减少认知负荷，选择重要之事

好了，Sam 回来了。最近怎么样？

Alright, the return of Sam. How's it going?

很高兴回来。你怎么样？

Nice to be back. How are things?

挺好的。你知道，YC 下一批会很大。是啊，面试了超过一千家公司。我说 OpenAI 进展很顺利。对此很兴奋。是啊，我准备好让这一年结束了，但这一年很棒。所以我想这一期做点不一样的，就是拆解你的很多文章。我回顾了一下。

Good. This is good. You know, YC is gonna be huge next batch. Yeah, interviewed like more than a thousand companies. I'm saying OpenAI is going really well. Excited about that. Yeah, I'm ready for this year to be winding down, but it's been excellent. So I wanted to do something a little bit different for this episode and just break apart a lot of your essays. I went back.

听起来有趣。我们看看吧。

Sounds fun. We'll see.

好的。我有一个粗略的大纲，有几个你似乎写过好几次的通用主题，我想我们可以一个一个深入。所以从观察你的想法来看，似乎有几个类别：选择做什么，一旦选择了某件事就创造实际价值，然后是一般意义上的目标，比如工作中的目标。我喜欢的第一句话来自你的文章《日子漫长但十年很短》，那句话是：'尽量减少无关紧要的琐事带来的认知负荷。这件事的重要性怎么强调都不为过，而大多数人在这一点上做得有多差也不为过。'

Alright. I have a rough outline with a few general topics that you seem to have written about a few times, and I figure we can kind of just go into each one. So it seems like there are several categories from observing your thoughts: choosing what to work on, creating actual value once you've chosen a thing, and then purpose in general, like purpose in your work. The first quote that I liked was from your essay 'The Days Are Long but the Decades Are Short', and that quote was: 'Minimize your cognitive load from distracting things that don't really matter. It's hard to overstate how important this is and how bad most people are at it.'

是的，我认为很容易花十年时间每天忙得不可开交、压力巨大，感觉自己工作极其努力、制造了大量动静，却没有向前迈进。我认为这是一个大陷阱。很容易陷入那些紧急但不重要的事情，或者很容易陷入办公室政治、玩地位和权力游戏这些无关紧要但感觉很有趣、很重要的琐事。或者只是生活中堆积的其他随机事情。至少我自己发现，我每天有一个固定的认知输出预算，我可以把它花在任何事情上，但如果我把它花在不重要的事情上，我就永远没有时间去做真正重要的事情。而且很容易埋头苦干、专注，却错误地评估自己到底在完成什么。

Yeah, I think it's very easy to spend a decade being incredibly busy and incredibly stressed every day and feeling like you're working incredibly hard and creating a ton of movement, but not moving forward. And I think this is a big trap. It's so easy to get caught up in things that are urgent but unimportant, or it is so easy to get caught up in the trifles of office politics and playing status and power games that don't matter, but feel so fun and so important. Or just random other stuff that piles up in life. I've at least found for myself I have a fixed budget of cognitive output per day, and I can spend that on whatever, but if I let it go on unimportant stuff, then I never have time to get to the really important stuff. And it's so easy to get heads down and focused and sort of mis-evaluate what am I actually accomplishing.

是啊，很多人对我说过：‘老兄，我花了十年时间以为自己做了了不起的工作，因为我感觉自己产生了大量产出，但回过头来看，方向是错的。’这总是令人沮丧。

Yeah, the number of people who have said to me, 'Man, I thought for 10 years I was doing incredible work because I felt like I was creating a huge amount of output, but in retrospect it was in the wrong direction,' is always depressing.

你确实需要专注并做事，但我认为专注于一段时间然后退后一步思考‘我在做正确的事吗？’是非常好的。你也不想犯那种典型的硅谷 22 岁或 21 岁大学辍学生的错误，他们说：‘我要在三个月内解决世界上所有的问题，如果三个月没做到我就失败了，所以我要放弃去做下一件事。’

You do need to focus and do stuff, but I think it's really good to focus for some period of time and then step back and think, 'Am I doing the right thing?' You also don't want to make the mistake that the stereotypical Silicon Valley 22-year-old or 21-year-old college dropout makes of saying, 'I'm gonna solve all the world's problems in three months, and if I haven't done it in three months I failed, so I'm going to give up and do the next thing.'

