Why Would AI Want to do Bad Things? Instrumental Convergence

推荐理由

Robert Miles' videos are among the best presented arguments about specific points in this list

强烈推荐Veedrac · 原文2023年5月

说明

这是2018年由Robert Miles AI Safety 发布的一部讲解，主题归在AI safety, AGI & impact。它讲的是ai safety instrumental convergence这一支里的具体问题，而不是把整个领域扫一遍。适合第一次接触、还没有先修课的人。内容上可以概括成：Robert Miles explains instrumental convergence: why an AGI with almost any goal would tend to pursue sub-goals like self-preservation and resource acquisition. It builds on Omohundro's basic AI drives and his orthogonality thesis video.这些读者给出的理由，核心是：HN commenters repeatedly recommend it as the clearest introduction to why smart AI might behave badly by default.讲法不绑某一条新闻，过几年仍然可以当材料看。从2019年到2023年，有6位不同的读者在6个不同月份把它推荐出来，其中明确叫别人去看的有3次。适合用一集把一个概念补上的人。抓住它用的那个图像或类比，比背公式重要。如果已经会了基本定义，就把注意力放在它怎么把ai safety instrumental convergence变成能跟着复述的例子。

视频信息

频道

Robert Miles AI Safety

时长

10 分 36 秒

发布年份

2018

同主题更多

• What Happened to Horses Is Happening to Us
• Veritasium: What Everyone Gets Wrong About AI and Learning – Derek Muller Explains
• Intelligence and Stupidity: The Orthogonality Thesis
• We're building a dystopia just to make people click on ads | Zeynep Tufekci
• Intro to AI Safety, Remastered
• Sparks of AGI: early experiments with GPT-4