PT-Contest results

Cars Computers Cats Cold Baseball Books Fire
Team #1 4 4 6 4 4 5 5
Team #2 4 5 7 4 4 6 6
Team #3 6 3 5 6 4 6 4
Total 14 12 18 14 12 17 15
Mammals Robots School Houses Total
Team #1 2 3 4 4 45
Team #2 3 6 5 6 56
Team #3 3 4 5 4 50
Total 8 13 14 14 151

The numbers in the above table represent the number of correct answers given by each teams "artificial person" in each topic area.

A Brief Analysis

Given that there were 10 questions in each of the 11 topic areas and that each team was given 2 questions in advance and contributed 2 questions of their own, each team should have been able to answer 4 questions in each topic area for a total of 44 correct answers at the end of the test.

All 3 teams did better than 44 correct responses overall. In fact the winning team's "person" gave 56 correct responses, 27% more than expected. This means that in addition to answering the questions that it should have, Team #2 was able to give their person a broad enough belief structure to enable it to answer questions that the team did not know about in advance. In other words they were able to give it enough background information to enable it to field questions it had never seen before. Given that their time was extremely limited (only several hours in real time) this seems truly remarkable.