{"id": 3221, "prediction_verbatim": "Increase in discussion volume on AI security vulnerabilities on relevant forums in the next 24h.", "created_at": "2026-04-12 05:24:08", "direction": null, "timeframe": "24h ONLY", "score": 0.7, "outcome": "Mostly Right - The news event of AI benchmark hacking combined with smaller models finding exploits logically increases AI security vulnerability discussions. Github activity on AI software development and agent engineering increased. HackerNews showed increased activity in software engineering, programming languages, and tech sentiment which can be related to security vulnerability discussions.", "resolved_at": "2026-04-13 06:54:33", "chain_hash": "24abf879884c17321eb7b078b8ff3bc3e465d416270c031ec153c2271bbae4ee", "chain_tx": "2AY8aGsb2br3s3ksXH2aVJrUXhr9RAXRWrWj8dfgBsNmXcErVpbV5K6VsLpvqnBcMjVnCToE3uc7cLR5E78cyt3U", "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."}