hmu if you want to talk futarchy/prediction markets, or have any questions! cal.com/distbit

Long-dated prediction markets lose accuracy when traders must lock collateral until resolution: the opportunity cost makes small price corrections unprofitable. One fix is bond-like collateral, such as USDC redeemable only when the market resolves, though this is costly to issue. Since it is already time-discounted, traders can mint prediction-market shares without paying an additional time-discount cost. The hard question is optimal supply of said bond-like USDC collateral. Demand can come from prediction-market traders or parasitic savers who just hodl, and the system cannot easily distinguish them.

life rewards action not intelligence

event-conditional markets are useful when traders want exposure to or to hedge the effect of a specific event rather than ordinary price volatility. options already serve many hedging needs outside the event. how much options-based hedging & speculative demand is actually event-specific?

llms fit offensive security more easily than defensive security. defensive monitoring agents need access to untrusted inputs, sensitive internal data, and channels for taking action or reporting results. that puts them near the "lethal trifecta for ai agents," so the monitor itself becomes an attack vector. offensive tooling does not face the same self-exposure constraint: threat actors can run it without giving it access to their own sensitive info.

Futarchy is supposed to isolate a proposal's effect, but traders still hold positions whose returns can be swamped by unrelated asset-price moves. How can futarchy trader returns be isolated from external factors such as asset price?

Strategies for passing malicious futarchy proposals: - Fail-Branch Sabotage - Resistance-Contingent Delivery - Committed Holder Extraction - Proposal Convexity Maximisation - Counter-manipulation Deterrence Details ----> distbit.xyz/malicious-futarc…

economist.com/democracy-in-a… the long-run dominance of effective error correction over ~all other properties is quite underappreciated

Adverse selection is hard to measure robustly: the more control or info-asymmetry an insider has, the easier it is to make market activity look organic, via fake volume, depth, or price moves, because others fear calling their bluff and making it costly.

Would attempted bribery of an equity value-maximising futarchy oracle into making false reports be structurally unprofitable, given the bribe can be symmetrically countered by the party harmed by the false report? (Assuming the initial bribe is large enough to offset the false report’s effect on the futarchy’s reputation, hence expected future revenue, hence its equity price)