What evidence exists that we (humans) can make accurate predictions for the states of complex systems decades or centuries in the future? It seems like "things

What evidence exists that we (humans) can make accurate predictions for the states of complex systems decades or centuries in the future? It seems like "things will be different" and "there will be catastrophes" are always solid bets, but modeling complex systems -- especially systems with many unknown variables -- always involves probability distributions. Unwarranted confidence always leaves me skeptical, suspecting ulterior motives. I would much prefer to hear about progress toward adaptability. Bad things will happen, no matter what. Let's plan to deal with the bad things so we can minimize future suffering.