/sci/ - Science & Math

>>9439026


If you were to poll people what restaurants they had eaten at in the past month, you would have data.

If you were to poll them on what their favorite restaurant is, you would certainly have a very different sort of data, that would require much different interpretations, and be applied to different uses, but still something of relatively similar kind.

If you were to ask people what they think the *best* restaurant is though, now were talking something categorically different.

As a general rule, the more people you involve in a decision making process (not in the broad sense of 'everyone on a team', as in *executive* decision making), the more adulterated the output will be. Or to put a spin on the classic riff: "none of us are as stupid as all of us".

A being can be better or worse at grasping Being. Usually worse. Evolution is not always by necessity a process of optimization after all, and the invention of civilization itself comprehensively changes the selection filters. Adding more onions to the mix so often means you're just adding more noise to a signal.

Which isn't even the biggest issue though, which is that *the mechanism itself*, of decision by consensus, *incentivises* hacking the process with less than felicitous signals for fun and profit.