Computing power is expanding exponentially as predicted by Moore's Law (which is really Moore's observation, but I quibble), and while there are signs it may start slowing it's still clear that the growth of computing power will continue to outpace "human intelligence" by a huge margin. The shortcoming is programing - it's already incredibly difficult (and expensive) to design and certify flight critical software (Design Assurance Level A - DAL A in the lingo). But even today, most of the coding is done by computer - someone draws a flow chart and that's turned directly into 'machine language' by computer software.
Given sufficient information, a computer can evaluate hundreds or even thousands of possible actions, and determine which one has the best probability of a successful outcome. The weakness is a computer can quickly get stuck when it has 'insufficient information' - something humans are somewhat better at (but still far from perfect). But electronics and the associated sensors are improving so fast that before long, there will simply be more information available than a human can ever hope to process - only a computer would be able to make sense of it.
I was in the industry for 40 years. I listened to people who swore they'd never get on an aircraft with FADEC engines. Same thing with FBW, glass flight decks, and less than 3 engines for long overwater flights. Yet all these things became commonplace during my career. Forty years ago, the idea of fully autonomous cars was wild science fiction - yet it's predicted that within 10 years we'll have exactly that. In fact, I foresee a future where automotive 'human drivers' will be discouraged if not outright banned - and it may happen far sooner than most of us would like...
Autonomous commercial aircraft will come. Aviation is understandably slow to adopt unproven technology (the FAA is on-record as stating 'artificial intelligence' is banned from flight critical software). But when CFIT and other forms of pilot error (and sadly, pilot suicide) make flying more dangerous than driving to the airport into your fully autonomous car - the industry will have no option but adapt.
BTW, many cargo operators are buying brand new, modern freighter aircraft. They'd be delighted if they didn't need pilots...