This article took a weird turn midway. I really enjoyed the beginning and the start.

My takeaway of the middle part is that the author had certain expectations about self hosting that weren't realistic and they jump to certain conclusions I would not have jumped to.

Yes it's true that we here are an extremely privileged bunch but over the decades I've seen tech become widely more accessible even before. People were paid a bunch of money just to write HTML in 1997 and in 2025 anyone can write a production ready website using the SOTA LLMs.

I am convinced the future is self-hosted (which is why I clicked on this link!) and as the cost of providing a service becomes less subsidized and starts to reflect the true cost to service a user, more and more people are going to self-host. From that lens, majority of the public not self-hosting is a cost-benefit, not a knowledge issue. Ergo, if tomorrow TikTok charged every person $500/mo, most people would figure out a way to run their own federated TikTok (but most likely just start using a similar but cheaper service).

Consider, there will be a future when Google will struggle to make money (gasp! how can that be?!) and many years before that day comes, all the "free" gmail accounts will be gone - because by then, email will cease to serve as a method for a company to mine novel information, I imagine there will be no "free" GMail although "free" email might still exist then just because the cost to provide 10GB email would be insignificant and worth good PR to someone (but at that point, not Google).