By this probabilistic reasoning, we shouldn't bother searching for "meaningful" polynomial time algorithms because they are practically non-existent in the sea

By this probabilistic reasoning, we shouldn't bother searching for "meaningful" polynomial time algorithms because they are practically non-existent in the sea of all exponential time algorithms. If anything, this argument should be a motivator to work on a problem because if an underlying structure exists, it would be rare and beautiful. Very nicely written article, showing us a glimpse of the inner workings of a great mind.