"x is with a likelihood of 95% between a and b" -- that is not what a confidence interval means. Confidence intervals are about as confusing as p-values and don't really solve the problem.