This is extremely salient. Check out Phoenix, AZ sometime in street view. It's a brutalist grid of wide roads (even in "residential zones") where every property is lined in a six-foot block wall. As a result, sight lines are excellent for drivers (encouraging high speeds) but terrible for homeowners. Kids can't reasonably roam free, neighbors rarely meet, and everyone is viewed with suspicion. Most of my neighbors are really decent people, but I see them so rarely we might as live in different cities.

I used to drive a 1996 VW Passat. If I drove reasonably on the highway, I could average about 57 mpg per tank. I thought that was great, so my next car was a 2006 VW Jetta. That car only got 40 mpg at its best, but that was pretty good comparatively, so I accepted it. The newer TDIs were rated even lower, and then mostly went away, so I got a Honda. It only gets 23 mpg. Each of these cars has also been shorter between maintenance cycles, and with a shorter overall expected lifespan. I don't know what's going on with the regulatory landscape that cars keep getting bigger, heavier, and less efficient (by mile), but it seems like there's plenty of room for improvement. The fact that this is proposed by automakers, though, makes me extremely skeptical.

In other words, a world designed for cars instead of people is less friendly to people.

We run into this with drone filming from county and state lands. The governments assert a right to the airspace above their parks, which is almost certainly unenforceable. However, the governments do issue permits, which one can imagine become significantly more limited in scope once an airspace lawsuit is filed. Ergo, the courts can say whatever they want, but in practice it's even more restrictive than that.

The penalty doesn't have to be increased, it just needs to be selectively enforced.

You'd need to have some unenforced penalty first though.

There is no penalty.

> against the value add of AI

Hasn't the surprising lack of value add been discussed with increasing frequency?

The problem is that the people who will put this in place rate capability on a linear scale: in their view the ability to write software is sufficiently magic, so such an ability is obviously good enough to recognize criminals. From their perspective, there are hurdles to be crossed (like probable cause) and an AI flagging a suspect feels like a magical intelligence crossing those hurdles and allowing them to continue in the process.

They don't validate the results of their fellow officers, or the validity of warrants, or anything else that predicates an arrest. Why would they start with this?

I don't think it's money. I think it's requirements and training pipeline restraints. The system is predicated on being able to throw bodies at the problem, but there is a distinct lack of qualified individuals to back that up. Personally, I didn't realize ATC as a possible career path until I was 36-- imagine my surprise when I found that I had already aged out.

The training is also not run particularly well. There's a single facility in Oklahoma that every prospective air traffic controller has to go through. I had a friend in college who graduated in the early 2010s with a four year degree in air traffic control. He waited several years for the FAA to tell him he could start training, a spot never opened up, and he moved on with his life and did something different. It's broken on a pretty fundamental level if we have a shortage of air traffic controllers but also people who want to do it can't get in.

> but there is a distinct lack of qualified individuals to back that up

Which means either the compensation is insufficient to attract and retain the necessary number of qualified individuals, or the FAA lacks the resources to train an appropriate number of qualified individuals. Either way, it's about money.

Who would want to work that job once they find out what the day-to-day is like? I had an intern who looked at that out of the Air Force but he found out what you get paid and what the expectations are for the job and he figured he'd try his luck on something easier and better-paying like life-preserving medical devices. On a related note, why do you think nobody who you'd actually want teaching public school actually teaches public school in the US?

I know this is a throwaway comment, but I can't let it pass.

> why do you think nobody who you'd actually want teaching public school actually teaches public school in the US?

We're currently doing school visits for our kid, in a low-performing school district, and the teachers and administrators we've met have been impressive. I've worked in education, and visited a lot of schools in another professional capacity, so I know the questions to ask, and things to look for. I have no illusions about there being absolutely terrible teachers out there (and I'll tell you some horror stories, if you'd like), and doubtless any (hypothetical) bad teachers at those schools are being kept away from prospective parents, but your statement is hyperbolic in the extreme. The problems in the US school system are legion, but "every single teacher is crap" is not remotely true.

I dunno.

I was at school for 12 years.

There were two good teachers.

The rest of them, and all the staff at those four schools, are hopefully spending the rest of eternity burning in hell.

I get you. I'd say I had three good teachers, one absolutely awful one (that I would likewise condemn to hell), and the rest... meh. I hated their classes at the time, but with an adult perspective I can say that they didn't do me any harm, some people well, and (in at least one case) more (albeit non-academic) good for me than I could have recognized as a kid.

Should we do better? You bet your ass. I have all kinds of ideas....

Nevertheless, both of our experiences put the lie to the GP's hyperbole. Bad as the rest might have been, you had at least two teachers who were exactly whom you'd want to be there.

Maybe it seems like I'm being pedantic, picking on GP's wording, but I'm really not. I'm trying to point that even those of us who had a bad time in education (and, to be clear: I did, too) experienced a few bright spots. It's important, if we're going to engage ourselves with any kind of reform, a) not to shit on the entire teaching profession, b) to consider what made those good teachers good, and c) think about how to support the quality people already in the system, and to attract more like them to it.

I think the comparison to public education is apt: often (at least initially) great people trapped in a terrible system. I suppose you can pay people to ignore a certain amount of misery on top of the job, but I do not believe you can (or should) completely obviate all brokenness in a system at the end of two weeks in a paycheck.

To play devil's advocate, ASEL into small deltas is significantly different than receiving full-stop IFRs late at night.

This small mistake (and it is initially small, just catastrophic) is a system breakdown, not necessarily a staffing breakdown. Though staffing is definitely a wider issue in the NAS.

Edit to add: looking at this incident closer it appears LGA was busy enough to make a single tower/ground controller an obviously bad plan. Still, systemically, there's enough low hanging fruit here, like ADSb in for the airport trucks or hold short line guard lights. I hope the takeaway isn't just "don't have controllers make mistakes".

Yea, if you listen to the ATC audio, you can hear that in addition to the normal high workload of handling both ground and tower, this guy had an emergency aircraft on a taxiway to deal with, too. A lot of holes in the swiss cheese lined up, but one of them clearly is ATC workload.

Perhaps in a scenario where there is an active emergency and one controller, protocol should be that ground and air frequencies are combined.

That would have given the jet a chance to hear the truck cleared to cross the runway they were landing on.

For a long & low trip like that, ATC staffing is a common story. Higher sectors not having the manpower leads to clearance delays, and this is a somewhat common workaround for that somewhat rare situation.