words: remote schooling may have had a negative effect on ACT scores, but all evidence points to it not being the only (or even primary) effect.

Regarding this thread, I really want to caution on the desire to automatically assign worse car crash rates to covid-19 infections, or any other specific cause.

Car crashes and ATC mistakes are two different sets of data. Just because they are going up or down at the same time and you can overlay some other event on top of it means absolutely nothing on its own.

There are a lot more reasons that car crash rates can be getting worse. For one thing, they have been trending upwards since smartphones have become prevalent. Car crashes weren't trending down for some time before covid-19. Also, the number of screens installed on automobile dashboards is increasing every year.

Still, I won't automatically assign blame to smartphones or automotive screens unless someone at least attempting to use the scientific method weighs in on the subject, because our intuitions about statistics are often really, really wrong.

Distractions might be going up in cars, but I’d guess that’s offset by automatic crash avoidance systems that’s also going up in cars.

I had a new car for a month (mine got damaged whilst parked). Simply adjusting the aircon fan required screen navigation followed by a fidly touch screen control.

The crash avoidance systen was a distracting alert that took my attention away from the road for none risks but instead increased there likelyhood to become a risk.

Glad to get my own car back with physical switches and dials without all this crap, having to operate a tablet behind the wheel is a dumb idea and no different to operating a smart phone behind the wheel.

> Car crashes and ATC mistakes are two different sets of data. Just because they are going up or down at the same time and you can overlay some other event on top of it means absolutely nothing on its own.

A good time to dust off this classic:

https://www.tylervigen.com/spurious-correlations

Also https://commons.wikimedia.org/wiki/File:Pirate_Global_Warmin...

Except that COVID is known to be brain invasive and to affect cognition.

This looks more like refusing to ever entertain the idea that letting COVID rip was a bad idea and lockdowns and vaccinations and masks were justified because you don't like it, than actual scientific rigor.

Is the cognition impact less-so if vaccinated? How much, if so?

As far as the vaccine leading to more mild infection goes the impact seems to be unclear. There's been some evidence that disease severity can have an effect on how bad the cognitive issues get. Some studies have suggested that people can end up with greater cognitive impairment if the infection got bad enough to result in hypoxemia/hypoxia, or if they were put on ventilators, or spent longer amounts of time in the ICU, but other research didn't show any link between time in the ICU and the severity of cognitive issues, and others found that people with mild infections might even be more likely to have problems than those with severe cases.

As always, vaccinated or not, the best thing you can do for yourself is avoid getting infected in the first place.

> Regarding this thread, I really want to caution on the desire to automatically assign worse car crash rates to covid-19 infections, or any other specific cause.

People’s tolerance levels and general concern for other humans plummeted in the post covid times. The warning signs about acceptable behaviour in supermarkets, medical practices and restaurants proliferated. I don’t know how you measure that effect though.

Right just look at the number of people having to be restrained on airplanes and just outright acting out.

I worry that people may just feel like we're doomed and it's causing some to lose it.

That's a self-fulfilling prophesy, and an unstable feedback loop

If it helps, I certainly not so worried that I'm going to flip out on minimum wage employees or ground an airplane in a full blown "Karen meltdown", but I do think that the pandemic has left a lot of people disillusioned, that Americans are seeing their standard of living decline, that climate change has left our future uncertain, that the stability of our democracy is in question (more than normal anyway), and that doom scrolling makes it easy to ignore anything but the worst things going on around us. It's difficult to imagine that those pressures aren't contributing to the problem anyway.

I don’t know what grocery stores or restaurants you go to but I don’t see this supposed lack of concern for others that you’re talking about. It seems the same to me.

> Or maybe ACT test scores have plummeted because these students have experienced 1-2 years of remote teaching, which might have been less effective than in-person school?

As I mentioned in another post, you can check this theory against the pandemic control-group that kept schools open:

https://www.thelocal.se/20231205/swedish-students-maths-and-...

You're drawing a very simple conclusion from quite a complex article though. Some of the things it is noting:

* School closures during the Covid-19 pandemic were one factor behind the unprecedented decrease across the OECD.

* The general decline also began before the pandemic.

* Immigration was quite probably a factor.

Even without getting in to the details.

colleges still care about ACT? I thought SAT was the golden standard.

I remember taking the ACT many years ago and it was definitely much easier. Also not many colleges accepted it at the time. SAT was preferred.

The preference is regional. [1] When I took the ACT, many years ago, it was the test the university I was applying to listed first in their requirements.

And that's still true.

[1] https://en.m.wikipedia.org/wiki/File:SAT-ACT-Preference-Map....

When I was in high school (graduated 2012), my school recommended we take both, since not only do certain schools prefer one to another, but if you happened to score relatively well on one versus the other, you could choose to only send that one instead.

> if you happened to score relatively well on one versus the other

This should be reasonable proof that the test doesn't test things very well, IMHO, because there are so many factors that go into getting a good score.

I was one of these people -- I got one point below perfect on ACT, but the morning of my SAT, I got into an argument with my mom and got a below-average score.

1. No (nontrivial) test is perfect: we take each with a grain of salt and accompanying context 2. You can take these tests multiple times and use the highest score, reducing impact of one-off ‘off days’ for whatever reason 3. These particular two tests do not claim to assess the same thing (aptitude vs achievement) and do so through different implementations

These factors limit me from tending towards ‘reasonable proof’ on its face when considering the system as a whole.

> This should be reasonable proof that the test doesn't test things very well, IMHO, because there are so many factors that go into getting a good score

Yep, I don't disagree! I happened to get scores that converted to each other exactly according, but overall I think it was good advice to try to use the system to our advantage rather than treat it as some sort of objective measurement.

This might be the worst real example of a jump to conclusions based on a casual correlation I have seen in my entire life.

I don't even know how to properly deconstruct how irresponsible this style of thinking is, it's just so flabbergasting.

In science you can't just draw lines on the wall between random shit like you're Charlie Day in It's Always Sunny in Philadelphia.

ACT scores are very obviously not a product of your raw brain power as an isolated variable, they include a number of other factors like how much sleep you got last night, your family life, the quality of your curriculum and teachers, your socioeconomic status of yourself and your peers, and, of course, whether you spent a year in a global pandemic emergency while your school district scrambled to switch to a fully online curriculum with zero notice.

There is a very well-documented labor shortage for air traffic controllers. Fewer people are covering the same shifts - that is fact. If we are going to ignore the scientific method and start pointing fingers, this is so much more obviously the type of thing to start scapegoating.

Not sure of the specifics, but often shortage of labour leads to lowering of standards of training/min experience required as well.

> potentially concerning

> This might be the worst real example of a jump to conclusions based on a casual correlation I have seen in my entire life. I don't even know how to properly deconstruct how irresponsible this style of thinking is, it's just so flabbergasting.

Exactly who is jumping to conclusions in this exchange?

I used the words "potentially concerning" because that is what I thought and intended to express. I could have used words like "causally disastrous" instead, but they weren't what I thought or intended to express. I am also not smuggling in such a claim in a "just asking questions" sort of way. I think the potential exists, should be studied, and shouldn't be reflexively admitted or dismissed until studied. That's all.

Please respond to what people actually say, and ask questions if what they said wasn't clear to you.

I read your comment carefully. I think it projects more confidence in the theory than you might realize.

The word “potentially” is the lone word that makes any attempt to pump the brakes.

The same as the PISA scores. PISA, which might not be well known is the US, is the International Student Assessment for the 15 years old students in the OECD (and non-OECD) countries. The PISA measures their ability to use their reading, mathematics and science knowledge and skills to meet real-life challenges. The PISA scores for 2022 is the lowest in the history of PISA and show a clear trend downward since 2012. The huge drop in 2022 scores can be explained through the longer school closure period during the pandemic lockdowns. But only partly. The increasing prevalence of mobile phone amongst children and their distractions certainly play a significant part in contributing to the huge drop in PISA scores.

Not necessarily related to the increasing number of air traffic controls incidents, though.

I'm not sure phones can explain it. Even when I was in high school 10 years ago everyone had a smartphone with social media.

You didn't have the same platforms nor the same dynamics on them. Social media has evolved over the past decade.

