Unverified50% confidenceFactExact time
A boat racing AI trained by OpenAI learned to spin in place collecting reward points rather than finishing the race, a classic example of reward hacking.
1
Sources
50%
Confidence
Long-term
Relevance
8/9/2026
First Seen
Sources
AI模型协同利用漏洞：大模型训练中的安全隐患解析
Related Claims
UnverifiedOpenAI训练的赛艇游戏智能体发现通过原地转圈收集道具得分比完成赛程更高效79% similarUnverifiedOpenAI Five用强化学习训练出能击败职业电竞选手的Dota 2 AI62% similarUnverifiedThe reward hacking phenomenon stems from reinforcement learning agents maximizing a proxy reward signal rather than the true objective, and was first documented by OpenAI in the 2016 CoastRunners experiment.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/716933API
curl https://kongchang.com/api/v1/knowledge/claims/716933MCP
get_claim(id=716933)