I think it's disappointment fatigue.
Pareto's principle is alive and well, the first 20% of developing a technology is the easiest and most productive, but the actual goal is deep in the weeds of the non-trivial 80% of the work that must be done.
Thanks to capitalism, the work will stall alongside the profits. We'll get 30-50% of the way there, and hit the wall, that being that if we want general purpose AI we will have to pay researchers and developers more and more money for more and more diminishing returns.
Eventually we will hit the point where even though we are 80% of the way there, the bean counters won't justify the expense. "It's good enough, just stop here" they'll say, and then all of the research will be shelved for 5-7 years, then open sourced, and then muddled with for 10 years or so and then the cycle will repeat itself all over again until the right genius with the right background in the right time and right location finally codes the right bit of code or uncovers a new branch that we've overlooked and this all blows up again.