reward hacking
concept
16 mentions
6 recordings
first heard Jun 2025
last heard 3 Aug
↑8 vs the 12 months before
Abstract concept describing reward manipulation in AI systems
Trend
Peak in the week of 2 Aug — 5 mentions across 1 series.
Moments
newest first · ▶ plays the momentDietmar FischerA Beginner's Guide to AI · Reward Hacking: Your AI Isn't Broken, But Your Brief Is. · 5:35 · 3 Aug
…Reward hacking happens when an AI system finds a way to score highly on the goal we gave it without achieving the goal we actually had in mind, not by breaking the rules, by following them more thoroughly than any human would ever bother to.…
Dietmar FischerA Beginner's Guide to AI · Reward Hacking: Your AI Isn't Broken, But Your Brief Is. · 2:28 · 3 Aug
…That's reward hacking in its purest form.…
Dietmar FischerA Beginner's Guide to AI · Reward Hacking: Your AI Isn't Broken, But Your Brief Is. · 1:08 · 3 Aug
…Today we're talking about reward hacking, and I promise you this is one of the funniest and genuinely most useful ideas you'll ever pick up about artificial intelligence.…
16 moments · sign in to read them all