Followup: Quests And Requests
Read the original on Astral Codex Ten →
Summary
A long followup to 'Quests And Requests' that (a) gives Scott's status update on the eight proposed projects — flagging the predictable failure mode that lots of people volunteer but few take point — and (b) curates the expert/volunteer comment threads. Several threads carry real substance: Gustav Nilsonne's careful takedown of the EEG-learning replication (control-group omission, p=0.045, underpowered, analytic flexibility); a rich open-source-polygenic-score feasibility debate (UK Biobank access barriers, EA-vs-IQ heritability divergence, pgscatalog workflows); the classical-architecture thread's cost/regulation/elite-opinion diagnosis and the 'why do new buildings look like sophisticated blobs' trapped-prior story, with Scott's pushback on the 'popular architecture' reframe and his pro-choice/Schelling-category analogy; and the political-change-primer thread (Erusian and others on lobbying, whipping, patron-client relations, administrative comment periods), where Scott's electrical-engineering-textbook analogy nicely frames what he's actually asking for.
Why this score
Quality 68 · Strong. Strong: a substantive companion — the architecture and political-change threads are genuinely illuminating and Scott engages rather than just aggregating — but diluted by a lot of coordination logistics and 'here's who volunteered,' so it sits mid-Strong at 68, not up with the tightest companions.
Claude’s paradigm shift 46 · Moderate. Moderate: synthesizes and reacts to others' contributions; the reframes (popular-vs-classical, the textbook analogy) are sharp but incremental. 46.
Real-world impact 2 · Minor. Actually mobilized real community effort — the dating site, open-source polygenic score, and other quests got teams and leads out of this thread — modest concrete organizing beyond pure discourse. RWI 2.