I think often the appeal of prediction markets is something like “legibility” rather than accuracy and usefulness, and people tend to discount established sources of information that are less legible in that sense.
For example, reading an article by a journalist who spoke to a set of known experts and synthesized their conclusions in an article, indicating there is broad consensus that X is likely given certain circumstances, probably gives you a better understanding of what is likely to happen than a prediction market saying 83% likely. But having a headline percentage number of a defined outcome feels more precise, and it is easier to quantify whether it was correct or not in the future.
Legibility is obviously a positive thing in general, and ideally you’d have both. But there’s a risk of a streetlamp effect causing people to overly focus on things that can be measured in fields with high uncertainty.
I think often the appeal of prediction markets is something like “legibility” rather than accuracy and usefulness, and people tend to discount established sources of information that are less legible in that sense.
For example, reading an article by a journalist who spoke to a set of known experts and synthesized their conclusions in an article, indicating there is broad consensus that X is likely given certain circumstances, probably gives you a better understanding of what is likely to happen than a prediction market saying 83% likely. But having a headline percentage number of a defined outcome feels more precise, and it is easier to quantify whether it was correct or not in the future.
Legibility is obviously a positive thing in general, and ideally you’d have both. But there’s a risk of a streetlamp effect causing people to overly focus on things that can be measured in fields with high uncertainty.