[Retric]

The complete failure of British Intelligence to predict the fall of the USSR shows they might have gotten better at counter espionage, but still failed at their primary job.

The tricky bit isn’t collecting intelligence, the tricky bit is being able to turn raw data into something worth all the effort of collecting it.

[mhh__]

True, but economics predicting 2*n of the last n crashes and Noam Chomsky predicting 17 of the last 0 revolutions indicates that it's an impossibly difficult problem than mere incompetence.

SIS do not publish their archival material so we'll probably never know exactly which side of that argument they deserve to be placed.

[Retric]

Predicting the timing of a crash is basically impossible, but predicting instability is more straightforward. To use a recent example the US stock market basically ignored COVID until mid February, but in January people where talking about an economic risks of a possible pandemic and the signs just kept getting worse.