AI 放慢节奏是公关失误，网络攻击才是真风险

核心论点 · 点时间戳可跳到原声

原话 · 已逐字校验

First off, it's orthogonal to safety and security. Like you can slowly build a weapon. That's not different than building a weapon.

首先，它与安全和安保是正交的。你可以慢慢地造一件武器，那和造武器没有区别。

Ali Ghodsi6:16

there's so much infrastructure that's insecure right and if you're going to unleash these agents they're going to find loopholes they're going to find exploits they're going to break in here and there

有太多不安全的基础设施，而如果你放出这些 agent，它们会找到漏洞，会找到 exploit，会到处入侵。

Ali Ghodsi20:39

I do think that there are people saying like, hey, if there was a regulation that would pace us, sorry to use the word, that would be good for us. Right. That would be good for us.

我确实认为有人在说：嘿，如果有一项监管能让我们放慢节奏——抱歉用了这个词——那对我们有好处。对，那对我们有好处。

Ali Ghodsi32:44

But at the same time, also, as I said, the time from CVE to actually weaponize exploit has like been going down from years down to like minutes now, just in like three, four years. So it's real. The cyber attacks are real. And this is, but there's also a great marketing ploy to, you know, whenever you train a new model, make lots of noise around how much of a, you know, crazy risk it is to the world.

但与此同时，就像我说的，从 CVE 到真正把漏洞武器化利用的时间，在三四年里从数年降到了几分钟。所以这是真的。网络攻击是真的。但这也是一种极好的营销手段——每当你训练出一个新模型，就大肆宣扬它对世界是多大的风险。

Ali Ghodsi33:45

One has an ontology of how that organization works, who the people are, how you get stuff done. Don't look at the org chart. Don't go ask that person. He will not get anything done.

其中一个人拥有关于这个组织如何运转、人是谁、事情怎么办成的 ontology。别去看组织架构图。别去问那个人。他什么都办不成。

Ali Ghodsi47:01

So how do they do it? They have an index, right? You never leave Google servers. You search for, it hits the index, the reverse index immediately gets you the 10 blue links within, you know, less than 100 milliseconds. We need to do the same thing for the AI.

那他们是怎么做到的？他们有一个索引，对吧？你从不离开 Google 的服务器。你一搜，命中索引，倒排索引立刻在不到 100 毫秒内给你十条蓝色链接。我们需要为 AI 做同样的事。

Ali Ghodsi49:02

Turns out actually the harness itself matters. Like if you use the same model but different harnesses, there's almost 2x different cost difference. Even exactly the same model. You know, same version but different harnesses. You get 2x difference in actual cost.

事实证明 harness 本身很重要。比如你用同一个模型但不同的 harness，成本能差将近 2 倍。哪怕模型完全一样。同一个版本，只是 harness 不同，实际成本就差 2 倍。

Ali Ghodsi56:08

So, by dollar, open source is like 5%. It's very little, but by token count, it's over 60%.

按金额算，开源大约占 5%。非常少，但按 token 数算，超过 60%。

Ali Ghodsi59:09

数字与实体

术语

pacing放慢节奏

前沿实验室用来包装安全议题的说法，主张主动放慢 AI 发展速度。

RSI递归自我改进

AI 能改进下一代 AI，形成自我加速循环。

CVE通用漏洞披露

公开记录软件安全漏洞的编号体系。

ontology本体

组织内部关于人、流程、权限的隐性知识结构。

harness运行框架

包裹模型、负责调度工具和上下文的软件层。

branching分支

在同一数据库上轻量开出多个分支，供 agent 并行实验。

收听指南