I think one of the main problems (pointed out in earlier posts) is that single pilot operations for heavies and/or LH/ULH requires a level of safety for the aircraft that needs to assume no pilot at all, or at best, no pilot for significant periods.
I am reminded of a particular phrase when I see gushing projections from academics and aircraft manufactures: Full Self-Driving. It will happen, I’m sure of it, but just like working fusion power, it’s been lots of false starts and a huge amount of work to come. There is also a parallel between aircraft and vehicle automation levels, in that the low-hanging fruit of levels 1 & 2 have been picked off, so level 5 (full) must be just around the corner? Unfortunately, we’re still stuck at level 3, looking at 4 which still needs an element of human intervention, which is unlikely to go well if the last few hours of the journey have been spent on social media because the transport is “reliable" in human estimation. 3 & 4 are unwanted and dangerous, IMO, and we really need to go directly to 5 because the human is already well out of the loop in 3 & 4.
Another issue is certification. What will happen to the DAL-A/B/C stuff which was so heavily regulated in the past but will effectively have to be done by some sort of trained AI? “Yeah, we use GPT-3 to talk to ATC, a genetic algorithm to optimise the route and Stable Diffusion to create a pilot avatar. We’ve done some testing and it seems to work most of the time..."