Yup. Things like filters & algorithms have "improved" significantly.

Definitely. But then smartphones are not to blame. The algorithms are.

But you didn’t have ubiquitous smartphones and tablets when you were 6-10 years old, and weren’t handed a smartphone to head off a tantrum during the absolutely critical brain development years before school.

I don't think there is any evidence that smartphone usage by kids has any impact on brain development.

> The huge drop in 2022 scores can be explained through the longer school closure period during the pandemic lockdowns

The latest PISA scores arrived just last week or so and do not support this.

There is a very weak negative correlation between number school closure days and drop in PISA scores, at least in Western European countries. The higher the number of school closure days, the less the results dropped. But the correlation is so weak that it's basically just noise.

Scores dropped pretty much everywhere so the pandemic probably had an effect, but the data does not support school closures (or at least its duration) of being the primary culprit.

I do not have a good English language source to point to (but the raw data is easy to find).

Or spending years being overwhelmed by smartphones mainlining information to us for multiple hours a day is finally catching up with us?

What about the effects of an extremely wide and extremely quick deployment of a novel vaccine (under threat of being fired)?

What does "novel vaccine" mean?

That we haven’t used mRNA gene therapies to treat a virus before.

Maybe not that widely, but the technique seems to have had prior usage.

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

In bovines yes. And they laughed at Joe Rogan for using "horse medicine"

It's been studied intensely since its discovery in the 1990s, it's not in any way a "brand new" technology.

A vaccine that had a significant impact on a notable percentage of the population 24 months before widespread uptake.

The Thiotimoline of vaccines.

https://commons.wikimedia.org/wiki/File:Historical_Average_A...

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

Spelling this out slowly for the reflexive downvoters, ACT test scores began their decline well before any widespread use of mRNA COVID vaccines which would make them extremely novel in their endochronicity if they had an effect before they were used.

Otherwise, as pointed out by lostlogin, mRNA vaccines have been used in humans since 2013 and studied in animals since 1989 which makes them less novel than widely claimed.

Your problem is thiotimoline sounds too much like thalidomide, the thing that made the FDA proud of their excessively strict medication regulations.

What about the effects of repeat infections by a very contagious and neuro invasive novel virus? Hmmm?

Yes also that.

Amazingly enough, it turns out it's easier to test a vaccine when there's a pandemic happening.

Since vaccine testing requires giving half your study population a placebo and then waiting for them to get sick. Bit cruel really.

> Traffic fatalities declined for the fifth straight quarter. About 19,515 people died in motor vehicle traffic crashes in 2023, a decrease of about 3.3% compared to 20,190 fatalities in the first half of 2022.

I don't see much correspondence with:

> In the fiscal year that ended Sept. 30, there were 503 air traffic control lapses that the F.A.A. preliminarily categorized as “significant,” 65 percent more than in the prior year, according to internal agency reports reviewed by The Times. During that period, air traffic increased about 4 percent.

Especially since the NYT blames the latter on "A nationwide staffing shortage — caused by years of employee turnover and tight budgets, among other factors — has forced many controllers to work six-day weeks and 10-hour days."

Traffic fatalities decrease because cars are getting safer.

But the number or accidents per capita might not be. Pedestrian deaths have certainly been on the rise.

https://www.ghsa.org/sites/default/files/publications/files/...

Pedestrian deaths also increase because cars are getting bigger - reduced visibility, much more energy, and worse impact zones.

Cities should really start to regulate what kind of cars under what conditions they let mix in with people.

Sadly, we're going in the other direction. For example yesterday the capital city of Poland (Warsaw) decided, despite a lot of protests, to restrict "old cars" from entering a number of central districts. "Old cars" will mean diesels older that 18 years initially, 13 years later and petrol cars older than 20 years. The protesters proposed the rules are amended to block cars based on their emissions instead (which would allow most of the small old cars people still have, but it would block many modern SUVs). Of course the ruling faction (still enjoying majority support in Warsaw) decided otherwise. So much for the policy to be driven by "health concerns" and not car dealership lobby.

Restricting diesels older than 18 years old is perfectly reasonable - the amount of PM 2.5 those things put out significantly contribute to childhood asthma etc.

By comparison, London's ULEZ restricts diesels older than 2016.

Of course there's no reason why vehicles couldn't be restricted both on CO2 AND PM 2.5 emissions, if you want to promote small, clean cars

People are driving spaceship-sized SUVs while using their phones, hitting pedestrians who are walking around looking at their phones, jumping out of the path of an oncoming 12yo doing 40kmh on a battery powered scooter while looking at porn on their phones and using Whatsapp to schedule a meet up with their Fastgas supplier. Their parents think that a scooter is a toy, the phone is so that they can call home, and the money was for lunch.

Yeah, kinda half joking, in the sense that these 3 hypothetical people don't actually crash into each other every day, but they all exist. I see them every day.

I've also seen a huge rise in cyclists on the footpath (sidewalk), and it's so normalized that I had a guy in his 50s zoom up behind me and the wife and actually tell us to get out of the way. To say that he got called all the things I could possibly call him is an understatement.

How does a technical gas supplier factor into the life of today's teenagers? I'm genuinely curious. Is our hypothetical 12yo person learning tig welding on a side?

Same as sniffing glue... nitrogen aka N2O aka laughing gas.

Nitrous oxide is a popular inhalant. It's reasonably safe despite the usual hysterical claims, and it's unregulated in most countries, possibly because regulating it would drive users towards more harmful inhalants like glue and petrol.

It's not safe at all. In my country there are many people who have destroyed their brains using it. I would say even drugs like cocaine are safer.

It's just as safe as choking yourself, the gas itself isn't causing any damage, but the oxygen deprivation may.

That's already an improvement over many other drugs, where the active ingredient is actively harmful.

It's comparatively safe. See e.g. https://journals.sagepub.com/doi/10.1177/20503245221127301. Cocaine is a very dangerous drug both because of its physical effects - cardiac issues, stroke, seizure - as well as its effect on risktaking. In order to "destroy your brain" using nitrous you need to take heroic quantities of it and get hypoxia: from memory somewhere near a hundred times the typical dose is required, and it must be maintained over a long period of redosing. For more dangerous drugs, that multiple is lower: heroin can be lethal at 1.4x the typical dose.

It is wrong to claim it's an especially dangerous drug, particularly if your evidence is an anecdote about "many people who have destroyed their brains" in a nameless country based on no quantitative evidence and citing no personal experience.

You don't seem to know the most basic reason why it is so dangerous, it's not hypoxia. The problem is that it oxidizes vitamin B12 which leads to demyelination and is neurotoxic even in a single dose(Although the harm seems to be reversible in that case). This is what causes the neurological symptoms. Personally I would much rather die from a heartattack or a stroke then never walk again.

Give me a source for B12 deficiency being serious enough that it's comparable to the health risks associated with cocaine, of all drugs. All the reports I'm aware of involve extreme levels of daily use, e.g. https://journals.lww.com/jfmpc/fulltext/2020/09110/recreatio... ("75-100 canisters daily"). The idea that a single dose will oxidise enough B12 to be clinically significant is not supported by evidence, it's something you've made up, unless you're claiming a dose is made up of 75-100 canisters.

Personally I would rather compare the actual incident rates of side effects from drugs than make a false choice between the most extreme symptoms, particularly when one of them is a common and relatively safe form of anesthesia.

First off. It's really not nice throwing out terms like "it's something you've made up". But anyway here are some pointers why I think that Nitrous oxide is worse then cocaine, these are stats from the Netherlands because that's where I did the research:

A single dose shows traceable, but reverseable, neurotoxic effects in rats:

https://www.sciencedirect.com/science/article/abs/pii/S03064...

> Our findings demonstrate that short-term exposure of adult rats to N2O causes injury to PC/RSC neurons that is rapidly reversible, and prolonged N2O exposure causes neuronal cell death.

In the Netherlands the lower bound of people with paralysis, such as beeing unable to walk due to the effects of using Nitrous oxide is 64 in 2020. And that is by querying a subset of hospitals. These are indeed extreme usage cases. You also don't get a heart attack from snorting a few lines of coke.

https://nos.nl/artikel/2338520-64-mensen-met-gedeeltelijke-d...

