I am not a pilot but I have an MS in Aerospace Engineering and an MS in Aeronautical Engineering. I have worked for over 15 years in the aerospace industry. My main occupation has been control law design for satellites - this is different from flight control SW for atmospheric flight but there are many common issues. Today I work in a company that measures the complexity of systems (air traffic systems, power plants, IT systems and SW).
Modern airliners - especially those that fly-by-wire - are full of SW (4+ million lines of code is a commonly cited figure). In practice, pilots "train to fly the SW", not the aircraft. I agree, it is a strong statement to make but when I see pilots posting in this thread that the AoA is not something you necessarily want to be displayed in a cockpit it kind of confirms my point.
But the point is this. I work in measuring the complexity of (SW) systems and I can state that when you have 4+ million lines of code:
1. There is a huge amount of circumstances (combinations of operating conditions) are never tested.
2. To test a SW system of that size you need another SW system that is at least as large as the one you're testing.
3. Because high complexity implies the capacity to deliver surprising behavior, SW systems of that size are almost bound to do so.
4. There are some misconceptions when it comes to systems of systems:
- if you have 100 components then you can get at the most 100 headaches. In actual fact, the number is orders of magnitude larger.
- the worst condition for a given system is NOT that which corresponds to all variables operating each at its operating limits. Sometimes, combinations of values well within the design bounds correspond to the worst-case scenario
- if you have 100 great components a system made of these components is also great.
5. High complexity implies high fragility. If we continue to manufacture more complex SW and more "intelligent" aircraft, these will cause an increasing number of accidents.
My opinion, based simply on my own professional experience and knowledge, is that we are in the hands of computers and that the trend will be to go in that direction even more. I used to work for a computer HW company in the late 1990s. We had a problem with our operating system on one of our models (which was already on the market). We put in one meeting room all our experts on OUR operating system. Their knowledge, when added up, was estimated to cover about 99% of OUR own product! Now 1% of a complex product/system is still something terribly huge. That was in the 1990s, now things are even worse!