Jim Carter wrote:
Autopilots do not take a degree in astrophysics to operate. That being
said, I also made the point that an hour or so of instruction "in the
aircraft" should be sufficient.

Time flies in the air. I don't think I would claim proficiency
in my autopilot (and it's interface to the NAV radios) until I'd
flown a couple of climbs, descents approaches and holds with the thing.
Time flies in the air. I think it was more than an hour. It is
however a tiny fraction of the IFR training time. And this was
despite the fact that I had the thing for a year before trying to
do IFR procedures with it and I still got stumped the first time
I tried to do an approach with it after my checkride (this I think
was the GPS not the autopilot but the result was the same... push
the big red button and hand fly it).

I still don't trust the thing. In VFR conditions I'll let it
motor me around the purple line, but in IFR I'm still hovering
over the big red button and scanning just as carefully as I would
if I was hand flying it.