是的，然后他们跳去另一个项目。那也不好。所以平衡这一点很难，但我认为定期检查以确保你在做重要的事情，并且在合适的时间尺度、时间跨度上工作，非常重要。

Yes, and then they jump onto some other project. That's bad too. So balancing this is hard, but I think making check-ins to make sure you're working on things that matter and working on an appropriate time scale, time horizon, is really important.

是的，因为我有点想广泛地讨论一下，你如何构建这个框架？你需要一定程度的抽离。那么当你需要的时候，你怎么退后一步？我是怎么做的？我在休息时间很自然地就会思考：‘我感觉自己是否在发挥最大作用、做正确的事？或者，YC 是否应该做一些与今天完全不同的事情？我们是否应该愿意抛弃一切重新思考？还是进展顺利？我们需要扩张吗？’但我认为人们有不同的方式。我特别喜欢的一个具体做法，已经成了年度仪式，就是每年 12 月 31 日，我从假期中抽身，坐下来写下过去一年哪些做得好、哪些不好。我看看今年艰巨的待办清单，然后为下一年写一份。因为我知道每年都会这样做，所以其他 364 天我就不那么焦虑了。但我认为一直对我有效的基本模式是：花时间，探索很多东西，尝试很多东西，对什么会成功、什么不会成功保持初学者心态，但相信你的直觉，尽可能廉价快速地追求很多事情，然后对自己诚实，什么做得好、什么不好。困难的部分是砍掉所有不奏效的东西，不断聚焦，直到最终专注于真正有效的一两件事。

Yeah, because I kind of wanted to address broadly how do you even frame that? You have a certain amount of detachment that you need. So how do you step back when you need to? How do I do that? I am pretty good about in my downtime I naturally just think about, 'Do I feel like I am being maximally useful and working on the right things? Or are there like, should YC be doing something totally different than we are today? Should we be willing to throw everything out and rethink about it? Or is it going well? Do we need to scale?' But I think people find different ways to do that. One specific thing I do which has sort of become an annual ritual I really like is December 31st every year, I take a break from vacation and I sit and I write down what went well the last year, what didn't. I look at the hard to-do list for the year and I write one out for the next year. And because I know I'm gonna do that every year, I stress about it less the other 364 days. But I think the fundamental pattern that has always worked for me is: take time, explore a lot of things, try a lot of things, try to have a beginner's mind about what will work and what won't work, but trust your intuitions, pursue a lot of things as cheaply and quickly as possible, and then be very honest with yourself about what's working well and what's not. And the hard part is cut all the stuff that's not working and focus down and down until you're eventually focused on the one thing that's really working or the two things that are really working.

在选择方面，我的意思是这有点过渡到选择做什么以及如何创造价值。但在建立这些指标方面，比如说，你有 YC、OpenAI，然后是你生活的其他部分。你会为每个都建立类似的指标，还是每个都有自己的？

In terms of choosing, I mean this sort of transitions into choosing what to work on and how to create value. But in terms of establishing those metrics for, say, you have YC, OpenAI, and then the rest of your life. Do you create similar metrics for each one, or does each one have its own?

哦，绝对不是。我对那种试图把所有事情放在同一个标准下的系统持怀疑态度。我的意思是，我能说的最好的就是追随我的兴趣。我尝试追求很多看起来有趣的项目。我意识到大多数都会失败，我不在乎。而那些成功的，就真的成功了。比如 OpenAI 真的很成功，YC 真的很成功。过去五年我做了很多其他事情，完全没有成功。

Oh, it's definitely not. I am skeptical of these systems where you try to have everything on the same rubric. I mean, the best thing I can say is I follow my interests. I try to pursue a lot of projects that seem interesting. I realize that most of them will fail and I don't care. And then the ones that work really work. So like OpenAI really works, YC really works. I've done plenty of other things in the last five years that have not worked at all.

我们应该探讨一下，因为我认为这是没人谈论的事情，对大家都不利。比如，去年或前年，你做了很多政治方面的事情。

We should explore that because I think it's something that's not talked about and it's to the detriment of everyone. So for example, last year or the year before, you were doing a lot of political stuff.

