That was not my experience. I was meeting web chat friends IRL in 1995 and none of us were the least bit afraid. IRC and the web were everything SnapChat is today and not considered any stranger than AOL IM from 5 years before that which were hardly different from the BBSes that came before the subscription services.

So, lots of people were meeting online in 1995. 5 years later couples who met on the internet were all over the place. When I moved to Silicon Valley in 2000, "I met my wife online" was hardly even a conversation starter. (I did meet my wife at The State of Insanity web chat boards in 1995.)

Also, from 1998-2008, the world invested precisely what it's now invested in the last decade of LLM AIs and was already adding over a trillion dollars annually to the global economy. AI, with that same spending over about the same amount of time, is still an economic black hole and no one has any answers that would satisfy anyone other than VC about how it's all going to pay for itself.

The internet, and particularly email, IM and the web, were going strong in the 90s and by the mid-2000s were absolutely dominant. This re-writing of history to suggest that AI deserves more runway to burn more hundreds of billions on things that will have no staying power like the trillion dollars in internet build out between '98-'08.

These things don't take time. The whole broadband internet took about a decade and about as much money as AI has gotten in this latest bubble. At least when the dot com bubble burst, we had infrastructure with staying power and real societal value, unlike AI which will leave us with next to nothing when its bubble pops.