In these arguments it's always very notable that not only do people not benchmark LLMs against people, but several I've discussed with have argued very strongly

In these arguments it's always very notable that not only do people not benchmark LLMs against people, but several I've discussed with have argued very strongly for not doing so unless they're benchmarked against above average people. While arguing that these same tests prove LLMs can reason. It never seems to land with them that their standards for "reason" would exclude large portions of the human population to some state of lesser being without the ability to reason.