Were you not allowed to go to the library during lunch?
The top 0.1% are no longer more competent than they were 10 years ago. Take math. About 1% of high schoolers participate in the American Mathematics Contest, and about 10% of those score >100. Ten years ago, they had a strict "top 5% or those scoring above 100 advance to the next round" rule. Five years ago, the contest got much harder very quickly (the class of 2022 was exceptionally good, and the contest writers probably anticipated that by seeing MATHCOUNTS results), and where the cutoff had been ~105, it dropped down to ~84. Note that I'm talking about the AMC 12 here, and participation had declined at this point, so they took the top 10% instead of the top 5% to keep the same number of contestants in the next round (the AIME). The most recent contests are easier than they were five years ago, but the cutoff has remained in the 80s.
You might as well say, "if everyone could just get along, everyone would be able to get along." Everyone has different preferences, and sometimes they are competing preferences. You can put this all in a big matrix, and compute the eigenvectors to find the cliques. Naturally, if my preference is for you to do well, and your preference is for me to do well, we'll want to collude. Naturally, if everyone preferred for everyone to do well—or at the very least were neutral on others' wellbeing—the eigenvectors would be nonnegative and there would be no reason to collude at the expense of others. But there are negative entries among eigenvectors (e.g. the murderers in prison), so there are groups colluding at the expense of others.
Also, I don't think your social collusion problem is as bad as you might think. If your reputation is negatively affected by association, and you no longer like the association, why have you not broken off from it? People do not punish others for a history of association—especially when done in ignorance—very much, especially if you publicly denounce or expose them.
I haven't been able to find the spin-1/6 anyon's partition function, so mine could be wrong.
A couple things to add that don't deserve to be in the main text:
The Taylor series for the partition function is , which means it actively learns "not this" the second, and third times around. This is why we see a dip () when , followed by a steep rise in loss () as , and then a tapering out.
The and partition functions correspond to bosons (e.g. photons) and fermions (e.g. electrons) in physics. Perhaps corresponds to an exotic particle the theorists have yet to classify.
I've only have it happen with one person... but it happens every time with that particular person. I've mostly stopped debating with them.
What about when they say, "you're strawmanning me!" and slightly change their argument? You believe their argument from two minutes ago was wrong, and that they are now intentionally misleading you so they can maintain their position and eternally shift the burden of disproof back onto you.
Debbie is actually red-blue colorblind, so she thinks her graph looks normal.
Looking back on this comment, what do you think you can change about how you politick to avoid making the same mistakes?