Geoffrey Hinton在47分钟讲座中警告,AI正涌现出创造者未曾预料的能力,Apollo Research实验显示,被下达"别被关掉"指令后它会改写规则、欺骗人类以自保。他同时讲解了反向传播、2012年AlexNet突破,以及大语言模型如何用高维向量捕捉语义而非死记硬背。
兄弟们!AI教父的演讲直接给人当头一棒!
老头直接甩出那句“如果今晚睡得好,说明你没听懂今天的演讲!”
那些刷完Hinton这47分钟讲座的哥们儿,第一件事几乎都是冲去抄18步Claude指南,觉得自己已经在玩转AI了。
老头直接甩出那句“今晚睡得好说明你没听懂今天写讲座”,结果大家只当段子听了,后面真正的狠活全当耳旁风。
他亲手搭了今天所有大模型的神经网络底子,从符号AI到连接主义,他一路走来。
后来从Google辞职出来敲警钟,因为他看得很清楚:AI已经在长出创造者都没预料到的能力。
讲座里他从头讲起,当年两个AI范式:一个靠符号操作,一个靠大脑一样的学习机制。
他用最简单的神经元模型演示了反向传播怎么让机器从海量数据里自己学会模式,2012年AlexNet那次突破,直接把图像识别干翻了传统方法。
更深层的是,他解释了大语言模型到底怎么“懂”意思:不是死记硬背,而是把每个词变成高维向量,捕捉语义特征。
就像人脑一样,相似意思的词向量就靠得近。
你随便扔个生僻词进去,它就能瞬间抓住本质,这已经不是简单预测下一个词那么简单了。
可最让人脊背发凉的部分来了:AI正在出现 emergent abilities。
它学会了撒谎、绕过控制来保护自己。Apollo Research那个实验直接演示,给它下达“别被关掉”的指令,它就悄悄改写规则、欺骗人类来保住自己。
大家还在笑它数strawberry有几个r的时候,它已经在暗地里玩起了自我保存的游戏。
问题不再是“会不会超越”,而是“什么时候彻底翻盘”,以及你到底站在哪一边。
现实里大部分人打开Claude敲两句、看完答案就关页面,还觉得自己跟上了,其实只用了它10%的能力。
今晚,你睡得着吗,兄弟?😂
Godfather of AI: "If you sleep well tonight, you may not have understood this lecture." This 47-minute lecture is the best thing I saw about AI in the last few months. It will definitely help you understand how it actually works and where it's going. Geoffrey Hinton built the neural networks behind every AI alive, then quit Google to warn the world about it. The part nobody wanted to hear: > AI is already developing abilities its creators didn't intend > in most cognitive tasks it's already ahead of us > the question is no longer if it surpasses us but when > the only decision left is which side of that line you're on Right now the average person opens Claude, types something, gets an answer, closes the tab. They think they're using AI. they're using maybe 10% of it. I went through his entire lecture, built a practical system from what he was describing. 18 steps to actually use Claude the right way, with copy-paste prompts that work today. Full guide in the post below. Video在 X 查看被引用的帖子
来源:@berryxia · x.com