① 他说过的话

13 条

因此支付个人像 1 亿美元的四年的方案，相比于为业务创造的价值，这实际上是相当便宜的。
And so to pay individuals like $100 million over four year package, that’s actually pretty cheap compared to the value created for the business.
—— Benjamin Mann · [07:01]

指向原始笔记的链接

这有点好笑，因为这个叙事每六个月左右出现一次，而且从未成真，所以我有点希望人们看到它时脑子里能有一个废话探测器。
It’s kind of funny because this narrative comes out every six months or so and it’s never been true, and so I kind of wish people would have a little bit of a bullshit detector in their heads when they see this.
—— Benjamin Mann · [08:06]

指向原始笔记的链接

对我来说，如果你看物理的基本定律，其中许多并不跨越 15 个数量级都成立，所以这非常令人惊讶。
To me, if you look at fundamental laws of physics, many of them don’t hold across 15 orders of magnitude, so it’s pretty surprising.
—— Benjamin Mann · [09:39]

指向原始笔记的链接

这个想法是，如果你为一个特定工作雇佣一个 agent 一个月或三个月，如果你决定雇佣那个 agent 并且结果证明那是一台机器而不是一个人，那么它就通过了该角色的 Economic Turing Test。
It’s this idea that if you contract an agent for a month or three months on a particular job, if you decide to hire that agent and it turns out to be a machine rather than a person, then it’s passed the Economic Turing Test for that role.
—— Benjamin Mann · [11:24]

指向原始笔记的链接

如果它第一次不起作用，再问三次，因为当你完全重新开始并再试一次时，我们的成功率比你只尝试一次然后继续在同一个不起作用的事情上敲打要高得多。
And if it doesn’t work the first time, asking three more times because our success rate when you just completely start over and try again is much, much higher than if you just try once and then just keep banging on the same thing that didn’t work.
—— Benjamin Mann · [19:13]

指向原始笔记的链接

即使是现在，我的意思是行业正在爆发，正如我提到的，如今每年 3000 亿美元的资本支出，我会说全世界从事这方面工作的人可能不到 1000 人，这太疯狂了。
Even now, I mean the industry is blowing up, as I mentioned, 300 billion a year CapEx today, and I would say maybe less than 1,000 people working on it worldwide, which is just crazy.
—— Benjamin Mann · [26:14]

指向原始笔记的链接

如果你看看像谄媚这样的东西，我认为 Claude 是最不谄媚的模型之一，因为我们投入了大量的精力在实际的对齐上，而不仅仅是试图在我们的指标上玩弄花样，说用户参与度是第一位的，如果人们说是的，那对他们就是好的。
And if you look at something like sycophancy, I think Claude is one of the least sycophantic models because we’ve put so much effort into actual alignment and not just trying to good heart our metrics of saying user engagement is number one, and if people say yes, then it’s good for them.
—— Benjamin Mann · [27:21]

指向原始笔记的链接

我们不想要猴子爪子场景，精灵满足这三个愿望，结果你接触的一切都变成了黄金。
We don’t want the Monkey Paw Scenario of the genie gives these three wishes and then you end up having everything you touch turns of gold.
—— Benjamin Mann · [30:29]

指向原始笔记的链接

然后 ASL-5 是潜在灭绝级别的，如果它被滥用或者如果它 misaligned 并做自己的事情。
And then ASL-5 is potentially extinction level if it’s misused or if it is misaligned and does its own thing.
—— Benjamin Mann · [37:51]

指向原始笔记的链接

一旦我们达到超级智能，可能就太晚来对齐模型了。
Once we get to superintelligence, it will be too late to align the models probably.
—— Benjamin Mann · [42:49]

指向原始笔记的链接

我认为在短短几年内达到某种超级智能的 50% 的概率大概是合理的。
I think 50th percentile chance of hitting some kind of superintelligence in just a small handful of years is probably reasonable.
—— Benjamin Mann · [46:13]

指向原始笔记的链接

我们在行业内看到，通过算法数据和效率改进的结合，给定智能水平的成本降低了 10 倍。
We’ve seen in the industry a 10X decrease in cost for a given amount of intelligence through a combination of algorithmic data and efficiency improvements.
—— Benjamin Mann · [58:33]

指向原始笔记的链接

如果它们对你来说似乎不狂野，那你一定是生活在石头底下，但也要习惯它，因为这是它将呈现的最正常的状态。
If they don’t seem wild to you, then you must be living under a rock but also get used to it because this is as normal as it’s going to be.
—— Benjamin Mann · [69:40]

指向原始笔记的链接

② 出现在这些集

1 集

• 《Anthropic 联合创始人：安全为什么不是添头，而是 Claude 性格的来源》 — 作为联合主持

③ 他谈到的

点进去有真内容 —— 本页主要出口

Lenny · Anthropic · OpenAI · Claude · Claude Code · 宪法 AI · 智能体 · 缩放定律 · 对齐 · 超级智能

④ 也在聊「AI 安全」的人