Section
Experiments
Various experimental projects exploring the intersection of AI, technology, and knowledge systems.
Study · August 2026 · Philoscopia
How do generative AIs respond when you bring them a philosophical question?
Six models, five languages, close to five thousand answers analysed.
Philosophical questions were put to ChatGPT, Claude, Gemini, Grok, Mistral and DeepSeek, and the answers analysed under a protocol fixed before any collection began. An answer only lays out reasoning when the question already carries a problem: asked about a personal situation, the models give practical advice almost every time, whatever the phrasing. And laying out reasoning is not yet weighing positions.
The second half tests a remedy: a philosophical debate map listing the questions with the positions competing on each, the framework behind Philoscopia. It triples problem framing and cuts the share of unmapped questions from 56 % to 15 %. Raw data and a reproducible experiment are published alongside.
In development · not yet released
The Laboratory of Artificial Mind
Mapping the space of philosophical possibles.
An exploration of the space of possibilities opened by thought experiments, with AI: vary one parameter of a classic dilemma and see whether the judgment moves, for humans and AI models alike. More soon.
See all philosophy projectsComing soon
More experimental projects and ideas will be showcased here.