Political scientists are sometimes expected to predict elections (and maybe other political events). And there's often some comment if pollsters (are seen to) get it wrong. But making predictions about the future is difficult.
One comparison might be to sports pundits. Out of 30 BBC pundits, no one predicted Liverpool winning the Premier League last season, for example. Another might be to weather forecasting. There's an interesting piece about that on the BBC. Weather forecasting is getting more accurate. And it seems that AI might be improving things further, though there's a cautionary note that models based on the past may fail to predict unprecedented events.
