New paper: the 'creative paradox' — training directly for creativity backfires
kenneth0stanley · x · 2026-11-18
Kenneth Stanley's team, led by Alon Albalak, published a paper on the "Creative Paradox": using past creative successes as training examples makes similar outputs more likely, yet reproducing past examples is the opposite of creativity — so directly training for creativity can cap the model's creative potential. They propose "Creative Priming," a decoupling method designed to sidestep the paradox and boost creativity without emulating prior successes.
More from Research
- Vals AI: frontier models mark civilians and ambulances as drone strike targets — scaling01 · 2026-11-18
- Researcher finds negative parallelism and participle tails as AI writing tells across 187 books — TuhinChakr · 2026-11-18
- Running a literature research project inside Grok bot, from question to grant aim — MaxUnfried · 2026-11-18
- SenseTime open-sources SenseNova-Loop-9B, a looped-architecture 9B matching 27B-class Qwen — ChengleiSi · 2026-11-18
- DeepMind & UChicago paper: AI agents gain more sway in deliberation when wordier and less cooperative — rohanpaul_ai · 2026-11-18
- DeepMind Paper: AI Agents Speaking for People Gain Influence by Being Wordy, Less Cooperative — rohanpaul_ai · 2026-11-18