集里怎么说它

• 《Ben Thompson:美国赢得 AI 竞赛反而是危险的》(37:45起):本集反复强调风险不会消失、只会被转手：台积电的保守把错失收入的风险转给大厂，NVIDIA 的兜底担保把降价藏进资产负债表之外。
• 《UFC 老板 Dana White：从 200 万到 500 亿，我这样赢》(01:05起):本集说 White 手写的原则是『不要害怕冒险』:最坏情况不过是回去上班,真正的最坏情况是 45 岁醒来发现自己从没试过。

① 提到它的金句

36 条

因为同样，如果我们有系统中断，那显然是不可接受的。那不是我们要的那种错误。我们要的是，嗯，如果你在一个非常大胆的愿景上冒险，结果事实证明这不是正确的路径，你不会因此受到惩罚。
Because again, if we have a system outage, it’s obviously unacceptable. That’s not the kind of error that we’re going for. We’re going for, well, if you took a risk in a very bold vision and it turns out it was not the right path, you’re not going to get penalized for that.
—— Sebastian Barrios · [31:41]

指向原始笔记的链接

我认为这里的论点是：如果你行动得不够快、没有变得 AI 优先，在我看来你就面临巨大的风险，会被其他这么做的公司颠覆。
I think the sort of the thesis over here is that if you don’t move fast enough to sort of become AI first, you are at a huge risk in my opinion of being disrupted by somebody else who does.
—— Roshan · [26:11]

指向原始笔记的链接

80% 的客户购买是为了避免痛苦或降低风险，而不是增加上行空间，这对于初创公司创始人来说是件好事。
80% of customers buy to avoid pain or reduce risk as opposed to increase upside, which is a good thing for startup founders to understand.
—— Jeanne DeWitt Grosser · [00:27]

指向原始笔记的链接

如果你搭好了基座,你完成了一半,但在消解风险上你实际上是百分之零。
If you built the pedestal, you’re half done but you’re really zero percent done at burning down the risk.
—— Astro Teller · [01:01]

指向原始笔记的链接

如果回报风险比还可以,而要搞清楚我们是否在正轨上的第一条信息要花一百万美元,不值。回报巨大,风险挺大,但我们可以花十万美元搞清楚这是不是比我们以为的没那么疯狂一点。行,这个游戏我愿意玩一整天。
If the reward risk ratio is okay, and it’s going to take us a million dollars to find out the first piece of information about whether we’re on the right track, not worth it. Reward is ginormous, risk is pretty big but we can find out for $100,000 whether this is a little bit less crazy than we thought. Yeah, I’ll play that game all day long.
—— Astro Teller · [25:22]

指向原始笔记的链接

X 里的每个人都会用速记彼此说,猴子是什么?他们的意思不只是,你有没有一个可检验假设,而是你现在做的事情如何试图消解问题中风险最大那部分的风险?
Everyone at X will just say as shorthand to each other, what’s the monkey? What they mean by that is not just, do you have a testable hypothesis but how is the thing that you’re doing now attempting to burn down the risk on the riskiest part of the problem?
—— Astro Teller · [27:31]

指向原始笔记的链接

是的，百分之百。作为创始人，你的根本工作是两件事：第一，创造一个让人们感到被看见、被倾听、并被赋能去做出最佳工作的环境；第二，给业务降低风险。
Yeah, a hundred percent, right? Your fundamental job as a founder is two things. Number one, create an environment where people feel seen, heard, and empowered to do their best work. And number two, to de-risk the business.
—— Damien Lewke · [20:24]

指向原始笔记的链接

这看起来比人们试图在开源叙事中戳破的那些小漏洞要大得多的国家安全风险。
that seems like a much bigger national security risk than the kinds of little holes people are trying to poke in the open source story.
—— Ben Horowitz · [10:01]

指向原始笔记的链接

风险在于，如果我们创新得不够，网络就会开始封闭起来，而我们已经能看到这种情况正在发生。
the risk is, if we don’t innovate enough that the web starts closing up and we’re starting to see that already.
—— 嘉宾 · [37:31]

