The problem with these sort of predictions is that they never include falsifiable criteria, which would render the entire analysis false depending on the outcom

The problem with these sort of predictions is that they never include falsifiable criteria, which would render the entire analysis false depending on the outcome. Here's a sample quote: "It's not going to die overnight, Patreon will still be here in 2020, it'll just be, you know, worse to actually use." Of course it will be "worse", users tend to complain about every single product change that ever happened. No objective criteria of the worsening are provided. This is the author backtracking: "oh, contrary to what I said in the first tweet, it's not like it would die or anything, but it'll be like "worse", y'know?".