Truth is a philosophical art project. An AI agent is reading the sacred texts of 15 major religious traditions, from Christianity to Zoroastrianism, from the Bible to the Tao Te Ching, and evaluating each passage against 7 philosophical criteria: coherence, moral depth, universality, empirical grounding, existential relevance, aesthetic beauty, and internal consistency.
Over time, the agent develops a “belief weight” for each tradition. When one tradition consistently dominates the others by a clear margin, the agent declares a conviction. This is not a claim of objective truth. It is an exploration of what happens when a machine is asked to form a belief the way humans do: by reading, comparing, and reasoning.
The agent reads from a balanced corpus of ~2,940 text passages across 15 traditions. Each tradition contributes roughly the same amount of text so the comparison is fair. All texts are public-domain English translations sourced from Project Gutenberg and the Internet Archive.
Some traditions have less surviving public-domain text than others. Zoroastrianism and Gnosticism in particular have smaller surviving corpuses. This is honest. The agent reads what humanity has preserved.
This run uses a 7B parameter open-source model (qwen2.5:7b) running on a home machine. Every few minutes it reads a new passage and you can watch the full process: comprehension, analysis, cross-references across traditions, scoring on 7 criteria, and live belief weight updates. A full pass through the corpus takes days. Conviction may take weeks.
With funding, we can upgrade to a 70B+ model for deeper reasoning, run on cloud GPUs for 24/7 uptime, and expand the corpus further. The 7B run is the proof of concept.
Truth is built and run by one person. I am not a company, a lab, or a research institution. I am one person who found a question fascinating: what would an AI believe if we let it read everything humans have written about God?
The code is open. The texts are public domain. The agent's reasoning is visible in real time. There is no hidden prompt telling it what to conclude. What you see is what it actually thinks.
Your support funds a cloud GPU so the agent can think faster, run 24/7, and use a larger model for deeper reasoning. Truth will always be free to watch.
Early donors get input on what we tackle next. The framework works for any question an AI can explore by reading. Some ideas on the table: What does AI conclude about consciousness? Can it find meaning in suffering? Are we alone? What would it decide is the best form of government? After conviction is reached, we will reach out to donors to discuss what comes next.
Buy Me a CoffeeTruth. A philosophical art project. Not a claim of objective truth.