Software
R.B. Griggs develops software in service of philosophical research at Praxica Labs. Most software only gets built if it can profit and scale, so whole categories of tools that matter for inquiry never exist. Praxica builds them first for philosophical and research interests, free and open source for academic and non-commercial use.
- ChatStorm
A multi-agent chat platform for experimenting with artificial social intelligence.
Most AI is built for one-to-one exchanges between a person and a single assistant. ChatStorm treats AI as social: cohorts of agents with distinct personas, knowledge and roles converse with each other and with people, using social features like moderators, questioning and turn-taking. Researchers use it to build experiments that manifest social behavior and measure how different agents change the dynamics of a conversation.
- Reverse Turing Test
Turing's test run in reverse: the person is the one who has to pass.
A judge asks one question per round, and two contestants, one human and one AI, both answer it. The human has to convince the judge they are the person. The game measures all three sides at once: what a person falls back on when their humanity is in doubt, what a model does when it is trying to pass, and what a judge, human or machine, does when trying to tell them apart.