Eh, it depends on what we count as "AI". I'm in a field where machine learning has been a thing for years, and there's been a huge amount of progress in the last couple of years[1]. However, it's exhausting that so much is being rebranded as "AI", because the people holding the purse strings aren't necessarily the same scientists who are sick of the hype.
[1] I didn't get into the more computational side of things until 2021 or so, but if I had to point to a catalyst for this progress, I'd say that the transformer mechanism outlined in the 2017 paper "Attention is all you need", by Google scientists.