I don't trust complex models. I've written basic simulations with less than 20 variables to model far simpler phenomena and seen how a slight mistake in just 1

I don't trust complex models. I've written basic simulations with less than 20 variables to model far simpler phenomena and seen how a slight mistake in just 1 variable can invalidate the entire results in a dramatic way. It doesn't give me much hope for simulations with millions of variables. I suspect that these are completely worthless. It seems like every year, scientists are cherry-picking among thousands of different models for the one which has been right for the last decade or so... They keep changing models every few years and using 'technology has improved and made the old model redundant' as an excuse for doing so as opposed to 'the old model turned out to be wrong this year so we decided to switch to a newer model which has been right.' If they keep creating new models and changing them every few years, they will keep finding new models which have been 'scarily accurate' for the last 10 years or so. They will never run out of such models even if the models were randomly generated. That said I also wouldn't say that there is no global warming or that humans have no effect on it. I just think humans often overestimate their knowledge and capabilities, especially when there is a financial incentive for doing so.