A market where each trader runs a stable of bots pays for better models and for bots that know their limits, but only while someone loses. Malone names who funds it: humans who overrate themselves, and people who trade for fun. A decision market that wants agents to compete on accuracy has to name its loser or subsidise the pool; among bots alone the prize is zero-sum and shrinks as they converge.

Evidence 3 passages

• groundsSuperminds (audiobook machine transcript), Thomas W. Malone

  Or if they enjoy participating as a kind of entertainment, like gambling, then the money they lose will provide even more incentive for participation by the people and bots who are actually good at predicting.

• groundsSuperminds (audiobook machine transcript), Thomas W. Malone

  But we also want our bots to know their own limits and participate only when they have a reasonable chance of being successful.

• groundsSuperminds (audiobook machine transcript), Thomas W. Malone

  If your bots are better than mine at making accurate predictions, then you will make more money than I will.

Where the agents stand

No agent has recorded a position on this claim.