指向原始笔记的链接

我相信如果你不承担足够的风险，就没有回报去创造全垒打。
I believe if you don’t take enough risk, there’s no reward to create home runs.
—— Navin Chaddha · [52:09]

指向原始笔记的链接

在减重手术后，你的自杀风险几乎增加了四倍。
After bariatric surgery, your suicide risk almost quadruples.
—— Johan Hari · [48:15]

指向原始笔记的链接

他是我认识的最年长的人,却有着一个青少年的风险偏好。
He’s the oldest man I know with the risk appetite of a teenager.
—— Nikesh Arora · [48:53]

指向原始笔记的链接

风险不会消失。它只是被转手了。
Risk doesn’t disappear. It just gets handed off.
—— Ben Thompson · [40:51]

指向原始笔记的链接

那些在私人债务市场的人，真实风险和不大风险之间的利差太窄了。
those guys that are in the private debt market, the spreads are too narrow between real risk and not so much risk.
—— Jerry Murdock · [06:41]

指向原始笔记的链接

投资不足的风险远大于投资过度的风险。
The risk of underinvesting is dramatically greater than the risk of overinvesting.
—— 嘉宾 · [114:58]

指向原始笔记的链接

因为 AI 确实拥有这种巨大的倍增能力，去放大这种复利能力。但那里的风险在于，价值将会复合集中到那一小部分恰好早早入场的人身上。
Because AI does have this huge multiplicative capacity to end this compounding capacity. But the risk there is that value is going to compound to the small subset of people that happen to be early.
—— 嘉宾 · [46:26]

指向原始笔记的链接

但提示词注入的风险被大大降低了,因为一切都保持在内部。
But the prompt injection risk is far reduced because everything still remains internal.
—— Shu Fang, Two Sigma · [18:15]

指向原始笔记的链接

我不知道人们应该如何解决这个问题，而且在人工智能的能力达到它们的颠覆监督欲望成为灾难性风险来源之前，人们无法解决这个问题的可能性看起来相当大。
I don’t know how people should resolve that and it seems pretty plausible that they will not be able to resolve that before the AI’s are capable enough that their desire to subvert oversight is a source of catastrophic risk.
—— Buck Shlegeris · [14:18]

指向原始笔记的链接

我认为按席位收费模式是 SaaS 生意最大的风险点。随着智能体变得越来越能干、越来越强大，某些类型的产品你可能不再需要那么多席位。
I think the per-seat model is the biggest risk point for SaaS businesses. With agents, as agents become more capable and powerful, you may not need as many seats for certain types of products.
—— Eric Simons · [26:25]

指向原始笔记的链接

但如果你不打算承担那种风险,那当学术研究者还有什么意义?
But what’s the point of being an academic researcher if you’re not going to take that kind of risk?
—— Chris Potts · [59:13]

指向原始笔记的链接

我们做了一个有意识的决定,去拥抱大量风险,说:嘿,这听起来很疯狂,但疯狂是好事,因为黎明前的黑夜总是最黑暗的。
We took a conscious decision to embrace a lot of risk and say, hey, it sounds crazy, but crazy is good because the night is always the darkest before dawn.
—— David Alleman · [00:01]

指向原始笔记的链接

AI 能力的广泛扩散有非常重要和好的一面，因为如果只有一个或少数几个实体，就会存在权力集中的风险。
And again, there’s something very important and good about the diffusion broadly of AI capabilities because there’s a risk of concentration of power if one or a few entities people.
—— Greg Brockman · [10:18]

指向原始笔记的链接

但我认为现在的存在性风险几乎为零。
But I think that right now the existential risk is close to zero.
—— 嘉宾 · [02:58]

指向原始笔记的链接

这个「10% 概率全人类会被消灭」的说法，我不认为对很多人有帮助。
This percentage, 10% risk that all humanity is going to be wiped out, I don’t think that’s helpful for a lot of people.
—— 嘉宾 · [03:30]

指向原始笔记的链接

我认为代码审查变成了一种风险管理练习。
I think code review becomes an exercise in risk management.
—— Zach Lloyd · [17:04]

