The future is improbable

In 2008, financial models predicted a housing market collapse. Traders, trusting these models, began selling their positions. Other traders, seeing the sell-off, followed suit. Within weeks, the prediction had fulfilled itself - not because the models saw the future, but because enough people believed them and acted accordingly.

We tell ourselves the first traders to sell had foresight or "saw it coming." However, maybe a truer interpretation is that they and their colleagues collectively made decisions that resulted in the very collapse they'd "predicted."

This pattern repeats everywhere predictions and human behavior meet. Poll numbers shift elections. Analyst reports move stock prices. Trend forecasts shape consumer demand. We don't predict the future, we negotiate it into existence.

The trouble is that we treat probability as though it applies to the future the same way it applies to fixed states. Consider a coin flip. A coin lands on a table, covered by the flipper's hand. Heads or tails? Your odds of guessing correctly are 50/50. The coin already exists in one of two states, but you simply don't know which.

Now imagine being asked the same question before the flip. Intuitively, it feels like the same problem. But it isn't. In the space between now and the coin landing, an infinite number of futures exist. The flip might never happen. The coin might land on its edge. A gust of wind might carry it away. The sun might explode. Most of these possibilities are absurdly unlikely, but that's the point—probability requires a closed set of outcomes, and the future refuses to stay closed.

This doesn't mean probabilistic models are useless. They're extraordinarily useful for decision-making within bounded scenarios. But we overestimate their accuracy when we forget what they actually measure. They calculate likelihoods within the eventualities we've imagined, not the eventualities that exist.

The future can only be dreamed of, and then made. Any prediction later deemed true is coincidence dressing up as foresight.

The principle worth extracting here is to take seriously one's own role in shaping what comes next. There is no predetermined outcome to be predicted successfully or not, only the results of the decisions people make; yours included.