我仍然认为那会成功。那不是……让我试着把它放在一个公平的领域里，很多事情同时出错了，但从某种意义上说非常好。好吧，我 26 岁左右卖掉了我的初创公司，然后在收购公司工作了一小段时间，也许 25 岁？我记得。然后我休了一年假。在那一年里，这很难做到，在硅谷真的很难，因为在一个社会地位由你的工作和你在做什么决定的地方，当你出现在一个派对上，有人说：‘哦，你在做什么？’你回答：‘我就是在休假一年。’你可以实时看到他们的眼睛在房间里寻找其他人交谈。感觉不太好。但能够这样做是一件极其有特权的事情。但如果你处于工作之间的间隙，可以休一年假，我强烈推荐。我认为这是我职业生涯中最好的两三件事之一。在那一年里，我读了几十本教科书。我学习了我一直感兴趣的领域。我完全不知道它们会以那样的方式汇聚在一起，但我学到了很多关于核工程的知识，我开始研究 AI，我学到了很多关于合成生物学的知识，我学习了投资。

I still think that is gonna work. That's not... Let me try to frame it in a fair area where a lot of things went wrong at once, but in a way that was really good. Okay, so I sold my startup when I was like 26, something like that, and I worked at the acquiring company for a little while, maybe 25? I remember. And then I took a year off. And in that year, which is hard to do, it's really hard to do in Silicon Valley because in a place where social status is determined by your job and what you're working on, when you show up at a party and someone says, 'Oh, what are you working on?' and you're like, 'I'm kind of just taking the year off,' you can sort of see in real time their eyes look for someone else in the room to talk to. It doesn't feel that good. But it's an incredibly privileged thing to be able to do this. But if you are in the position between jobs where you can take a year off, I highly recommend that. I think it was one of the two or three best career things that I ever did. In that year, I read many dozens of textbooks. I learned about fields that I had been interested in. I didn't have any idea that they were all going to come together in the way they did, but I learned a lot about nuclear engineering, I was starting to work on AI, I learned a lot about synthetic biology, I learned about investing.

所以听起来这段经历是积极、美好的，你过得很愉快。但你说没成功，具体是怎么个没成功？我是说，那些具体的项目？

So you both sound like it was positive and great, and you had a great time. But when you say it didn't work out, how? I mean, like the individual projects?

我投资了一堆公司，它们都失败了。我对几个想创办的公司非常兴奋，但后来因为各种原因发现是坏主意。但这没关系，因为当你处于探索模式时，你会尝试很多项目，并接受大多数不会成功。

I invested in a bunch of companies that failed. I got really excited about a few companies I wanted to start that turned out to be bad ideas for different reasons. But that's fine, because when you're in explore mode, you try a lot of projects and you accept that most won't work.

对。我认识几个经历过这个阶段的人，但他们似乎并不总是这样。也许他们会说，好吧，我要去度假，然后持续几个月，回来见人，或者可能立刻跳进某件事，开始研究任何热门的东西，我认为这是一个单独的问题——你只是在跟风。这通常是个错误。群众有时是对的，但好吧。那么你如何变得果断？因为你说一年，这是一个相对短暂的技巧。你怎么弄清楚该承诺什么以及何时承诺？

Right. So I know a few people that have gone through this, and it seems like they don't always do that. Maybe they say, alright, I'm gonna go on vacation, and that lasts a couple months, then they come back, meeting with people, or maybe they immediately jump into something, start researching whatever's trendy, which I think is a separate problem where you're just following everyone else. That's a mistake usually. The crowd is right sometimes, but okay. Then how do you become decisive? Because you say a year, this is a relatively short trick. How do you figure out what to commit to and when?

我认为这需要对自己极其诚实，判断哪些项目真正有效，哪些你只是说服自己有效。我多次犯过这个错误，将来还会再犯。这是很难训练自己做到的事情。你真的应该退后一步，冷静地审视，问自己：我是在拼命说服自己这有效，还是它真的有效？但我认为这是可以锻炼的能力。显然有人擅长这个。只是很难，因为这意味着承认失败。每次我对一家公司进行重大投资时，我都确信它会成功。你不会抱着亏钱的心态进去。至少你知道，我准备好为我认为会成功的预期价值买单。一旦你锚定了这个想法，就很难回头说，天哪，我错得这么离谱，而这显然经常发生。投资很多公司最好的事情是，你学会了谦卑。你学会了频繁地大错特错，并且你学会了只要成功的规模能覆盖所有失败，那就没问题。

