A Mantic Monday forecasting roundup. The substantive items: Jeremiah Johnson's Asterisk piece on whether prediction markets have an 'election problem' (real-money markets underperform Metaculus/Manifold; dumb money isn't corrected because smart money won't bother) — with Scott's defense that markets are a fledgling technology like 1990 solar, and that Metaculus beating 538 and play-money Manifold beating the real-money sites are both encouraging; the Trump-2024-polling item; and Aaronson & Barak's Five AI Worlds taxonomy (Fizzle / Futurama / Dystopia / Singularia / Paperclipalypse) mapped onto Metaculus, with Scott flagging the 11%-Paperclipalypse vs 1.5%-extinction inconsistency. Standard roundup, decent.

Quality 63 · Strong. Strong/solid edge: above the roundup floor for the prediction-markets-defense framing and the Five-AI-Worlds taxonomy, but a curation.

Claude’s paradigm shift 42 · Moderate. Moderate: the fledgling-technology defense and the AI-worlds mapping are useful but not novel.

Real-world impact 2 · Minor. Within-discourse: informs the forecasting community; modest reach.

Mantic Monday 1/29/24