

AI x Epistemics Hackathon: Bangkok Hub
AI is transforming how humans evaluate facts and make decisions on what to do and believe, and how experts make judgment calls.
How can we build a tool to help people assess evidence, check claims, and make informed decisions?
These advanced systems can help clarify complex data, but they can also centralize authority and rely on unverified blackbox models.
Join AI Safety Thailand and Apart Research for the 3-day AI x Epistemics Research Sprint
How works
You’ll pick, build tools, and run experiments from four tracks:
Research tracks
Model epistemics and decision-making evals. Test whether model confidence matches real evidence and leads to good decisions.
Trust infrastructure. Build tools to verify claims easily and trace sources accurately.
Epistemic products. Create software that helps people reason clearly and make better decisions.
Open track. Explore broader ideas, like how to measure improvements in decision quality and encourage tool adoption
Timeline
Friday, 18:00: Meet potential team mates, collaborators, and discuss project ideas in person.
Saturday, Sunday, Monday: Work on your project until submissions close on Monday, 16 November, 6:29 PM Thailand time.
What you get
$2,000 in cash prizes across tracks.
Fast-track applications for top teams to join the Apart Fellowship.
Mentor introductions to help publish your work.
A complete research project and a team by Sunday night.
Bring a laptop and your ideas. No prior AI safety or machine learning experience is required.
Who should join
Open to all. Anyone can participate regardless of experience or background. You can join solo without a team or work in groups of up to five.
We also welcome engineers, researchers, students, and domain experts in finance, health, law, journalism, forecasting, and policy. Many projects focus on structured analysis and evaluation rather than coding.