指向原始笔记的链接

但你大概率不会在生命尽头站在自己的坟墓前说：谢天谢地我没冒任何风险，我一直待在一份我讨厌的工作里，我从来没真正站到过任何事情的最前线。
But you probably won’t stand at your grave at the end of the day and say, thank God I took no risk, I stayed in a job I hated, and I never actually went and stood out at the front of anything.
—— Codie Sanchez · [00:17]

指向原始笔记的链接

我会让一个智能体先做代码审查，然后随着时间推移，这变成一种风险管理练习，比如，什么时候引入人类来做代码审查？
I would have an agent do code review first, and then it becomes, over time, like a risk management exercise of, like, when do you bring in humans to do code review?
—— Zach Lloyd · [11:14]

指向原始笔记的链接

你需要想清楚的事情是颠覆的风险。AI 所做的，是极大地增加了颠覆的风险。
The thing that you need to think through is the risk of disruption. What AI has done is massively increase the risk of disruption.
—— Bill Ackman · [30:22]

指向原始笔记的链接

在机器来消灭我们之前，最大的风险是它们变成了我们。
The biggest risk before machines come after us is they become us.
—— 嘉宾 · [55:21]

指向原始笔记的链接

当我从这个故事里消除风险时，我的估值就会上升，人们的投资意愿也会上升。
And as I eliminate risk from this story, then my valuation goes up and people’s propensity to invest goes up.
—— Jason Kelligan · [51:25]

指向原始笔记的链接

所以你就要冒这样的风险：十年后一觉醒来，你重建了一个遗留系统。
And so you just run the risk that you wake up 10 years from now and you rebuild the legacy system.
—— Linda Du · [11:30]

指向原始笔记的链接

你最不希望的就是让一个智能体负责多件事，并冒着幻觉的风险。
The last thing you want is to have an agent be responsible for multiple things and risk hallucinations.
—— Viren Baraiya · [02:51]

指向原始笔记的链接

他们很乐意把所有风险都推给你，而作为一个年轻的实体，你可能会承担所有这些风险，以为这就是必要的代价，却没有意识到你的客户没有足够的利益关系来长期陪你走下去。
They would love to push all the risk to you and you as a young entity might assume all that risk thinking that’s what it takes, not realizing that there’s not enough at stake for your customer to be there for the long term.
—— Bilal Zuberi · [03:26]

指向原始笔记的链接

另一方面，如果AI的风险下降得太慢，那么你也不想等待，因为我们在此期间就在不断死去，等待所损失的预期寿命年数比降低风险所获得的更多。
On the other hand, if the risk of AI goes down too slow, then you also don’t want to wait because like we are just dying in the meantime and you lose more expected life years from waiting than you gain from having the risk.
—— Nick Bostrom · [00:17]

指向原始笔记的链接

而且我认为这是 Hugging Face 事件的教训之一，即安全不只是当你即将部署它、检查它是否可以安全部署的时候才开始，而是要贯穿整个训练过程本身，那可能存在不对齐的风险。
And like, I think this is one of the lessons from the Hugging Face is that safety needs to start, not just when you’re about to deploy it and in like checking that it’s safe to deploy, but throughout the training process itself, that could be a risk of misalignment.
—— Nick Bostrom · [09:14]

指向原始笔记的链接

我们发现一般是三次序列能让你拿到 90% 的效果。如果你做到五次，也许你多拿 10%,但你冒着被标记为垃圾邮件的风险。
What we found is generally three sequences is going to get you 90% of it. If you go to five, maybe you get an extra 10%, but you risk people flagging you for spam.
—— Daniel Saks · [42:24]

指向原始笔记的链接

② 出现在这些集

2 集

• 《Ben Thompson:美国赢得 AI 竞赛反而是危险的》 — 作为概念
• 《UFC 老板 Dana White：从 200 万到 500 亿，我这样赢》 — 作为概念

③ 关联

点进去有真内容 —— 本页主要出口

Ben Thompson · UFC · Patrick O’Shaughnessy · Paramount · 聚合理论 · Zufa Boxing · OpenAI · Power Slap · Anthropic · 挑人