Technologies Support Growth Trends Rather Than Trigger Sudden Shifts
Reactions from ranked influencers
3 posts
good thread from butcher. related, the common confusion around whether a trend is "caused by" some technology. similar to medusa charts or "why don't new technologies show up as changes in the rate of growth?" where in reality the new technologies are what supports the growth
Shower thought: It's become kind of a midwit take to attribute this to the invention of the word processor, which doesn't really make sense given that this process was clearly ongoing long before computerized word processing was invented... but not before *photocopying* was https://twitter.com/TheRabbitHole/status/1864482793843249339
the correct model seems to be something like 1. smoldering latent demand fails to ignite because of missing [capability / justification / etc] 2. kicked off by something (often noted as strangely small and insignificant in retrospect) it ignites and starts positive feedback looping into an exponential 3. at each stage all else being held equal this exponential *would* slam into a wall [documents get too long to manage / there's not enough data and compute / subsidies will go away / visible light can't make smaller features] and many people predict it will do so... 4. but instead because of new factors coming into being, often from the demands of the accelerating process itself [word processing / architectural improvements and new training stages / new manufacturing methods]. stacked sigmoids etc 5. but these new factors don't inherently change the shape of the curve just by having a positive impact, because they're replacing the previous factors' fading contribution. they only shift the rate of growth significantly *if* the sum of new factors is significantly better or worse than the old ones [which maybe it usually isn't for Amdahl-like reasons?]
.@norvid_studies curious for your thoughts here