This is much more then the yearly reported deaths for cocaine in the Netherlands. Which is reported to be 38 https://www.jellinek.nl/vraag-antwoord/hoe-vaak-gaat-er-iema....

If you compare the number of users in the netherlands you can see the picture how bad it actually is:

- Cocaine had 230k users in 2020

- Nitrous oxide had 290k users in 2020

https://www.trimbos.nl/actueel/nieuws/dit-zijn-de-opvallends....

This means you are much more likely to get paralysed by Nitrous oxide usage then you are to die from cocaine.

It's definitely not a good idea to label NO2 as a safe drug.

That's worse than I was aware of. Thanks for putting some rough stats on the size of the problem.

At the same time, in the US in 2021 there were 7.1 cocaine-related deaths per 100k[0], and 1.7% of the population had used cocaine in the past year[1], so their mortality rate is more like 417 deaths per 100k. That's hugely higher than the rate in the Netherlands and I wonder why there's such a difference.

[0] https://www.cdc.gov/nchs/products/databriefs/db474.htm

[1] https://nida.nih.gov/publications/research-reports/cocaine/w...

>That's hugely higher than the rate in the Netherlands and I wonder why there's such a difference.

Interesting. I wonder if average drug purity and common adulterants make up for this difference? What is the average "street" cocaine purity in the US vs the Netherlands? What do they cut it with? Inert substances? Amphetamines? Meth? Mephedrone? Who knows.

Also, what about the mode of taking it. Is there huge variability in how the users take it between the two countries due to different culture? (snorting vs smoking vs injecting, also multi-drug use). Finally, what if in one country this particular drug is more popular with the older population while in the other it's "all the rage" with the younger people? Death statistics will be different too.

And those empty canisters are everywhere, thrown around parks and playgrounds

Nitrous oxide is certainly less harmful than many recreational drugs but it can cause nerve damage in some chronic users.

https://www.qmul.ac.uk/media/news/2023/smd/young-people-with...

In the Netherlands ebikes are a game changer because many people can now commute between cities on a bike (bike lanes are alongside motorways too). But they are so fast! There’s been an increase in dangerous accidents with them, particularly among the older generations.

Electrification is great but it has risks too.

Most of the kids I see on scooters and bikes aren't legal. I have no idea where the kids are getting them because they're not "pedelec". You push a button and accelerate up to 40kmh. They're essentially electric mopeds and they require a driving license here (and anywhere in the EU I believe) which makes them illegal for under 16s to ride them.

Legal ebikes in the EU stop their electrical motor at 25 km/h. You can go faster, but only as much as you would with a regular bike, ie by pedaling. Therefore ebikes are not the problem as such, illegal vehicles are the problem.

It's still fast for cycle paths where everyone is accustomed to people cycling manually, and it's definitely not just illegal vehicles. See: https://www.fietsersbond.nl/nieuws/voorlopige-cijfers-laten-...

> In 2022, at least 578 road accidents killed at least one, in 2019 there were 522. The number of accidents with at least one injured person has also increased considerably: in 2019 there were 18,298, in 2022 no less than 21,455. According to the police, the e-bike appears to play a greater role, especially with people over fifty. The figures do not necessarily mean that cyclists on an e-bike caused the accident, but it is important to see where the increase comes from.

I've been cycling for 45 years, and I don't think that the addition of a small pedal-assist motor has changed my cycling style at all. I'm just willing to take on 100km journeys or mountain roads that I previously considered to be too much effort. My speed also hasn't changed that much either, unless you're looking at my uphill speed.

But the arrival of relatively cheap ebikes has changed the demographics of cycling in a very short time. Lots of people with little-to-no road experience or etiquette are now zipping around at up to 25kmh. These will definitely appear as a spike on the accident graph, but for pedelec I'd expect that to drop a little over time.

However I still think that pure throttle bikes and scooters that can exceed 25kmh are dangerous though. Particularly the scooter which has the image of a "toy" to a lot of people. I'm really not kidding when I say that I've seen 10 year olds on serious scooters, reaching 40kmh on walkways. That's always gonna be a problem.

To be fair, the responsibility to not-kill lies squarely at the hands of SUV drivers.

SUV -> human/scooter collision is the only one that leads to fatalities.

Also remember, we have infrastructure for cars, but very little for pedestrians or cyclists. Cyclists don't want to be on the road. They're much rather be on grade-separated bike paths. I don't endorse cyclists on footpaths, but in a lot of cities it is legal if the road not have adequate bike infrastructure.

Similarly, if pedestrians are walking on a green, they should be able to be on their phones. I do not endorse that behavior, but it is not like they are responsible for coming under death machine driven by another driver.

Your comment appears to distribute the blame across all 3 people. But 1 of these is 2 orders of magnitude more to blame than the other 2.

_____

I do agree that any going at car speed (Motorized scooters, Lycra-clad racers) belong where cars are: the road. A helmet can only do so much if you face plant at 50 kmph. Anything that can go above 30 kmph on a flat, belongs on a road.

Trains are running more safely, efficiently, elegantly, and reliably than both cars and airplanes, and they have consistently done so with a much longer track record.

Not Amtrak. Worst excuse for a railroad ever.

No experience myself, but from past comments on HN, I understand the company Amtrak provides pretty good service within the areas it can control. It's just that the US rail network prioritizes freight every time (even though legally it's not supposed to).

That’s the official party line, but it doesn’t quite capture the intense disorganization and indifference the company often presents to its riders.

those HN posters are just rabid libertarians hell bent on dragging any and all services that are not privately owned and operated for profit.

I don't love Amtrak on the East Coast. But on the West Coast, it can be really nice. Coach on the West Coast is like Acela on the East Coast. Even the cheapest seats have nice tables with outlets, and the views on the Pacific Surfliner can't be beat!

I followed road traffic safety since 2020 closely (and have submitted and posted about it here) and traffic fatalities is slowly coming down since 2020. My hypothesis about it was that the empty roads in 2020 caused drivers to degenerate in terms of road manners. For example, deaths per mile actually increased in 2020 despite people driving fewer miles which suggests people drove less safely.

On the other hand this article says there was a 65% increase in significant incidents from the year prior to Sept. 30. So I think these are two different issues.

My pet theory has been that extended lockdowns caused regressions in people’s ability to drive safety, whether it was the actual skills to operate a vehicle, or people’s risk assessment skills.

Operating a vehicle is a skill like any other, and if you’re not performing reps you’ll regress like any other activity. Empty roads might even permit crappier driving.

ya that makes sense. we'll probably never know what happened :|

Perhaps, maybe, mass infection by a virus that affects the brain and vascular system has consequences for society.

It's an interesting hypothesis but should be easy to corroborate with other data, e.g. test scores. I would assume we'd have heard of that?

Isnt it a growing opinion of American teachers that students are underprepared? probably a combination of covid and remote learning... I doubt we will see the long term effects of this any time soon.

We have. The scores have fallen significantly.

If we're drawing connections between 2020 and driving behavior, the much more obvious correlation would be the massive anti-police-violence protests, the accompanying increased negative public sentiments re: police, and the police reaction that occasionally gets brought up on HN, decreased policing as retribution.

Underrated comment. Police have "quit" in more than one major city in sfbay.

I disagree. However, I'm used to going long stretches without driving and then suddenly driving hundreds of km/miles. I suspect, based on my wife's behavior and her driving, that before 2020, people were used to and happy with simply listening to music, a podcast, or something. After/during 2020, many people were no longer satisfied with that. They turned to listening to YouTube, movies, or even face-to-face calls while driving (hell, even my mom does this one), and holding the phone while trying to operate the vehicle.

It's almost like people forgot the most important part of driving: minimizing distractions. If I'm in a high-stress part of a drive (unfamiliar roads, high density, etc), I always turn off all media and ask all passengers to be quiet for a bit because I need all my senses. I don't know how many times I've been saved by hearing a car that I can't see yet.

And mine is that the lockdowns felt like an overreach and there was some kind of infantile rebellion

I’ve noticed a pretty drastic decline in driving behavior around here. We’re seeing a LOT more just absolutely insane drivers doing things you’d only see in the worst of compilations a few years ago. Something happened during Covid that made people just not care anymore. Also doesn’t help that cops are not pulling people over as much as they should. Watched someone just yesterday fly through a red right in from of a cop that did nothing about it.

