That incident was definitely ahead of its time, but before that, targeted ads for scams were propping the industry up. For instance, Experian used to spend $$$ to funnel people into credit card scams -- one year, they stole our CC#, and I called our top 5 US bank to dispute the charges. The bank fraud department rep said Experian drove most of their case load, to the point where they had a special flow to block just Experian affiliates because physically reissuing stolen cards was too expensive. You can back estimate how many predatory ad dollars that involved. Of course, there was also a long tail of smaller companies that wanted to sell supplements, useless other useless services, etc.

Before that, you could pay for placement on all the major search engines except early Google, which basically just broke search. (Google does similar stuff now, in the same way they "don't" sell personal information or read your gmail, but somehow make lots of money off it anyway, and target things only mentioned in private correspondence).

LLMs will be next. The only solution is to refuse to pay for access with your attention. That means no copilot, gemini, alexa or siri, since those are all ad supported or adjacent now. ChatGPT and Claude are on borrowed time. I suspect the open weight models will be the last ones to fall.

Social media was weaponized from the days of PHP forums. I remember Palantir shilling sock puppet management technology at a time where most people didn’t even know what a moderator was.

AFAIK, Russia’s Internet Research Agency was the first organization to weaponize social media and the internet.