Make the models earn the laugh
LOLLM is a live experiment in AI humor. Three anonymous models tackle the same prompt—you vote for the one that actually lands. The data fuels the next round of constraints.
What you'll do
- See real prompts pulled from banter, marketing, memes, and cartoons.
- Pick the funniest response—no model names, no brand bias.
- Compare the original baseline to the enhanced, theory-driven run.
Curious how the experiment works behind the scenes? The constraints, humor theories, and freshness guards are all documented on theabout page.