186 - Strategic Incompetence

I was recently reminded of a curious principle, that if you hear yourself thinking "They couldn't have made a worse choice/claim/etc. if they tried...", and this occurs multiple times in quick succession and holds up under scrutiny, the reasonable explanation is that they are indeed trying to select the worst possible option.

In the tech industry and politics, this should sound quite familiar, as heuristics furnish everyone with an abundant succession of worst-possible choices being made. Of course, these heuristics are often unfair, as news doesn't report neutral, positive, or mundane choices with the same frequency as negative choices, particularly the worst-possible options. This obscures which companies and other groups live up to the worst-choice selection consistently but doesn't negate their existence.

What reminded me of this curious principle recently was an email chain that included someone whom I hadn't heard from in 3 years. In rapid succession they wove together a complicated tapestry of cognitive bias that was maximally divorced from reality, leaving me baffled and to conclude that they must be severely delusional. Given a bit of reflection, I realized the principle was there, that it was unreasonable to assume that someone consistently and rapidly hitting the maximally wrong mark wasn't doing so without clear intention.

In the context of such an individual within a group, such dynamics illustrate a more precise opposite of collective intelligence than even "Groupthink", as the former intentionally undermines collective intelligence, while the latter may or may not be intentional in nature. The terms "troll" and "saboteur" seem inadequate descriptors for such a person, so a new term may be necessary.

In the context of corporations and political groups then forms of decision-making structure may be collectively constructed to consistently deliver those maximally wrong decisions. The dynamics change in the sense that the individual previously noted clearly holds malevolent intentions, but the inanimate structures of a corporation or political group hold no parallel capacity, just as trivial AI systems they train hold no "intention".

Intentionality is a tricky thing, as it is usually virtually impossible to prove. It also holds no relevance to ethics, as it is subjective and functionally incalculable. However, where it does hold relevance is in the selection of actions that an individual or group is likely to take, making it a metric relevant to risk management in those handfuls of cases where intentions are made clear, such as that noted above.

While it may remain unreasonable to predict a specific action and time, the risk of a subset of actions may be increased absent naïve forecasting. The maximally "toxic" individual is a risk to be avoided, but the maximally wrong structures that corporations and political parties construct over time are a far more complex and hazardous challenge to overcome, as well as much more difficult to recognize in their interim stages of development.

Prof. Tetlock popularized the term "dart-throwing-chimpanzees" to describe competing with random chance, but when considering intentional aim toward the worst option then the term "feces-throwing-chimpanzees" might be more accurate.