jcountry

The HUGE problem here is what about CRM?

How many times has the other guys caught a small mistake of yours, how often have you caught his or hers-in just one trip. I bet that number is a lot higher than all these automation advocates would ever imagine.

My main issue with any sort of automation of our jobs deals with improvisation. The best "AI" on earth can't figure out a capcha which any 3rd grader can easily solve.

Situations no one has ever anticipated don't happen often. DC10s don't lose all hydraulics every day. Planes don't land in rivers often, or fly through volcanic ash every day. But when they do, humans solve these problems with novel solutions. Programs would not. Those are called "low probability/high consequence" events. I can think of several more which automation would fail at.

I have become even more opposed to the idea of automation since I started flying the Airbus. I have had to jump in and fix stuff way, way more often on the AB than any of the previous generation planes which didn't "just fly themselves."

Don't trust the "experts" on AI. Those are mostly guys who have books to sell, TV appearances, or contracts for software (they know won't work)

I have been learning technology since the first PC came out which could be plugged into a TV. I remember the "paperless office" predictions from the 70s and 80s. I also know that the dumbass FAA has dumped at least 30 or 40 billion into the idiotic idea of replacing ATC with a neural network..... That idea was the next big thing in the early 90s-and we have just as many controllers today. That job should be much easier to automate than ours.

I'm no professional expert on any of this stuff, just a long term observer. I'm telling you that pilotless airliners won't happen for several generations.