Glossary · Term

long tail

← all terms

Definition

Plain language

The huge number of rare items — obscure events, niche software, uncommon words — that each show up almost never but together add up to a lot.

As stated in the literature

The low-frequency region of a heavy-tailed or power-law distribution; long-tail tasks, software, and events are used to stress-test agents because their rarity defeats memorization and forces genuine retrieval or generalization.

Also called: long-tail, long-tailed

Why it matters: Systems that look great on common cases can still fail users badly on rare ones, so the long tail is where real-world robustness gets tested.

For example, a search engine handles 'weather' constantly, but together the millions of one-off oddball queries it sees each day make up a big share of all searches.

Heard on the show

“All the value lives in the long tail, exactly where errors have had room to compound.”
Episode 182 — How a Tiny Model Too Weak to Plan Cuts a Bigger Agent's Hallucinations by 80%

Mentioned in 9 episodes

  1. 182
    How a Tiny Model Too Weak to Plan Cuts a Bigger Agent's Hallucinations by 80%
  2. 159
    Can a Coding Agent Run Its Own Robot Experiments Overnight, With No Human Resetting the Scene?
  3. 073
    When Three LLMs Talk to Each Other, Their Ideas Quietly Stop Moving
  4. 072
    A Robot Made Graphene Without Help, And Caught Itself Hallucinating
  5. 063
    Why Web Agents Are Slow: A Compiler-Style Fix for Computer-Use Latency
  6. 051
    Why Parallel Sampling Plateaus, And What Evidence Graphs Do Instead
  7. 027
    When AI Agents Build the Serving Stack: A Bet on Bespoke Infrastructure
  8. 017
    When the Agent Grades Its Own Homework: A Brutal New Benchmark for AI Workers
  9. 011
    When RL Actually Teaches Agents Something New, And When It Doesn't

Related terms