The VR vs AI comparison is interesting to me, because I think both technologies have come in “waves”. However, I think this is the last VR wave - it’s going to be on a steady gradient to ubiquity now - whilst I believe AI will winter again and there are many more waves to come, and decades (centuries?) to pass before AGI.

Reasoning being:

VR is just making what we have better. Better screens, better refresh, better batteries, better lenses etc etc. I don’t see any roadblocks.

AGI, by contrast, is not going to be a better DNN. Harder to convince people but thinking is: brain neurons are vastly most sophisticated than digital; we don’t even fully understand what neurons do; we don’t have anything other than a vague understanding of what the brain does; it is apparent that we engage in plenty of symbolic reasoning, which DNN do not do; DNNs are fooled by trivial input changes that indicate they are massively overfitting data; from what I’ve heard from researchers at top AI companies/institutions DNN design is just a matter of hacking and trying stuff until you get that specific results on your given problem, so I don’t see where DL research is actually headed; improvements are correlated with compute power increases, indicating no qualitative gains in the study of learning.

I’m incredibly impressed by DL’s achievements but I believe at best current methods could serve as data preprocessing for a future AGI.

I’m actually quite glad that AGI is so far off, because I don’t think that it’s likely big tech companies will use it responsibly.

VR OTOH is very close and is going to change everything (and IMO is likely a necessary step towards AGI).