I'm a programmer, not a psychologist. I have, however, some slight familiarity with IATs as a purported measure of unconscious bias; they have never impressed me as being particularly reliable as such, given what seems to be their questionable repeatability, not to mention the ease with which they are manipulated, and the way most of the results so obtained tend to hover just outside the margin of error. I won't even talk about the tendentious nature of the investigations themselves, because experience suggests there's no point in so doing.

You disappoint me, sir. I had such hopes of finding something new and interesting, only to discover that your mere vagueness led me astray.