Perhaps chess was a bad example, since it's not the human that contains Deep Blue's rulebook. But this excuse is used for more than just difficult AI problems; it's used for justifying moral intuitions, research inscrutability (like I thought I would see in grad school), and professional skills (like medical diagnosis, on which fairly simple expert systems beat out real doctors, and which analysis of images actually has extremely simple algorithms doctors didn't know they were using).

In the majority of cases, it's a simple matter of refusal to do the introspection necessary to identify what it is you're actually thinking, and/or an unwillingness to take the hit to status (either because you lose your mystique or because you open yourself to scrutiny you don't feel you deserve). "You" in the general sense, of course.

To give a particularly apt example, I was a libertarian for a long time before I first encountered rational justifications for laws and taboos regarding sex. And it's because their proponents aren't doing the introspection necessary to find these more general arguments.

The very fact that people found my point on homosexuality insightful, and that Morendil had never encountered the crucial underlying issue, despite (what I find to be) a sincere commitment to understanding it, is proof that a huge class of people (on a major debate, no less), were unjustifiably appealing to this inability to explain. That needs to stop.