My sense is that, by definition, evidence of ways the future will be different than the present is always weak, particularly if the evidence you're looking at is data.