The graph he shows assumes a strict exponential growth. It's rather hilarious to see a projection that assumes a phenomenon will be at its fastest rate of chang

The graph he shows assumes a strict exponential growth. It's rather hilarious to see a projection that assumes a phenomenon will be at its fastest rate of change when it hits the strict limit (100% of the population). If he would project with a sigmoid growth curve he would get a different answer (much more plausible), but it would be less fun (definitely later than 2531).