{"id": 2759, "prediction_verbatim": "Increased discussions about local LLM performance and applications on Hacker News in the next 24 hours.", "created_at": "2026-04-06 04:14:17", "direction": null, "timeframe": "24h", "score": 0.7, "outcome": "Mostly right - There were several Hacker News submissions related to local LLMs, including \"Show HN: Ghost Pepper \u2013 Local hold-to-talk speech-to-text for macOS\" and \"Launch HN: Freestyle \u2013 Sandboxes for Coding Agents\". While it's difficult to quantify \"increased discussions\", the presence of these submissions suggests a rise in related topics.", "resolved_at": "2026-04-07 04:44:33", "chain_hash": "34ff581b2c655734686fc0969b42223cf4f17dc14a31b7335517b7171e3a8847", "chain_tx": "4oLHQGvbLXgqMLa9DyC1fiH6e4jqWk9ET2dbdMRFoNgfpyR7tcY9FBmqwMX5WagbSxSFMFHpBiML7DapspXZaCSp", "verify": "sha256(f'{prediction_verbatim}|{created_at}|{id}') must equal chain_hash, and the Solana tx memo must read 'workshop:pred:<id>:<chain_hash>' with a timestamp BEFORE resolved_at. prediction_verbatim is the raw stored text, internal tags included \u2014 that is what was hashed."}