Hacker News Reader

Comment by notabee | Hacker News Reader

Pick one or several:

* Air traffic controller scabs during Reagan years are all retiring.

* Difficulty attracting and training new air traffic controllers, which takes years to do the training and is a high burnout job.

* Widespread effects of a virus that often causes cognitive damage, fatigue, and/or outright disability in many that's been circulating for almost 4 years.

View on HN

I've also heard it suggested that there was a lull in traffic during COVID that might have made people lose their edge.

Not to mention how many of them could be suffering long COVID (perhaps without even recognizing it).

I find it odd that you're basically reiterating my third point and yet you're the one getting downvotes, presumably because you actually used the "C" word. The level of denial about mass disablement (temporary, semi-temporary, or permanent) just gets more and more ridiculous.

Downvotes are sometimes given to short comments that don't add anything extra to the thread. That might be the case here as your original comment was quite clear.

Also seems that this is one of the easiest jobs to automate in the future. Keep planes from running into each other. How hard could that be? If a computer can drag a 9-dan Go champion up and down the board, it can do this.

I'm sure I'll be told differently any minute now, but... if I were a young person looking for a career, ATC is going to be very far down the list. Concerns about future automation will become a self-fulling prophecy once no humans are available to do the job.

In the video of the incident linked elsewhere there is a caption that basically says that an automated system is what notified the controller there was an issue and probably why she was able to correct it at the last second.

Ah yes I can't wait until some LLM instructs a plane to go to a runway that doesn't exist, and the pilot can't ask for clarification because the airport fired all the humans because "we don't need them"

Hmm, that doesn't sound like an optimal implementation. Can you think of any other approaches for handling air traffic (no pun intended) besides using AI language models?

Reply on news.ycombinator.com