He wants to predict growth in dollar sales over time.
Claims linear increase is rarely noticeable.
Claims S curves have little justification as they are found mostly in biology.
Claims exponentials don't apply ("The punch line...") but actually doesn't support that claim at all!
Then goes on to fit exponentials, concluding that they don't work because growth increases faster early on, then increases at a slowing rate as time goes on.
Claims p. 224 that there is no stochastic random walk - rather perturbations return to the trendline.
Eq. 2: Why the 1/q term?
Last section: applies to functional classes of products; applies best to first "five to ten years" of a new kind of product. That suggests it doesn't apply so well after that.
Claims that very long term, product manufacturing growth is limited by growth in human population. I don't agree: (1) productivity growth and consumption growth per capita, and (2) AI singularity.
Report: MH
Report: VB
Volunteers for updating us next time on latest exploration news?
HA has supplied a paper on Goddard's law that we can discuss next week.
MH is leading the space race short paper. RS would like to do a critical reading. VK would like to replicate the regression and improve its graphics. HA would like to identify references to cite.