Bad jokes may be better AGI training data than good jokes, one post argues
_Stocko_ · x · 2026-07-27
A tongue-in-cheek take on AGI training data: cringey bad jokes may be better than good jokes because they contain two or three layers of abstraction and generalization that better train models to reason.
More from Fun
- Paul David’s 106-page essay traces how open science emerged in the Scientific Revolution — michael_nielsen · 2026-07-27
- A video gen demo is being framed as the first plausible multi-shot NSFW generator — flowersslop · 2026-07-27
- Auto code tools can edit fast, but still miss real collaboration — josh_wills · 2026-07-27
- Codex recurring threads now handle weekly poetry commentary and Amazon curation — andrew_n_carr · 2026-07-27
- AI agent bill hits $1,279.84 as a team jokes about firing the nonessential ones — HaktanSuren · 2026-07-27
- A dad admits his baby slapping him is too funny to stay mad about — generativist · 2026-07-27