The larger problem is that we don't really know what we should be assessing in the first place in this new world. Most solutions addressed to fairness concerns end up testing for (i) speed and (ii) ability to memorize/regurgitate--precisely those things that A.I. can and will do better in practice.
