Prototyping a prediction-accuracy tracker on HN comments with an LLM
Darpinian · x · 2026-11-13
Responding to the common claim "I've been right before, so trust my future predictions," Darpinian notes it would be easy to prototype a verification tool on Hacker News comments: the entire database is open, and the only obstacle is the cost of having an LLM check every comment's predictive accuracy in hindsight. He later jokes the product could be called "Truth Social."
More from coding & agent
- Agents Are Burning Tokens: Why Context Engineering Beats Prompt Tricks — Pavan_Belagatti · 2026-11-13
- OpenOrange launches: run Cloudflare Workers, Durable Objects on any cloud, 4.8× cheaper compute — GabGarrett · 2026-11-13
- Two AI voice agents settle a blender refund over the phone with no humans involved — kimmonismus · 2026-11-13
- LangChain launches annual State of Agent Engineering survey — hwchase17 · 2026-11-13
- Simular launches Box: one cloud computer shared by Claude, ChatGPT, Codex and Gemini — rohanpaul_ai · 2026-11-13
- Open-source legal conflict detector runs on encoder models only, 133-clause contracts in under 20s — ladyofcode · 2026-11-13