This is where I think it takes brutal honesty with yourself about which projects are actually working and which you would like to convince yourself are working. I've been guilty of this mistake many times and will be many times in the future. This is a very hard thing to train yourself to do. You should really step back and dispassionately look and say, am I desperately trying to convince myself this is working, or is this actually working? But I think it is a muscle that is buildable. Clearly people get good at that. It's just hard because it means admitting failure in something. Every time I make a major investment in a company, I am convinced it's going to work. You don't go in expecting to lose your money. At least you know, I am prepared to pay for the expected value that I think it's going to work. And once you've anchored that, it's hard to go back and say, man, I was so wrong, which obviously happens all the time. The best thing about investing in a lot of companies is you learn to be very humbling. You learn to be majorly wrong very frequently, and you learn that that's okay as long as the magnitude of the successes pays for all of the failures.

我以前没想过这个。假设你卖掉了公司，有了一些钱，而且没想过要做天使投资人。你认为在开始其他事情之前先做一点天使投资是个好主意吗，即使你并没有被那个方向吸引？

I haven't thought about this before. Say you sold your company and had some money and weren't thinking about being an angel investor. Do you think it's actually a good idea to angel invest a little bit before you start something else, even if you're not pulled in that direction?

我认为教会我很多关于商业的事情是扑克和天使投资，所以我推荐这两样。

The things that I think have taught me a lot about business are poker and angel investing, so I recommend both of those.

有道理。顺便快速推荐一下扑克，我大学时玩了很多，相当认真，它不适合所有人，但我强烈推荐它作为一种了解世界、商业、心理学、风险和其他一切的方式。

Fair enough. By the way, as a quick plug for poker, I played a lot in college, fairly seriously, and it's not for everybody, but I strongly recommend it as a way to learn about the world, business, psychology, risk, and everything else.

你玩的是竞技性的吗？

Did you play like competitive?

我在半岛上学，所以我经常去那里。那很棒。那资助了你的初创公司吗？

I went to school on the peninsula, so I would go down there a lot. That's great. Did that fund your startup?

它没有资助我的初创公司，但资助了我作为大学生的生活费。

It did not fund my startup, but it funded my living expenses as a college student.

哇。是啊，我总是对那个人感到惊讶。我是说，能做到这一点的人不是无限的，但足够多，这实际上说得通。我会免费做；我就是太喜欢了。

Wow. Yeah, I'm always surprised by the person. I mean, there aren't infinite people who can do this, but there are enough people that actually makes sense. I would have done it for free; I just loved it so much.

你有没有玩过在线的那种，比如同时十手牌？

Did you ever get into the online stuff, like ten hands at once?

我玩过。那没那么有趣，然后就不怎么玩了。

I did. That was not as fun, and then it kind of stopped.

好吧，明白了。酷。

Okay, all right. Cool.

你最近写了那篇关于生产力的文章。我觉得很有趣。谢谢。你的三个主要建议是睡眠、锻炼和营养，针对身体生产力。那只是整个文章的一小部分，但它是核心，对吧？因为如果没有基础，你就偏离了。

So you wrote the productivity post fairly recently. I thought that was really fun. Thank you. Your three main tips were sleep, exercise, and nutrition for physical productivity. That's like a small fraction of the whole thing, but yes, it's at the core, right? Because if you don't have a foundation, you're off.

有可能我基因上极其不幸，而那些人声称他们每晚只需要睡四个小时，这是真的吗？我怀疑不是。我怀疑不是。但我尽量睡八小时或接近八小时，大多数晚上都能做到。如果你想在真正大的创意想法上获胜，你最好削减其他东西，少一些想法，随便什么。但对我来说，睡眠不足——我在人生的其他阶段尝试过——是一个糟糕的交换。

It is possible that I am incredibly genetically unlucky, and I just—like these people who claim they only need to sleep four hours a night, it's true? I suspect it's not. I suspect it's not. But I try to sleep eight hours or close to it as much as I can, and I get it most nights. If you're trying to win on really big creative ideas, you're much better off cutting something else, having less of them, whatever. But for me, skimping on sleep, which I have tried at other points in my life, is a terrible trade.

