Since you didn't really experience, I'll tell you that Email, chatrooms, Ebay, Napster, being able to check stocks online, stupid stuff like that was amazing at

Since you didn't really experience, I'll tell you that Email, chatrooms, Ebay, Napster, being able to check stocks online, stupid stuff like that was amazing at the time actually...and a lot of the business models which would later go on to make money were being talked about at the time...but the technology had not caught up to the economic feasibility. You couldn't have a Youtube in 1999 because there wasn't enough bandwidth. Now what I think could be happening, using the 90s vs today heuristic you're suggesting...is that there is a massive amount of over-enthusiasm in AI where people think it's basically a magic box that you put stuff in, and you get Lt Commander Data on the other side...and there is no understanding or differentiation between what the different forms of machine learning, and what is technologically feasible, how hard it is to build, etc, that is going to cause a bubble because people don't want to miss out on AI, even though they don't understand what it is. That being said, there are many different forms of web and data tech...mobile, cloud, IoT, etc that you can invest and diversify in now...whereas the late 90s was a lot more monolithic - PC's and websites.