Let us work with a thought experiment and I'll try to explain the bayes way of thinking as we go along. I have a dice whose number of faces is a question for you to figure out. What I will tell you is this - when I rolled the dice 9 times the following numbers showed up -
3, 4, 2, 5, 4, 6, 4, 7, 4
Alluding to the above quote referring to evolution, I can argue that I used a 373-faced dice and there's no way to prove me wrong. But it just. doesn't. feel. right. that I would have used that particular dice. How do you prove me wrong then? What is the number of faces of the dice that I have used? Can we make a reasonable guess atleast?
Take your time to solve the problem.
On a side note, isn't this how the world works? All we have at hand is the data and not the process which got us the data. It's upto us to find out which process was likely responsible.
'Common sense' dictates that the dice must have at least 7 faces or else we would never have seen a seven in our data. Techincally speaking we've done a bayesian update on our hypotheses space. But we are jumping to incomplete conclusions. Let's take a step back and solve the problem the right way, the bayesian way.