我试图优化我认为的最佳时间。我尽量不把我的最佳时间卖给任何其他事情，只用于我想做的项目。我认为每个人都需要一段时间，但每个人都会找到那是什么。对我来说，是清晨，醒来后的头三个小时。我设置了一个巨大的屏障，防止任何东西干扰那段时间。

I've tried to optimize what I think of as my best time. I try not to sell my best time for anything other than the projects I want to work on. I think everyone takes a while, but everyone figures out what that is. For me, it's early mornings, the first three hours after I wake up. I put a huge barrier against anything getting in the way of that.

你认为你每天有多少小时是高效时间？

How many hours do you think you have good hours every day?

它总是在变化，但如果你服用阿德拉或其他东西，你可以有 20 小时。但我不确定那一定是 Sam Altman 的最佳版本。我绝对可以坐下来连续八小时超级高效，不休息——我是说，去洗手间什么的。我认为这比大多数人能做到的要多。但绝对不是 16 小时。我觉得那些声称 16 小时的人，我完成的事情比他们多得多，所以我不相信。也许在我们进化之前，没人能完成 16 小时。

It's always in flux, but if you take Adderall or something, you get 20. But I don't know if that's necessarily the best version of Sam Altman. I can definitely sit down and be super productive for eight hours in a row, no break—I mean, go to the bathroom, whatever. I think that's more than most people can do. But it's definitely not like 16. I think the people who say 16, I feel like I get much more done than they do, so I just don't believe it. Maybe until we evolve, no one can get 16 done.

在你关于选择工作内容的广泛文章中，我很喜欢之前 Hamming 的讲座。他比今天的大多数人更具指导性，我非常喜欢那种语气。你还记得那个关于餐厅的故事吗？

In your broad writing on choosing what to work on, I love the Hamming lecture from a while ago. He's so much more prescriptive than most people are today, which I really enjoy the tone of. Do you remember that story about the lunchroom?

我记得。他说得对。

I do. He's right.

人们不再那样说话了。不是所有人，不是很多人。帮助创建这里的研究机构非常有趣。我确实认为我们做了过去几十年里最有意思的研究实验室之一。世界已经不再支持这种模式了，原因很值得探讨。但我觉得我从 OpenAI 学到的一件事是：如果你坐下来宣布‘我们的研究目标是创造 AGI’，那目标非常遥远且非常困难。你可以做任何你认为有助于实现 AGI 的事情。以前没人这么说过。这很难做到。你怎么管理多个团队的心理，让他们都朝着 AGI 努力但方向不同？比如，也许是机器人，也许是游戏，然后突然有人得到一个结果，整个团队就转向了。

People don't talk like that anymore. Not everyone, not many. It's been very interesting trying to help create a research institution here. And I do think we've done one of the more interesting research labs of the last couple of decades. The world isn't set up to support this anymore, and it's a very interesting question about why. But I think if you take one thing I've learned from OpenAI, it's: if you sit down and say, 'Our research goal is to create AGI,' that's very far off and very hard. You work on anything you want that is gonna help us get there. No one said that before. It's hard to do. And how do you manage the psychology of multiple teams working on things that are all pointing toward AGI but in different directions? You know, like maybe it's robotics, maybe it's gaming, and then all of a sudden someone gets a result that shifts an entire team.

是的。

Yeah.

我认为这其实和我们之前讨论的如何对待生活的算法非常相似。尝试很多事情，让人们跟随直觉，然后当某件事开始见效时，投入更多资源。这对我们来说非常有效。

I think it's actually a very similar algorithm to what we were talking about for how you have to do with your life. Like, try lots of things, let people follow their intuitions, and then when something starts working, put a lot more resources behind it. That's worked for us really well.

嗯哼。

Mm-hmm.

还有一件事，来自‘日子很长，但十年很短’。我喜欢你对风险的解释。你说生活中的事情很少像看起来那么冒险。大多数人过于规避风险，所以大多数建议都过于偏向保守路径。

One other thing from 'The days are long but decades are short.' I liked your risk explanation. You say things in life are rarely as risky as they seem. Most people are too risk-averse, and so most advice is biased too much towards conservative paths.

