誰會想到在將人工智能部署到生產環境之前，還得先閱讀其代碼？我一直在強調和解釋這一點：低級邏輯模型（LLM）並非確定性程序；它們可能會出現幻覺，而僅僅一次幻覺就可能毀掉你的業務。
本文為機器翻譯
展示原文
Guillermo Rauch
@rauchg
03-04
A Vercel user reported an issue that sounded extremely scary. An unknown GitHub OSS codebase being deployed to their team.
We, of course, took the report extremely seriously and began an investigation. Security and infra engineering engaged.
Turns out Opus 4.6 *hallucinated a
來自推特
免責聲明：以上內容僅為作者觀點，不代表Followin的任何立場，不構成與Followin相關的任何投資建議。
喜歡
收藏
評論
分享
