The conclusion is that exponentials look better for longer-run trends, if you do fair comparisons. And that linear being a better fit than exponentials in recent data is more about the error-model than the growth-model, so it shouldn’t be a big update against exponential growth.
Great post! I was mainly concerned with the p-values heading haha. I wonder if Thomas Philippon will follow up on all of the attention his paper received.
There’s an excellent critique of that paper on LW: https://www.lesswrong.com/posts/yWCszqSCzoWTZCacN/report-likelihood-ratios
The conclusion is that exponentials look better for longer-run trends, if you do fair comparisons. And that linear being a better fit than exponentials in recent data is more about the error-model than the growth-model, so it shouldn’t be a big update against exponential growth.
Great post! I was mainly concerned with the p-values heading haha. I wonder if Thomas Philippon will follow up on all of the attention his paper received.