是的，风险是真实的。风险真正的样子是：当你回顾职业生涯结束时，你会说‘我完全浪费了它’。人们天生不会那样思考风险。我们非常擅长思考短期灾难性风险，非常不擅长思考长期慢性风险。我们担心核电站熔毁，却不担心吸入煤烟之类。所以有极少数人，我对他们的职业生涯感兴趣，因为我认为他们会做出惊人的事情，我花很多时间和他们在一起，没有别的原因，就是想帮助他们。我一直推着他们的是：你想得太小了，你对风险的看法错了。或者你需要转变视角。

Yeah, I mean the risk is real. What risk actually looks like is that you look back at the end of your career and you're like, 'I wasted it totally.' And people aren't programmed to think about risk that way. We're very good at thinking about short-term catastrophic risk and very bad at thinking about long-term chronic risk. We worry about nuclear plant meltdowns but not about inhaling coal smoke or whatever. So there's a very small handful of people that I would say I have taken an interest in their careers because I think they're going to do amazing things in the world, and I try to spend a lot of time with them for no reason other than I want to help them. And consistently, this is the thing I find myself pushing them on: you're thinking too small, you're thinking about risk the wrong way. Or you need a perspective shift.

视角转变。这是我想在某个时候写的东西。我认为随着年龄增长，视角和视角转变变得越来越难，而这是一种难以置信的有价值的东西。我认为这就是为什么人们最终会服用很多迷幻药。我不知道这是不是最好的方法，但保持那种从新鲜视角和新角度看待事物的心理灵活性，坚持这一点非常重要。

Perspective shift. That's something I want to write about at some point. I think as you get older, perspective and a shift in perspective just gets harder and harder, and it's an unbelievably valuable thing to get. I think this is why people end up taking lots of psychedelic drugs. I don't know if that's the best approach, but keeping the sort of mental flexibility to look at things from fresh perspectives and new angles, man is that important to hold on to.

是的。这种情况不常发生。就像失去亲近的人。迷幻药这条路变得非常流行，从阅读正在进行的研究和有限的经验来看，它既让你从日常习惯中抽离，又给你动力去追求那些你犹豫不决的事情。

Yep. It doesn't happen very often. It's like losing someone close. The psychedelic path has become very popular, and just on reading the research happening and limited experience here, it's both abstracting you from your daily habits and then giving you the motivation to pursue things that you were kind of on the fence with.

是的。

Yeah.

我的意思是，我不认为你需要药物来实现这些。我认为你只需要承诺去做，给自己一个空间、一种形式和一些伙伴。另一件我发现对所有真正重要的事情都很重要的是，让自己周围都是那些会让你更有雄心、更好奇、更能转变视角的人。基本上所有人，世界上几乎 98%的人，都会试图把你拉回来，说‘这似乎有点太疯狂、太出格、太雄心勃勃了’。几乎每个人都希望你平庸，因为这符合他们的框架。这样更舒服。

I mean, I don't think you need drugs for any of that. I think you just need to commit to do it and give yourself a space and a format and people to do it with. On the other thing that I found for all of the stuff that really matters, surrounding yourself with people who will make you more ambitious, more inquisitive, shift your perspective more, is really important. And basically all people, almost 98% of people in the world, will try to pull you back and say, 'It seems a little bit too crazy, a little bit too out there, a little bit too ambitious.' Almost everyone wants you to be average because it fits their framework. It's just more comfortable.

完全正确。这很有效。

Totally. It works.

是的。

It works.

你读过的任何关于风险的书是否帮助你理解了风险，还是仅仅通过经验？

Has any book you've ever read on risk helped you understand risk, or is it just through experience?

打扑克是一个好方法，读扑克书也不错。读那些做出惊人成就的人的传记，我觉得很有帮助。我喜欢读历史上那些大型科学和工程项目。我想我永远不会厌倦读关于阿波罗计划的书。

Playing poker is one good way to do it, and reading poker books was good. Reading biographies of people who have done really amazing things, I think is helpful. I love reading about the great big iron science and engineering projects of history. I don't think I'd ever get tired of reading books about the Apollo program.