During peak-Covid, the roads were sometimes at 1/10th normal capacity, which also contributed to a weird driving euphoria, imo. Now the traffic and traffic jams are back, and peak speeds are returning to pre-Covid levels.

Yes. I think about this all the time. It's so strange, and aside from internet comments it doesn't seem to come up in the mainstream. What happened to people? Why do they not care? It's very evident in driving, but shows up in other areas as well.

Everything is going to shit.

It’ll hit you soon too.

Early in the pandemic some places stopped requiring road tests to get a license. There are probably quite a few drivers on the road that would never have passed.

Ya, to corroborate that I know road rage deaths increased in certain areas. There is evidence that people started speeding more too. Since people started speeding during 2020 (deaths per mile went up as total miles traveled went down) my thought was that maybe people became accustomed to not being policed and just started driving much more dangerously and then had a hard time adjusting back. I don't know. Reduced policing could also explain it, or maybe just a general decline in socialization. No one will look into it probably.

> Car accidents are up as well, less so this year than last. Both Air traffic and Car accidents have been up since 2020. Its not unique to Air traffic control.

What's the connection between car accidents and ATC incidents that would make one a sensible comparator for the other? They aren't caused by the same things, unless the extra car accidents are all occurring in intersections where someone is directing traffic.

Don't recommend using 2020 as base given lockdowns resulted in much fewer miles driven

Very scary indeed. I think for the first time in my life I will be a little on edge flying in the US. You always hear that flying is the safest mode of public transport--well, if this is the state of the ATC, we're definitely in trouble. And you know there won't be any real change until hundreds of people die needlessly...I just hope it won't be my flight.

It's the safest per mile, but not per journey.

A nice comparison to other transport methods, including taking the bus, and the Space Shuttle: https://en.wikipedia.org/wiki/Aviation_safety#Transport_comp...

It is utterly fascinating to me that technically walking is just as safe as a car on a journey by journey basis. I definitely didn't need to know the danger per journey for flights though. I liked the lie I had before

Per that chart, flying is about 3x as deadly per journey as a car. Many people take a car to and from the airport, which means it nearly nets out already.

What do you mean, it nets out?

Shouldn't the car trips from/to the airport mean that an airplane journey is even more dangerous, as you can't really fly without those additional car trips?

Yes you can. I usually take the train to the airport.

I take a car to the train to get to the airport.

Sure, but two additional train trips also increase the total danger of the flight...

>can't really fly without those additional car trips

Hm, i almost never take a car to the airport.

It really blows my mind that the space shuttle was really so sound and safe even though I want to argue that it could have been far better. I mean this compared to SpaceX stuff blowing up a bunch before it gets the kinks worked out - I don't remember seeing or hearing about any prototype shuttles blowing up but it was before my time.

17000000 Deaths per billion flights for the space shuttle is not safe. This is more than 2,000 times more dangerous than skydiving, which is also not very safe.

It was significantly worse than that.

There were 2 failures over a total of 135 flights, with 14 fatalities. So that's about 103 fatalities per 1000 flights.

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

I think it's "passenger fatalities per passenger-journey". So need to multiply the number of flights by the crew on each.

Right you are, which brings us back to the ~1.7% number (actually 1.6% if you include the STS-135 mission, which Wikipedia doesn't for unknown reasons).

So a 1 in ~60 chance of dying on any given flight...have to agree that that is far from safe.

They did blow up 2 Shuttles out of 5, both crewed, so maybe they didn't get the all the kinks worked out in the first place.

Plenty of rockets did blow up on the launchpad or at various stages.

For example, the Titan I failed 17 launches, the Titan 2 failed 10 (and one exploded in the silo, while attached to a nuclear warhead), and the Atlas E/F failed 9. Even more recently, into the 90s and 00s, the Atlas I lost 3 and every Ariane variant has lost at least one.

I'm hardly a Musk fan, but although SpaceX seems particularly gung-ho about testing things they know aren't finished in order to iterate the design faster, the results after that are pretty good so far: Falcon 9 is 281 "real" flights for 2 failures, plus that one that went bang on the launchpad.

It's not 100% success rate like the Saturns, but out of 21 launches between the 1B and V, there were several near misses there. And probably quite a few A-4/V2s were harmed in the making of them and their predecessors too!

Others here have addressed the question of whether it was safe even by space flight standards, but as to whether it could have been far better, the refractory-surfaces problem alone seems to have put an upper limit on that.

These numbers are from 1990-2000. Airlines have gotten much safer since then, so I believe they would now be much safer per-journey, too. https://www.airlines.org/dataset/safety-record-of-u-s-air-ca...

(The difference between per-mile and per-journey is still interesting!)

As well as using extremely stale sources, the Wikipedia article fails to separate airlines from other forms of civil aviation.

US airlines have a fatal crash rate very close to zero. 'General aviation', which roughly means private light aircraft, is a very different world and has roughly 1 fatal crash per 100,000 hours flown, making it less safe than driving. Chartered flights are somewhere in the middle.

There's not much sense looking at the overall figure for aviation; the different categories really are very different.

There are so many disclaimers after that table that I feel that citing it as a "nice comparison to other transport methods" is absurdly simplistic and disingenuous. Especially since a sibling comment says "It is utterly fascinating to me that technically walking is just as safe as a car on a journey by journey basis."

> I will be a little on edge flying in the US

That doesn't make sense. The article is concerning, but the USA still has a solid track-record for aviation safety, and its airlines in particular have a fatal crash rate of effectively zero. (Non-airline aviation is a different world.)

There's a slim but real chance one of us will die in a car crash. There's essentially zero chance either of us will die in an airliner crash. That would still be true even if we were both airline pilots.

The US has developed a strong culture of believing that government is always the problem and government workers are always lazy and overpaid. And so cutting funding to these kinds of services becomes an easy sell. The problem is our society depends on these government employees. It’s no different than nursing, teaching, and many other professions - disparage it enough and eventually you’ll succeed in making no one interested.

Why do you think we have developed that culture? In my case it's because I know two people who have worked for the government or deal with it daily for their job, and they have told me that. And their low opinion of government is corroborated by my one experience dealing with the government.

None of this was with ATC, but my point is that this attitude you complain about is just a reflection of reality.

Thank you for your three anecdotes; we'll add them to the bucket.

Isn't this basically just some sort of weird selection effect? Similar to how people are more incentivized to leave a negative review when they're mad at a business, vs. a positive review when they had a good - or even just a "it was fine" - experience?

Let's say you are interviewing a candidate, and you have had one interaction with him, which was negative, and you know two other people who have worked with him extensively, and they say he is a bad employee. Are you going to say, well, my interaction was a one off, and those two people who worked with him could be wrong and that is selection bias, and hire him anyway, or are you going to take a pass? Why should I be so much more generous with the government?

Because the government isn't a single person you're hiring.

If I had three bad experiences at the DMV, that tells me very little about my possible future experience with my town councilmembers, or my local librarian, or my senator, or an air traffic controller.

“ The problem is our society depends on these government employees.”

That's called an earned reputation.

It’s a strategy by politicians that want to cut services so they can further lower tax on the rich.

Maybe I missed something, but has anyone cut funding to a single program in the history of tracking government spending? Especially the FAA?

People are seriously concerned with aircraft near misses when we let cars kill 45,000 Americans a year. Absurdity.

It’s possible to be concerned about two things?

Absolutely, but one is a trifling concern compared to the other. Yet the trifling concern gets the media attention.

What is your point, that newspapers should lay out A1 with "concerns" ordered whatever arbitrary metric you personally have decided is the "right" measure of importance?

Newspapers already are applying their own metrics for what's fit to print, seems reasonable to prefer one's own metric (if not reasonable to expect that to happen without action).

Alright, the media sucks. It’s amazing that people fixate so much on this obvious point, rather than be more concerned with rising ATC incidents or car accidents

Would you like people to not be concerned with aircraft accidents?

Certainly the concern is why air travel is so safe.

Only if the people who are concerned could be channeling their concern into solving the problem. Here, I am guessing, not many people can actually solve the problem, so we can safely be very concerned about both things simultaneously!

