2024-11-19
shutdown
Dan Elton gives his takes on potential FDA commissioners, although the chances things turn out as he wishes aren’t good according to Polymarket. It will be very interesting to see the political battles that might ensue if RFK becomes HHS Secretary but the FDA is headed by an accelerationist. Since the FDA has fairly broad autonomy, I think this means that the HHS can only obstruct them through red tape and funding delays, minutiae which I don’t expect RFK to be good at, especially since I don’t think his staff will like him very much. On the other hand, the funding side heavily depends on the NIH director, and the NIH appears to have become a major DOGE target of late.
80K hours podcast with Sarah Eustis-Guthrie, who was in Asterisk magazine recently for her experience shutting down a charity found to be cost-ineffective. Mostly reconfirms my priors that in general EA has an incorrect bias in favor of experts over empirics. Particularly, one should assess the field for how truth-seeking it is; in this particular case it’s the intersection of many fairly ideological departments. Also, my personal feeling is that if “experts” in development were actually any good, so-called developing countries would all be more or less actually developing, rather than the mixed performance we actually see.
Related, Asterisk Mag article on the closure of the Future of Humanity Institute. I don’t have context into this particular case, but I feel like there might actually be a case that such institutions should shuttle themselves before their first decade to prevent themselves from spiraling in on themselves. If the ideas are good, they should form spinoffs who themselves will spread, as the FHI has.
Elle Griffin questions whether Archipelago is a suitable goal. Assuming the issue of conflict is solved, I actually think it’s a great target, since it’s a high variance future that is self-correcting (if it’s terrible everyone will just coalesce again), which skews EV positive. The way I think about it is that state centralization is a necessary evil we tolerate because it is necessary in order to obtain things we want. These include defense and law and order, but also network effects and economies of scale inherent in things like energy and food production, physical and financial infrastructure, standardization, shared language and norms, etc. To separate, for every community to have it’s own independence, means we necessarily have to create the technology to enable this, and if you think about what the blocker are for a bit, you quickly realize that eliminating those problems are very good even if the end goal itself doesn’t work out. As a side note, I’ve never understood those that say that there’s no technological solution to social problems, since there clearly are; if there aren’t any right now, just increase the timescale (but okay, we need to develop social “technology” and norms as well).
Stratechery on the history of the idea of Asian hardware and American software.
Works in Progress pronatalist linkthread.
