> Keep in mind this is rank speculation 20 years into the future. It's a bit amusing that the comment right next to yours (albiet a cousin not a sibling) is in

> Keep in mind this is rank speculation 20 years into the future. It's a bit amusing that the comment right next to yours (albiet a cousin not a sibling) is insisting that what I'm describing is already happening. The (naive) weaponization of this tech is already underway. What's more, the tools and methods to turn the categorization work in this paper into a goal-oriented optimization procedure already exist -- see for example the Diplomacy work from Meta; again, focus on methods, not the application. The tech is here today, the barrier to use is low, and the incentive structure to weaponize exists. What will take 10-20 years is the trickling of that impact into social processes and then the retrospective reflection of what happened. Note, for example, that Facebook was founded in 2004, but it took 10-15 years for the impact on social processes to reach a point of significant material impact on global society and politics. > If this were 2004 OP would have you convinced 90% of your day would be spent culling spam from your inbox. It's not about the amount of time spent; it's about what happens in the time that is spent. Spam, and online attention markets more generally, is an incredible apt analogy. Even for the most strict definition of "spam", the market for prevention is pretty huge. Anti-email-spam tech alone accounts for ~$5B/yr in spend with a 21% CAGR. Going all-in on anti-spam tech in 2004 and capturing even of sliver of the market would make you very wealthy today. If you include spam-like content on social media platforms, that number certainly more than doubles. On headcount alone, Meta told Congress that they have 40K employees directly working on trust and safety, and TikTok says they have about the same number. That's just headcount, and at just two companies. Beyond that, the following question will prove only more prescient as time goes on: where does "spam" stop and "algorithmically curated content designed to part consumers from their dollar" start? Even today, the average person in places like the US will spend ~2.5 hours consuming algorithmically curated feeds on social media (in service to online advertising revenue). And many aspects of non-social media now have some aspect of quantitative optimization as well. Contrary to your take here, with the benefit of hindsight, I think the "Eternal September" doomers of the early naughts significantly under-estimated the impact of information technology on how people spend their time and how society spends its resources. (And, anyways, your strawman is a bit too hyperolic -- no one was seriously arguing that 90% of anyone's day would be spend culling spam from their inbox, least of all me...)