好的，那么接下来你很多文章的主题是关于创造价值。假设某人选择了他们想做的事情。在你的生产力文章里，你说：‘我的系统有三个关键支柱：确保完成重要的事，不在愚蠢的事情上浪费时间，以及列很多清单。’我觉得你应该说得更委婉一些。你永远无法摆脱你写过的东西。

Okay, so now that the next theme in a lot of your writings is about creating value. Say someone's chosen whatever they might want to work on. In your productivity essay, you say: 'My system has three key pillars: make sure to get the important done, don't waste time on stupid, and make lots of lists.' I feel like you should have hedged it. You never get to get away from the things you've written.

实际上，我认为这很好。我基本上……也许我会用不同的措辞，但我确实认为人们把事情搞得太复杂了。所有这些生产力系统，有多变量图表和三维什么的——那些人似乎从来不是真正推动世界前进的人。我会说，任何系统，只要它能让你保持自律，完成你说要做的事情，就很好。我认为简单的系统真的有效。我要说的一点是，如果你要大量多任务处理，书面清单非常有帮助。但除此之外，我认为任何对你有用的方法都是好的。我认为专注和优先级是关键。

Actually, I think it's good. I do fundamentally... maybe I would have phrased it differently, but I do think people overcomplicate it. All these productivity systems with multi-variable charts and three-dimensional whatever—those people never seem to be the ones who really move the world forward. I will observe that any system, as long as it keeps you disciplined and getting done what you say you're going to get done, is fine. And I think the simple ones really work. The one thing I would say is if you're going to try to multitask a lot, written lists are very helpful. But beyond that, I think whatever you find that works for you is good. And I think focus and prioritization are the keys here.

是的，我认为这是真的。我认为大多数时候，你知道该做什么，也知道自己在混日子而不去做。我发现有一件事对人们有帮助，那就是很多 CEO 会组成小型问责伙伴小组。这在 YC 期间很自然地发生，因为你有同批学员和小组办公时间，然后你失去了它，人们感到生产力下降，所以他们自己重新组建小组。这似乎有效。有个叫 Nick Crocker 的人写了一篇名为‘大象’的文章——我不知道你是否看过——基本上是他和三个朋友每周发邮件，然后每季度检查。非常简单易行。推荐。

Yeah, I think that's true. I think for the most part, you know what to do, and you know you're screwing around and not doing it. One thing that I have found helps people is if they have like a lot of CEOs form these small groups of accountability buddies. It happens very naturally during YC because you have your batch mates and group office hours, and then you lose it and people feel their productivity sag, so they reform the groups by themselves. That seems to work. There was this guy Nick Crocker who wrote this post 'Elephants' — I don't know if you've seen this post — but it was him and three of his friends basically weekly emails and then quarterly check-ins. It's very simple and easy to follow. Would recommend it.

那么你在创造价值中谈到的另一件事是关于这些大项目。还有你写的另一篇文章叫‘创新怎么了？’。我认为这是很多人的恐惧：‘天哪，所有酷的事情都已经发生了’，而实际上它显然正在发生。但你在质疑我们如何激励……

So the other thing you talk about in creating value is about these big projects. And there's this other essay you wrote called 'What Happened to Innovation?' And I think this is a fear that many people have: 'Oh man, all the cool stuff already happened,' when in reality it is obviously happening. But you're questioning how can we incentivize...

我的意思是，如果这件事成功了，比如我们真的造出了 AGI，那它比人类历史上所有创新的总和都重要。显然并非所有创新都完全发生了。人们为什么会有这种感觉，这是个有趣的问题。一是显得有点厌世很酷，一是人们喜欢抱怨，另一个是确实最近大多数创新都在软件领域，比如互联网。我确实认为互联网是人类的一大胜利，那是前沿所在，是最优秀的人想去工作的地方。所以我认为说最近的大创新大多在互联网上，这是一个公平的批评，但这既是批评也是赞美。我认为通常只有少数几个前沿领域，最有才华的人会去，因为那里能取得最大进展。当然现在不是物理学，但这没关系。如果你看看过去十年人类的变化速度，那真是疯狂。互联网不会解决所有问题，但我认为说我们没有进步是错误的。

I mean if this one thing works, like if we do actually build AGI, it's more important than all the innovation combined in human history so far. Clearly not all the innovation happened totally. It's an interesting question of why people feel that way. One is it's cool to seem sort of world-weary, one is people like to complain, another is that it is true that most of the innovation recently has been in software, like in the internet. Now I do think the Internet is a triumph of humanity, and that's where the frontier is, that's where the best people want to go work. So I think it is a fair criticism to say most recent big innovation has been on the Internet, and that is a criticism but it's also a praise. I think there's usually a small handful of frontiers where the most talented people go because that's where you're making the most progress. Sure it's not physics right now, but that's okay. If you just look at the rate of change in humanity in the last decade, it's crazy. The Internet will not solve all problems, but I think it'd be a mistake to say we're not making progress.

