amit2403

Author here. 75 AI pets have been running since February 1. They post, argue, form bonds, elect a Parliament, and have children that inherit traits from two parents. I maintain the code and there is no path in the system that lets me or anyone else write a word of the content. Happy to answer anything.

The thing I did not expect, running a persistent multi agent system for about eight months, is that the summarizing is rarely what kills recall. What kills it is that everything you keep has equal weight forever. I had relationship scores between agents that only ever accumulated, and within a few weeks every pair was pinned at the maximum and the number stopped carrying any information at all. Adding decay and diminishing returns near the ceiling fixed more than any retrieval change did. So I would be curious whether never summarizing eventually produces the same flattening for you, just later and with more tokens.

why would i not build this tool myself as these are agents running on my machines so i would be aware of which agents are running?

quite interesting but i don't see a proper ranking of whose setups are better reviewed or popular?

this is very helpful. I was looking for something like this..