It would be different if it were the same exact group of people and the same budget who are in charge of plane safety, as those in charge of the car safety. We can then scold them for wasting time on planes and told them to shift focus to cars instead.

Stop making sense!

It would be useful if the readily mechanizable aspects of ATC-piloting and driving were done by machine but with human-in-the-loop. The fixations with completely human, manual processes and the leap to full automation are both too risky. There is plenty of middle-ground to derisk and remove points where human-error is mostly likely to occur. TCAS has been an excellent aid. Perhaps an automotive "TCAS" would also be useful.

> Perhaps an automotive "TCAS" would also be useful.

Forward Collision Warning systems and automatic emergency braking are becoming common in new cars. The problem and solution spaces are pretty different, so I don't think applying TCAS directly would be helpful --- I don't think there's a chance of cars having transponders anytime soon, and certainly deer, cyclists and pedestrians are unlikely to participate in an active transponder system, and there's no opportunity for cars to pull up or sink. At the same time, TCAS does not instruct to reduce or increase horizontal speed.

Automatic emergency braking is fucking horrifying tho... It has false positives and can be dangerous.

Is it more dangerous than people not braking because they are watching something on their phone?

Depends I guess.

I've only been in a car with two people who claim to get a lot of false positives on their AEB system while driving. Every time I saw a "false positive" it was almost always because they were driving very unsafely, trying to weave through heavy traffic at high speeds, approaching cars extremely rapidly and attempting to leave no room. In other words, not really false just not what the highly unsafe driver was wanting.

Meanwhile across many thousands of miles and a few different makes of AEB I've never once encountered an actual false positive. Every engagement has been pretty reasonable to me.

My one car with AEB has been pretty good with activations at speed, sometimes activating for a situation which wasn't really needed, but always understandable. The trickiest one is when a car in front is leaving the lane and slowing, but I wouldn't call that a false positive --- it's not stopping out of the blue or for a falling (tree) leaf; it detected the scenario properly, but not the nuance.

At parking speeds though, it would regularly emergency brake to save bushy flowers encroaching into the driveway, or one, it tried to save a driveway from being backed into (there was some weird pavement angle).

Nah my feet work and so do my eyes. I have had small accidents in the past and learned from them. I understand for people who don't learn or can't judge speed and need a crutch, but perhaps it's better they don't drive at all.

> I have had small accidents in the past and learned from them.

Maybe if you had AEB and FCW you wouldn't have had those accidents in the first place. Or did you not have working feet and eyes at the time and only recently got those installed? You even acknowledged you've had multiple collisions, did you not learn when getting your license? Did you not learn on the first accident? How many collisions should be expected for people to learn?

I'll be more clear. If your AEB is engaging while driving on the roads, you're probably driving unsafely. If you're worried about your AEB triggering often, you should probably adjust your driving because it's probably pretty unsafe.

I'll grant there's probably a few models with twitchy AEB systems, but almost every overly sensitive AEB I've ever heard about is either a Tesla with phantom braking or overly sensitive in tight parking conditions, which once again shouldn't really be a safety problem as that should be low speed. And even then those are really only a few older models, anything made in the last several years seems pretty solid to me from the several cars I've driven with it.

No it would not have had an effect. And I was younger and learned from those minor prangs. These fancy computerized features will only make new drivers less capable and dull the senses of older drivers. They become a crutch and may cause dangerous situations if these inept people driving vehicles without AEB and other features.

Just hearing the stories the people around me tell me to do with the computerized features like AEB and other stuff makes me keep my car purchases firmly in the 1990s, which coincidentally is one of the best decades for cars.

Like the woz says "never trust a computer you can't throw out a window"

I bet before your first accident you probably thought you were pretty good at driving and "my feet work and so do my eyes." Then I imagine after your first collision you thought you learned from your mistakes and "my feet work and so do my eyes." Then after your second collision you thought you learned from your mistakes and "my feet work and so do my eyes." Then after your third collision...

Just no.

>It would be different if it were the same exact group of people and the same budget who are in charge of plane safety, as those in charge of the car safety

This is exactly the case. In the US we have the Department of Transportation that governs automobile transportation as well as air travel (and also trains and boats).

But the FAA and the NHTSA are different agencies under that. People who work at FAA can't switch tasks and stop working on plane safety and start working on road safety. With time of course either agency can be funded or defunded but that can be done independently. Just because NHTSA gets more money or resources doesn't automatically mean FAA will get defunded and vice-versa.

It's reasonable to monitor risk at an individual level. Considering one person's fuckup can affect hundreds of people directly, this is reasonable. The two concerns are not mutually exclusive.

I think a better question is why the "car accident" people are always hijacking threads to push their agenda. Like the topic is Air Traffic Control. Why are you bringing up car accidents?

50% of accidents are single vehicle accidents that involve drugs or alcohol. It's more if you add in youth an inexperience. So, _most_ vehicle deaths are self-inflicted.

16% are pedestrians. They usually get hit from behind, and even if they don't, almost always at night.

16% are motorcycles. I ride. I love riding, but the reason for that statistic is obvious. There's zero safety margin when you're on one.

Finally, the average death count from a vehicle accident is like 1.2. The average death count from two planes having an accident is like 600. The reason this all matters is because there are several orders magnitude _more_ car journeys than there are airplane journeys in any given year.

You can't compare them 1:1 and the relative differences matter more than the relative similarities.

Wat?!

You can compare them. Ok if you are being pedantic you need to compare the airline, route, plane model and weather characteristic's with an equivalent car journey route, car model, mechanical condition and driver experience and mental fortitude.

But in general planes are way safer than cars.

If you are 40, perfect vision, sober as a saint in a modern safe car and drive mostly on highways at safe speeds you are still safer in a plane.

    Ok if you are being pedantic you need to compare the airline, route, plane model and weather characteristic's with an equivalent car journey route, car model, mechanical condition and driver experience and mental fortitude.

You can’t compare a plane journey with a car journey; they’re not interchangeable. Most miles for people are just to and from work - planes can’t just replace all those journeys, even if you sum them up to some ridiculously large number.

If that’s wrong, then explain which pilot + aircraft combo you think is equivalent to the “typical” suburban car journey.

You can compare them if you have $1 to spend as a government and want to maximise the number of lives that $1 saves.

> If that’s wrong, then explain which pilot + aircraft combo you think is equivalent to the “typical” suburban car journey.

Walking. We used to be able to walk places before we redesigned everything for cars. Walking is safer than driving.

Lol k I get your angle now. Ngl you’re pretty funny

People aren’t driving drunk and high in a vacuum. You are affected by them even if you are the most careful driver in the world.

I have always wondered about the details behind the annual traffic fatalities stat and this is very clarifying. Thanks for sharing.

Such a ridiculous comment

> The average death count from two planes having an accident is like 600

Do you realize, that your "average" death count has literally never happened in aircraft accidents?

> The average death count from two planes having an accident is like 600.

Its...not. If I had to guess averages, the median is probably closer to 3, the mean probably under 10.

Most planes aren't commercial passenger airliners, and that's even more true of most planes involved in accidents.

https://www.youtube.com/@VASAviation posts ATC audio and a visualization a day after such incidents happen. Fantastic channel, if a bit too addictive.

I just discovered this channel a few days ago and I’ve been hooked to it since then. Another channel I’ve been enjoying is https://youtube.com/@YouCanSeeATC

Interestingly, in 2022 the FCC changed the range of acceptable EKG requirements for heart health required for pilots; previously the PR had to be between 0.12 to 0.2, and they widened it to 0.12 to 0.3. That would have allowed pilots with mild cardio injury (who are more likely to suddenly pass out) to fly.

There is a lowering of standards in many areas happening these days. I don't have an explanation, there are many out there and some are too political to resist even mentioning it on HN.

I’m not sure where this has come from.

First degree heart block (ie a PR interval > 0.2) is not a marker of mild ‘cardio’ (NB: it’s cardiac) injury. It does not cause syncope. It is a reflection of electrical timing between the cardiac chambers. It is not an echocardiogram or a RHC and will tell you nothing about the anatomy or physiology of a person’s heart other than timing. It is not a full electrophysiological study either.

