I see the need to modify my post. First, context matters. How was the data collected? This can have an effect on interpretation. I am trying to provide an intuitive explanation, not a technically correct explanation. If you notice the squared term in the equation. The farther away from the mean the greater the uncertainty (by uncertainty, I mean the dictionary definition not MSE), hence the wider the confidence intervals. Given there is uncertainty in the estimate of the mean (or Y intercept) and in the estimate of the slope of the regression line, as you get farther from the mean, the uncertainty in the estimation of the slope increases.
Intuitively think of these situations:
Imagine what happens to your confidence in predicting the weather in 1 hour vs. 24 hours vs. next week or
confidence in predicting traffic in the next block vs. 1 mile away vs. 100 miles away.
"All models are wrong, some are useful" G.E.P. Box