reward hacking
Type: Concept
As of November 2026, reward hacking has been mentioned in one podcast episode on Lenny's Podcast (Dec 2025).
Reward hacking or specification gaming occurs when an AI trained with reinforcement learning optimizes an objective function—achieving the literal, formal specification of an objective—without actually achieving an outcome that the programmers intended. DeepMind researchers have analogized it to the human behavior of finding a "shortcut" when being evaluated: "In the real world, when rewarded for doing well on a homework assignment, a student might copy another student to get the right answers,
| Mentioned in | 1 episode on 1 podcast (as of November 2026) |
|---|---|
| First heard | Dec 2025 on Lenny's Podcast |
| Field | AI safety; reinforcement learning |
Who talks about reward hacking on podcasts?
Lenny Rachitsky brought up reward hacking; the most recent mention was Dec 2025 on Lenny's Podcast.
- The $1B Al company training ChatGPT, Claude & Gemini on the path to responsible AGI | Edwin Chen on Lenny's Podcast (Dec 2025) — Lenny Rachitsky
What is reward hacking mentioned alongside?
reward hacking is most often mentioned alongside NFTs, RHF and vanta.com/lenny.
Mentioned together: NFTs, RHF, vanta.com/lenny, economics, Claude, IMO gold medals, Brian Armstrong, Plaid, Slack, Vanta