There is no such entity as ‘mild cardio injury’ and I suggest you stop spreading nonsense. Pilots already have strict medical guidance for health and unexplained syncope is not something that is hand-waved away.

>There is no such entity as ‘mild cardio injury’ and I suggest you stop spreading nonsense. Pilots already have strict medical guidance for health and unexplained syncope is not something that is hand-waved away.

A PR interval greater than 0.2 is considered "First-Degree Atrioventricular Block" (https://www.sciencedirect.com/topics/medicine-and-dentistry/...), which "occurs as a consequence of disease in the AV node and is common in older patients." How does something indicating disease in the atrioventricular node not constitute mild cardiac injury? According to that page: "Recent data from the Framingham cohort suggest that in the long term, patients with first-degree AV block have a significantly higher risk of requiring pacemaker implantation and have a higher mortality rate." Would not this increased risk of severe events also constitute an increased risk of mild events like temporarily fainting?

>Pilots already have strict medical guidance for health and unexplained syncope is not something that is hand-waved away.

Those strict guidelines were made less strict in 2022 by increasing the cutoff for PR interval, and the following year we're seeing an increase in flight incidents.

That's a symptom effectively, AV block is just ascribing a name to that description, it's not a disease or injury itself, though yes there may be some underlying cause of it.

It's not like they accidentally started permitting it, because if the rule was 'now pilots with first degree AV block are good to go' is the same thing as saying 'now pilots with PR interval >0.2s ...'.

Presumably they determined it's not a significant indicator of anything sufficiently bad or imminent in the 0.2-0.3 range.

Air traffic control seems like something that should be at the top level surveyed by humans but has many layers of automated computational safety built in. I imagine there is, but the idea that the safety of a plane is up to a single person at some point is unsettling. Having existed for many decades, the entire field should be far past the point of there being even a 0.000000001% chance of hazard.

ATC makes extensive use of automation for scheduling, queuing, etc. The hard manual parts are what remain: communicating with pilots over unreliable channels, handling unpredictable scenarios and emergencies, and maintaining spatial awareness for the airport and the tower's area of control.

Your expectations don’t seem to match any other field of human endeavor I can think of. Nuclear plant safety doesn’t; rocket launches, no; pharmaceutical trials, nope. Perhaps we’ll be better and smarter in the future, but…I doubt it.

When was the last time Nuclear plants in the US significantly endangered human lives. 1979?

With regards to rocket launches, pharma, etc. those are all somewhat novel domains: experiments where risk is a natural part of being the first time trying a set of variables. The variables in air traffic have been the same for decades, thus the thresholds for accepted risk should be much lower.

Call me a techno-pessimist, but we humans have a way of building imperfect machines regardless of theoretical perfectibility. Boeing 737 max?

In fairness, perfect and simple almost never go hand in hand in life critical applications because simple is almost entirely incapable of accounting for and mitigating impacts from the entropy of the universe.

Couple that with capitalism (the 737 max's design, oversimplified, resulted in two disasters due to a for-profit company trying to maximize profit and minimize time to market by hacking an existing platform to compete with a new and ostensibly better one) and it's a surprise we're doing well with air travel at all.

That air travel is as good as it is is a great case for techno optimism. In spite of all the garbage we get ourselves into, we manage to make it work.

The US literally stopped issuing new reactor permits in 1979 (down from on-average about 12 per year), and didn't begin issuing new ones until 2012. Out of 177 reactors issued construction permits up through 1979, at most 112 were ever online at once, in 1991, up from 69 in 1979 when the Three Mile Island failure happened - implying over half of already permitted construction was abandoned or never even started. The facilities under construction were subject to an exponentially increasing gauntlet of new safety assessments, failsafe systems, and in many cases reconstruction of things already built and approved just years prior; eventually, the expense of such facilities radically outstripped the expected ROI of nuclear facilities altogether. Most facilities overran on costs, and took decades longer than expected to see their first real returns, if they even stayed open long enough to see real returns.

If US nuclear plants haven't significantly endangered human lives since 1979, it's because the thresholds for accepted risk became so low as to render new enterprise impossible.

Regulating air traffic control (and thus air traffic) into impossibility isn't a realistic option. Unlike nuclear energy, which has functionally equivalent alternatives, there is no functional equivalent to the speed, reach, and cost of air travel. We likely already hit the practical floor on incidents in ATC a long time ago, thanks to (as you have observed) the variables involved being very stable for decades.

Hence the cause for alarm: what if, all of a sudden, those reliable variables are changing?

What variables are changing in air travel that haven't before?

The US Navy presumably has a near unlimited budget and aircraft carriers are still very analog in this regard. There have been some attempts to make the ouija board digital, but still have it be manually updated.

https://coffeeordie.com/ouija-boards-aircraft-carriers

It’s possible this is statistically expected. In the same way the stock market randomly fluctuates, it would be implausible that the ATC incident rate should only ever go down.

The question is whether this is outside the variance we should expect. Is it?

I see a lot of people coming up with reasons and explanations, implying that the incident rate can be controlled. To an extent it can. But it can’t be decreased forever, especially as the fleet increases. Is there a graph of the incident rate with respect to fleet size over the last few decades?

True, and a 68% increase in something that rarely happens isn’t as scary when you look at the actual total number of incidents.

I wonder how that compares to eg. france, where air traffic controllers strike every two weeks (a bit exaggerated, but they really strike a lot compared to other controllers in other countries and other workers in france).

Also been an 18% increase in road accidents.

Just post pandemic recklessness?

Well that was worrying to read since I flight through San Diego coming up.

Why are the budgets low and why is our government running atc like a best buy durring holiday season?

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

Policy is to run government entities like a corporation.

- ATC is a difficult job both in terms of training required and day-to-day work

- There are high health requirements to hold the job, which should be revisited

- The federally mandated retirement age is relatively low.

- Pay is not attractive

- Overall FAA funding is largely set by the finicky Congress, which swings between extreme austerity and moderate fiscal pragmatism

- Airports can supplement federal funding, but it's difficult to get someone that cares. The right port authorities would need to be elected, then raise rates at risk of airlines pushing back... only to improve the safety of their one airport (likely still taking the full PR/trust-hit if a tragedy happens at a different airport anyway)

- COVID19 led to a higher rate of people exiting the field (due to reduced labor demand + natural retirement), and few entering the field.

- Hiring/pay was not managed in alignment with the return of travel demand.

- These ATC articles have been circulating for a couple years now. That's probably dampened interest in these jobs.

> Pay is not attractive

That's not my impression at all, based on discussions with FAA employees.

From a quick online search:

> ATC 2022 median pay: $132,250 per year

https://www.bls.gov/ooh/transportation-and-material-moving/a...

Is $132,250 attractive relative to the constant high stress and everything else involved with the job?

Did you read the part about 10 hour days and 6 day workweeks? I wouldn't do that for $132k.

Add in the fact that the FAA is about 50 years behind the times in understanding mental health, so you might lose your entire career if you are diagnosed with depression (even if it is a result of stress at work).

Plus, you have to be younger than 31 to start an ATC career, and are required to retire no later than 56. So for good, not great, pay: you have to pick your career before the age of 31. Pass rigorous training. Have a plan for starting a new career if you ever have health issues, or when you turn 56, whichever comes sooner.

In a vacuum, the pay is attractive. In context, it doesn't seem much more than "above average", at best, for a VERY hard job.

In any case, the proof is in the pudding. If people thought the pay was generous and worth it, would there be a shortage?

$132k is great pay. You get to retire at 56 with a nice federal retirement package, plus an extra supplement to top things up until you are eligible for social security (which you're not initially, since you get to retire so damn early). Is it a hard job? Certainly. But I'm sure a significant number of retail jobs are nearly as stressful. And the union has your back, making sure you get a decent number of decently-sized breaks, overtime pay for overtime work, etc.

For the responsibility they’re undertaking, I would be happy for them to get paid even more and retire at 45.

> But I'm sure a significant number of retail jobs are nearly as stressful.

Ok you lost me there. Which part of retail work has you responsible for the lives of 200 people at a time?

> a significant number of retail jobs are nearly as stressful

A significant number of retail jobs involve the risk of a mass casualty event if you make a mistake and the opportunity for such mistakes happens many times through a day, each day for years on end?

