That was the dot-com bubble: even though the internet was a big deal, pets.com and Webvan were not viable internet companies as they took on something the ecosystem wasn't ready for. They did, however, anticipate, and market to, a future where their business became viable, in the form of today's Amazon. 1990's Amazon was purely a bookseller, and perpetually viewed as sitting on a knife-edge.
Now we have a second hype cycle for VR, and while it's much better stuff this time, that doesn't mean it's actually the future. The hype around AI is similar. There was hype over previous AI techniques, too: expert systems, for example, had a period in the 80's where they were promised to do everything, assuming "everything" meant encoding a huge decision tree. Now it's the neural networks that will do "everything". NN also had a hype cycle in the 90's, but without the data and computational power, they weren't able to achieve the results we're getting today.
Developments in AI today are enumerated in achievements that become banal moments after you first hear of them: Watson beat a Jeopardy! champion. AlphaGo beat a top-level Go player. We can detect cats in images. Cars can self-drive a course better than trained race-car drivers. And so on.
It's good to be excited about the future. The marketing hype is not evil, it's necessary - it creates a dialogue, focusing people to think in terms of new technology. It's not the same as its realization, which tends to be more of a "one morning I woke up and realized I was in the future" self-reflection.