Unsurprisingly, it turns out algorithms are not politically correct
The relationship between generative algorithms and political correctness is extremely complex. First, because obviously, we are talking about two very different fields: a generative algorithm is a complex mathematical construct that operates on the basis of statistical correlations, creating concepts from phrases and words, giving rise to vectors that relate them, and operating by means of inferences in that generated probabilistic space. Political correctness, on the other hand, is a complex, arbitrary human construct, describing language, policies or measures designed to avoid offense to, or disadvantages for, certain members of particular groups in society.
Not much of a connection there; even less so, bearing in mind that political correctness is a relatively recent phenomenon that has been progressively gaining importance since the 1980s, and even then, only in some countries. When we feed a generative algorithm from information available on the web, it is perfectly possible that it receives information that many would consider politically incorrect, or, since political correctness is a mechanism that is generated voluntarily, it deduces correlations or vectors that ignore it completely.
On the other hand, the absence of political correctness puts very strong pressure on companies. Ignoring it or…