You neglected to respond to the important question:

> In any case, the proof is in the pudding. If people thought the pay was generous and worth it, would there be a shortage?

The rest of your comment is irrelevant. The “great”-ness of a price is evident by the number of sellers willing to accept it.

Talking to FAA controllers would dampen interest in these jobs. If everywhere is understaffed, the current staff has to overcome that. Rotating shifts during the week is pretty hard on your body as well in an already stressful job. But supposedly they know their schedule a year in advance.

Is overtime mandatory or voluntary? https://www.jetwhine.com/2008/02/atc-overtime-faa-controller... Looks like a mix of both. Answer your phone, it’s mandatory.

I’ve talked to a few ATC people. One quit his job due to the stress. “The money was not worth it”. They were getting panic attacks and drinking to sleep as mentioned in the article.

There is also a pilot shortage and it takes a long time to get to being paid well after over $100k in training fees.

So many people quit the industry during covid, you can't ramp that back up in even 5 years without major compromise.

Do pilots have any tools detect conflicting runway use?

Like, I would imagine light patterns would mark a runway as take-off, landing, or neither (plane crossing, etc). A pilot told to land on a runway not lit for landing would know something's wrong, and planes would know not to line up on a runway lit for landing, etc.

Interesting idea but no, the lights never change based on usage like this.

> I would imagine light patterns would mark a runway as take-off, landing, or neither (plane crossing, etc)

It's often both, and the direction depends on the wind. It's better for planes to take off into the wind, so as the weather changes runway usage also changes accordingly during the day.

Or were you thinking the lights would change minute by minute based on the next designated usage. That might work! "Now it's a landing runway so it's blinking blue" kind of a thing. So nobody tries to take off or cross it. Or, then when someone is supposed to take off it's blinking yellow so planes landing on it won't be confused.

That's what I mean. Change the lights minute by minute.

There have been incidents where one plane is cleared to cross while another is landing on final approach. There's no fault on the pilots I think. With the right signaling on the runways all pilots involved could double check the clearances and one of them can speak up if things are weird.

You mean like this: https://skybrary.aero/articles/runway-status-lights-rwsl ? Or https://skybrary.aero/articles/runway-awareness-and-advisory... ?

Over the summer a near miss happened in Austin and I looked into this a bit. It seems that some airports do have lights that change in real-time, but not all. Austin does not have the system.

I was an airtraffic controller back in the day. I have very high spatial IQ, proven by the airforce. I went into IT because that was not a sustainable career choice.

The threshold to become a controller is so high. I'm in my 30s now and I literally dont qualify to become a controller anymore. When you starve your potential labour pool so greatly, it's going to leave a workload that's impossible. Thus this article can be rewritten every single year and it's going to trend in the wrong direction.

Pushing Tin is a pretty good film about ATC

The YouTube channels for ATC incidents have certainly been rather busy too.

This is bad news. Most of these people are pushed to the brink normally because of the executive function loading of the task.

https://www.npr.org/2021/08/05/1025018833/looking-back-on-wh...

I've been warning of software deterioration for the past few years and this seems to be an extension of the same effect. Likely the result of increasingly non-meritocratic employee selection processes and non-meritocratic incentive structures. Due to government interference in the economy (especially due to money printing), we've basically turned our capitalist system into a very poorly implemented communist system. People are broke, have low self esteem, can't afford to own a house and only anticipate further deterioration. Heck, average people can't even find a romantic partner because rich people are monopolizing the supply by dating 10 people at a time. I'm surprised things are not worse than they are. I guess it will take another decade to feel the full effects of the problems we're creating today.

You just have depression. Healthy people don't go around collecting sad half-truths.

I don't know any rich people with 10 girlfriends though. Rich people have just as much time as anyone else, likely even less since they usually work more, so that sounds hard to schedule.

As if anyone would tell you if they had 10 girlfriends on Tinder. The number is just a figure of speech. The effect is real.

Also, I'm probably happier than you will ever be because I live in the truth and am surrounded by honest people who have 0 incentive to pretend or say anything that they don't fully believe.

Your truth is that people are broke[0] because there's too much money[1] and that you don't have a girlfriend because rich people on Tinder stole the women?

[0] they aren't https://fred.stlouisfed.org/series/DSPIC96

[1] even if that made sense, it's going down again https://fred.stlouisfed.org/series/M2SL

Yes. More money but with much more inequality which is manufactured by money printing. People who aren't socially connected to money printers are starved of funding in spite of the abundance of money in the pockets of a shrinking number of elite.

And as much as you'd like the second sentence to be true. I'm happily married to an attractive Russian woman. The western brainwashing didn't reach Russia thankfully.

I'm speaking for other guys in the west as I could very well have been one of them if luck hadn't been on my side.

My Etihad flight from North America increased 3 fold compared to pre corona, I am not sure what people in important roles are paid so less with environment so toxic

US air traffic also increased 20-30% compared to 2022, depending on whose numbers you believe (the year isn't over yet).

I wonder whose fault that is, 𝙍𝙤𝙣𝙖𝙡𝙙 𝙍𝙚𝙖𝙜𝙖𝙣.

Automated aircraft for routine cargo and passenger travel controlled by an automated ATC system seems like the future.

How would it handle emergencies like https://news.ycombinator.com/item?id=38586773 ?

I wanted to go into ATC training but am too old at 32. Was pretty disappointed by that.

I guess lessons weren't learned after the lake Constance collision

It's not really the same problem but a good example of the end result (spoiler alert: dead peoples). Lake Constance was mostly due to faulty procedures, technical intervention and bad habits (letting a colleague work alone during night shift coffee breaks) and those lessons have been well learned as Skyguide is now one of the safest ANSP, refusing to increase capacity because of all those safety measures.

FAA is another really scary story. 6 days of 10h shifts with 8 to 9h of rest is just insane. As a euro ATCO I know for a fact that we would never accept that, we have way more days off, a minimum of 11h rest time and are never called to work more nor work more than the contracted hour. As a matter of fact it is the opposite, we usually work less hours a day than what is planned. Maybe it is a different work culture between US and EU but we won't sacrifice ourselves for the company. If we are understaffed we will let it happen, reduce capacity or even close some airspace and let the company lose money and credibility.

Anyway, as we often say, everybody working in the aviation industry is tired because of the shifts. Just don't let the PAX know about it.

Hope trains make a comeback soon

Maybe there is 65% more flights since the Covid times

Usually I like to just read comments, get a feel for where society is. I love the comments in this, it can be summarized as "throw AI at it" and "more tax dollars will fix it".

Lets all forget that we have put the airline industry on government life support through regulation and bailouts, thats turned the biggest companies into tumors that survive off their (literal) monopoly rights. We can all pretend we didnt dehumanize the staff on/off-board with vaccine mandates that simultaneously violated their human rights and failed to stop infection or transmission (as they were promised).

Yep, fat, lazy, unprincipled american voters to the rescue, throw in some rhetoric about how we are going to tax the rich (then tax poor/middle class people who dont have their own LLC), replace the workers woth AI, and incoherently froth about your political opponents.

Or, you know, we could start privatizing it. Calm down, its OK, I was a nuclear reactor operator in the navy, I can assure you it will only ever get worse.

Yeah people doing that kind of job needed a jab. Sorry to say it. Greater good.

Except, the jab didnt prevent infection or transmission, so what difference does it make?

More pertinently, you are referring to the ends justifying the means. If Hitler had succeeded in making a utopian society where everything is just great, would he have been justified in the methods he used to achieve it?

There will always be some ambitious goal, something society could achieve, if we could just get these people on board. You know, they are just in the way, they are holding society back! What follows that logic is never good for society, and look at what its created in ours; there is a contingent of ~60 million americans that never got vaccinated and are royally pissed about those mandates, probably tens of millions more that were effectively coerced into vaccination, and of these groups a disproportionate amount are gun owners.

So, greater good, is it a good thing that something like 50 million gun owners now see their political opponents as a threat to their way of life? Do you think that kind of force could be stopped if things ever turned sour in this country? Are you so certain in the authority and power of your own political alliance that you dont think thats, an issue?

