Mastodon上的一场讨论表明，虽然标题“如果有人建造它，每个人都会死”可能耸人听闻，但对AI风险的根本思考更为细致。作者指出，目前的AI模型尚未被完全理解，导致其行为不可预测，可能引发现实世界的问题。这种不可预测性足以成为将AI远离危险应用和个人的理由。 AI
影响 强调在危险应用中需要谨慎对待AI不可预测的性质，并对其有更深入的理解。
排序理由 Mastodon上关于AI风险细微差别的讨论，引用了一个特定的文章标题。
在 Mastodon — mastodon.social 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
As for it killing everyone, I’d say their thinking is more nuanced than their writing or title would make you think, but you’ll have to work at it. The chance seems non-zero, but there are some big leaps in there (important to note they’re not saying it needs to be conscious). Ev…
To me, that’s enough to say it shouldn’t be allowed near anything dangerous. And many things are dangerous, including a lot of people. # AI
In fairness to “If Anyone Builds It, Everyone Dies,” if you ignore the anthropomorphizing, weird tone, and janky structure and keep reading, there is some good stuff within. Here’s what I feel I got from it: - It is not possible at present to determine what an # AI will do in adv…