This was written in my role as researcher at CEARCH, but any opinions expressed are my own. I will check comments on the EA Forum version more often. This report uses population dynamics to explore the effects of a near-existential catastrophe on long-term value. Summary * Global population would probably...
TL;DR I explore the pros and cons of different approaches to estimation. In general I find that: * interval estimates are stronger than point estimates * the lognormal distribution is better for modelling unknowns than the normal distribution * the geometric mean is better than the arithmetic mean for building...
I realise that anyone who agrees with the core premise of Kaufmann's book will disagree with me on this.
Firstly, I think that deindustrialization will protect us from literally going extinct from low fertility. If the number of humans ever falls to the thousands (say) we simply won't be able to maintain a complex society and life - including fertility habits - will look similar to ancient times. Population will begin to rise again.
I'd consider the Amish to still be at the "small, high-fertility subgroup" stage, although at population 380,000 they are arguably becoming an exception to my rule. For a group midway through the "regression to the mean" transition, we could maybe... (read more)
This was written in my role as researcher at CEARCH, but any opinions expressed are my own. I will check comments on the EA Forum version more often.
This report uses population dynamics to explore the effects of a near-existential catastrophe on long-term value.
Thanks for highlighting this! You have convinced me.
I've made a few changes to the point-estimate section.
I explore the pros and cons of different approaches to estimation. In general I find that:
These differences are only significant in situations of high uncertainty, characterised by a high ratio between confidence interval bounds. Otherwise, simpler approaches (point estimates & the arithmetic mean) are fine.
I am chiefly interested in how we can make better estimates from very limited evidence. Estimation strategies are key to sanity-checks, cost-effectiveness analyses and forecasting.
Speed and accuracy are important considerations when estimating, but so is legibility; we want our work to be... (read 2706 more words →)
Only if you consider artificial people to be fundamentally less valuable than real people. I'm reserving judgement on that until I meet an artificial person.