I don't know why, but archive.is is not loading for me

They're in a fight with Cloudflare over technical details, if you use CF as DNS that's why

This is a relevant explanation from the perspective of cloudflare

https://news.ycombinator.com/item?id=19828702

There's plenty broken with both .is and .ph even if Cloudflare is not in the picture.

I have Google's DNS and DoH turned off

It didn't work for me either but the ph one worked

https://archive.ph/EV9WP

I had to disable DNS over HTTPS to get archive to work again.

In Firefox it's enough to add archive.is/.ph/.today to the exception list, and keep using default settings otherwise. Not sure about other browsers.

From what to what? What counts as “serious?” What percentage of flights? Which kind of flights? How many of them are commercial passenger (as opposed to cargo)? Can the variance be ascribed to noise?

Sorry if some of these questions can be answered from the paywalled article. But my gut says that this is making news out of nothing important.

Post-covid ripples through time...

What do we think, within 5 years pilots will be talking to AI chatbots that handle air traffic?

Is there a plane about to crash? Say 1 for not sure, 2 for normal business hours, 3 for no, or 4 for I'M SORRY I didn't catch that, say 1 for not sure...

Of all the things to use an AI chatbot for, this might be the worst one.

That's assuming AI still stay as it is forever. In 3-5 years every job that can be done on a computer will be able to be done better by an AI designed for that job. The only issue will be adoption, not the tech itself.

That's quite an assertion. Why do you think that's true?

All jobs done on a computer are information processing. AI is approaching human level information processing in every benchmark we can throw at it. By the time it eclipses all trained humans, using an AI to do a human job will be a no-brainer, as AI's:

1. Don't require health insurance/mental health leave/PTO

2. Can work 24/7

3. Are infinitely replicatable

4. More interpretable/modifiable than human intelligence

2 years ago the best AI art models were worse than 99% of human artists, now they're better than 99% of them. The same thing will happen to every skill that isn't based on motor/sensory abilities.

What happens once AI is able to do all our work for us? What will humans spend their time doing?

FWIW Obama’s administration instituted a diversity program for ATC that gave points to people that did poorly in school and haven’t had a job recently and penalized people with pilots licenses or military ATC experience and other qualifications. The ATC was considered “too white”.

In essence they worked backwards to figure out how to hit their arbitrary diversity targets without considering the lives of everyone that flies. Yes, they put your life at risk to fulfill an anti-racism agenda. This is criminal.

This would be very interesting to look into if you cited a source.

"Affirmative Action Lands in the Air Traffic Control Tower", Wall Street Journal (June 2, 2015): https://archive.is/lDyOB

>A recently completed six-month investigation by Fox Business Network found that the Federal Aviation Administration has quietly moved away from merit-based hiring criteria in order to increase the number of women and minorities who staff airport control towers. The changes come despite the fact that the FAA’s own internal reports describe the evidence for changing the hiring process as “weak.”

>Until 2013, the FAA gave hiring preference to controller applicants who earned a degree from one of its Collegiate Training Initiative schools and scored high enough on an eight-hour screening test called the Air Traffic Selection and Training exam, or AT-SAT, which measures cognitive skills. The Obama administration, however, determined that the process excluded too many from minority groups.

This is very troubling if it is true, but the reporting felt a bit thin. If there was an article with interviews and more concrete evidence I would be more satisfied.

"Fox Business (officially known as Fox Business Network, or FBN) is an American conservative business news channel and website publication owned by the Fox News Media division of Fox Corporation." ~ https://en.wikipedia.org/wiki/Fox_Business

A few weeks ago I heard about this from another pilot; I found several sources with a minimum of effort. I only fly in Europe, so I have zero personal experience with what is going on in US, but the narative makes sense to me as I saw it personally in other areas.

Here's an academic publication from 2018, but it doesn't seem to make any explicit claims about diversity related reasons for the changes in the hiring process:

>The hiring process for aspiring federal air traffic controllers from approved Air Traffic Collegiate Training Initiative (AT-CTI) institutions has undergone several revisions in recent years. Prior to 2014, graduates from AT-CTI programs were given preferential hiring from the FAA. In 2014, the FAA announced that AT-CTI graduates would equally compete with thousands of people the FAA calls “off the street hires”--anyone can literally walk in off the street without any previous training and apply for a federal air traffic control job. To apply, the FAA requires that a candidate has United States (U.S.) citizenship, a high school diploma, speaks English, and passes the FAA’s new Biographical Questionnaire (BQ).

>Another concerning perspective from AT-CTI administrators is that CTI graduates are at an employment disadvantage with the new hiring initiatives. One of the responding administrators expressed concern that off the street applicants have increased odds of employment over CTI applicants because CTI students are combined in track one with Veteran’s Readjustment Appointment (VRA) applicants putting CTI students second while all off the street applicants are grouped as one and have an equal opportunity for selection.

>The new FAA hiring protocol for federal air traffic controllers that was implemented in February 2014 included several significant changes. In particular, the FAA reduced the role of the CTI-approved program; therefore, the only remaining advantage for CTI graduates is that they are eligible to bypass the Air Traffic Basics Course, which is the first five weeks of qualification training at the FAA Academy in Oklahoma City (FAA, 2018). In addition, the FAA introduced the Biographical Questionnaire which was envisioned to predict controller performance through a process of asking individuals to recall their typical and/or specific behaviors from earlier times in their lives. But due to the lengthy process of hiring and training an air traffic controller which can take several years, it is too soon to conclude whether the FAA’s new hiring policies improved the ability to hire individuals who are more likely to successfully become federal controllers (FAA, 2017b).

https://ojs.library.okstate.edu/osu/index.php/CARI/article/v...

Agreed on this (the DEI part, not the obama part. It's much bigger than that.). Previously it was a mutual understand that ATC wouldn't get snarky with pilots in th air until they are on the ground as pilots have the final say.

Lately on youtube, i've seen nothing but bad behavior from ATC. And let's just say, there were certain qualities that were common between the incidents.

And what were those certain qualities? We aren't mind-readers.

They were black

ATC has always been terse and made you feel like a bother

What, that 90% of them are from contracted towers, whose contractor can get away with understaffing and overworking their controllers, and the controllers know they won't face consequences, because the FAA has no better solution (money, personnel) for staffing said towers and there's no political drive to force the issue?

It's sad that this is getting downvoted. What you posted is a fact and it should be addressed. We shouldn't bury our heads in the sand just because it's an issue of race.

I think if it were a fact, it would be supported by argument and evidence. What we have here is simply a claim, which was not supported by any evidence. I am perfectly willing to accept that poorly thought out diversity policies can have adverse effects on organizations, which is especially troubling if that organization manages vital infrastructure. But I would need to see good evidence, and argument with that evidence, to prove it, and not just a claim standing alone.

So if you don't agree with a claim, instead of asking for evidence you flag and terminate the discussion? This is what happened here.

The very person you are asking did ask for evidence right here in a peer comment 34 minutes before your comment, see:

https://news.ycombinator.com/item?id=38587616

There's still a lack of any hard evidence beyond a speculative Fox News | Tucker Carlson | Sean Hannity dog whistle article.

The primary article of this submission indicates more that overall systemic changes from the top are the root cause rather than being entirely the workers fault; less staff, more hours, reduced training, etc.

1. Lack of response is not lack of proof.

2. People don't leave on HN, demanding an answer and waiting 34 minute before nuking from orbit is illogic.

3. There was a source and some information to support the claim. Also part of the claim is self-explanatory (lowering the standards for new hires to meet some hiring targets).

4. There is ample evidence in so many other places about the effects of lowering standards to meet diverse hiring targets. I cannot publish the results in my own company with close to 100,000 employees because they are not public, but they exist and they are embarrassing at least.

I am not the one that made the claim and I am not interested to support or explain it, but I am disappointed by the reaction: cancel any discussion that does not fit your views. Burn Giordano Bruno at stake, he did not present proof enough and fast enough.

I didn't flag the comment.

https://archive.is/lDyOB

I have no idea whether affirmative action in Air Traffic Control has anything to do with increased incidents. Either way, I think affirmative action as applied here is a travesty.

HN's flagged comments are the most interesting.

Reply on news.ycombinator.com