Artificial Intelligence and Aviation

Can AI be trusted and is safe to fly commercial aircraft.

The algorithm flies. What happens next depends on whether the humans who built it, certified it, deployed it, operate alongside it, and trust their lives to it choose to do the difficult work that its safe integration requires.
That work is not glamorous.

It does not make for compelling product launches or impressive demonstration videos. It looks like data validation tables and simulator protocols and liability insurance schedules and regulatory dockets and training records.
It is, in every sense that matters, the most important work in aviation right now.

Aviation has, more than any other industry, earned the right to be taken seriously when it says it takes safety seriously. It earned that right by doing the hard work, over many decades, of converting terrible accidents into the most reliable large-scale transportation system in human history.

The AI transition is asking it to do that work again, in a domain it does not yet fully understand, under commercial pressures that are at least as intense as any it has faced before. Whether it will do that work the second time as well as it did the first is not a question this book can answer. But it is the question on which everything now depends.