I actually just posted this
link on another article thread in the career builders section. A little long winded but I think it highlights some of the challenges that would have to be overcome by designers.
No one knows the future, but we do know there are some "significant" hurdles to allowing automation to the level many in the media so confidently throw around.
Many of the strongest proponents, simply don't understand the nature of the profession or underestimate the demands put on existing pilots. I also question how much they understand about the capabilities of AI in reference to those demands.
This misperception may be a by-product of how we talk about the profession. Most pilots avoid going into the harsher realities of flying with the flying public as a professional courtesy but are we actually doing a disservice by causing others to underestimate the very real and present threats associated with flight and allowing those who overestimate the capabilities of AI to oversell its ability?
When pilot error is the cause of an accident, it is quickly measured and broadcast to a large audience. Is there a similar public tracking of automation/system/human factors/wx/sensor/etc errors that are corrected by pilot action, threat mitigation or in-flight management? Without that level of information, it's tough to know how to measure an AI's capability in comparison to pilots.
I worry the industry's temptation to reduce costs (avoiding training and wages to develop trained and skilled pilots) by considering higher levels of automated flight may be a risk to implementing technology that has not been fully tested or vetted. Investing in people/pilots and their unique capabilities has almost always resulted in better results than a "go it alone" approach to automation.
It also signals a much better future to the many thousands of aspiring aviators who are considering such a large investment....