Economist Aaroth: fast agreement dynamics survive without Bayesian updates or accurate priors
Aaroth · x · 2026-11-02
Economist Aaroth threads through new work on agreement theorems and information aggregation. He gives a clean counterexample: with binary observations XA, XB and outcome Y = XA xor XB, sharing observations determines Y exactly, yet a mere beliefs conversation converges on Pr[Y=1]=0.5 — agreement is not the same as information aggregation.
The key obstruction is that observations are complements rather than substitutes; for substitutes, agreement implies aggregation. His own contribution removes seemingly strong computational and epistemic assumptions: you need neither Bayes updates nor accurate priors to recover the fast dynamics, and the result is robust. Notably, conversation length is completely independent of the complexity of the objects being discussed.
Related event: Economist Aaron Roth Shows Agreement Theorem Holds Without Common Priors(4 posts)→
More from Research
- GeBNLP 2026 goes virtual Nov 9 at AACL with keynotes from Savoldi and Gautam — debora_nozza · 2026-11-02
- Berkeley's Evolving Agent Teams beats production FlashInfer by 1.2x on CUDA kernels from scratch — berkeley_ai · 2026-11-02
- Swarm-based auto-research engine runs 500+ experiments on one agent thread — vincent_koc · 2026-11-02
- ICML paper: finite context windows impose impossibility results for adversarial regret — burny_tech · 2026-11-02
- Labs now train on mini-swe-agent to game evals, says Xeophon — xeophon · 2026-11-02
- WASABI: Blei lab trains diffusion models to sample Boltzmann distributions with 3.3x fewer energy evaluations — skoularidou · 2026-11-02