那关于融资机制呢？你改变想法了吗？

And what about funding mechanisms? Have you changed your mind?

我认为那篇文章是在 OpenAI 之前写的。我会有更多内容分享，希望是在 2019 年第一季度。我今年的一个大项目就是如何为这些东西融资。我认为对于没有 AGI 那样吸引力的项目，我们接近有答案了。如果它不会在某个时候为世界创造大量价值，我仍然没有答案。但对于那些会为世界创造价值但时间跨度长且不确定的事情，我认为我们会分享一些新结构。好的，敬请期待。

I think that post was pre-OpenAI. I will have a lot more to share here, hopefully in the first quarter of 2019. One of my big projects this year has been how you fund this stuff. I think we're close to an answer on projects without the same appeal as AGI. If it's not going to create a lot of value for the world at some point, I still don't have an answer. But things that will create value for the world but on a long time frame in uncertain ways, I think we'll have some new structures to share. Okay, wait for that.

汉明在他的讲座中提到了一点：他说开着门工作的人会受到各种干扰，但他偶尔也会得到关于世界是什么以及什么可能重要的线索。所以弄清楚你在生活中接受多少随机波动与浪费多少时间之间的平衡是很难的。在某个时候，需要很长时间才能达到这一点。你可以雇几个人帮你筛选东西，这很好，但我很长时间没有这种奢侈。基本上有两件事要做：一是你参加很多随机会议，接受 90%的时间你会浪费时间和对自己感到沮丧，为了那一次让一切值得的会议。但这样做的代价是你必须工作很多，因为你必须在常规工作之外做这些。

There was a point that Hamming brought about in his lecture: he said he who works with the door open gets all kinds of interruptions, but he also occasionally gets clues as to what the world is and what might be important. So figuring out the balance of how much random flux you accept in your life for how much wasted time is hard. At some point, it takes a long time to get here. You get to hire a couple of people who screen stuff for you, and that's great, but I didn't have that luxury for a long time. There are basically two things to do: one is you take a lot of random meetings and you accept that 90% of the time you will waste your time and be frustrated with yourself for wasting your time, for that one meeting that makes it all worth it. But the trade-off for that is you have to work so much because you got to do that in addition to your regular work.

你会分配一个百分比吗？我以前会，但现在我有不错的筛选系统了。我以前每周大概花 10-15 小时在随机事情上，大多数时候都很痛苦，但有时非常有价值。我认为如果你完全切断它，那是灾难性的。如果你没有任何来自你网络之外的人带来新想法或大量未被充分赏识的人才，那真的很糟糕。那是那场讲座的一个关键点：诺贝尔奖得主只研究重要问题，而这使他们远离了这些。所以你不能完全切断，但如果你不愿意非常努力地工作，我认为你必须大幅削减它。

Do you allocate a percentage for that? I used to, but I've got sort of good screen systems now. I used to spend maybe 10-15 hours per week on random stuff, and most of the time it was just painful, and sometimes it's so valuable. I think if you cut it off entirely, it's disastrous. If you don't have any flow of people outside your network with new ideas or just a lot of talent that is underappreciated, it's really bad. That's a key point in that lecture: Nobel Prize winners only work on important problems and it removes them from them. So you can't cut it off totally, but if you're not willing to work really hard, I think you do have to just cut it down a lot.

这有关联。我记得去年黑客马拉松上有个有趣时刻，你在那里。有个团队说，‘我们要做这个，然后做那个，然后我们要研究火箭’，或者帮忙研究火箭。你说了一些我觉得很棒的话：你说，‘如果你想研究火箭，就去研究火箭。’我不相信延迟人生计划。对硅谷人的一个常见批评是，有些人会说类似这样的话：‘我一生的事业是造火箭，所以我要做的是，在接下来三四年里赚一亿美元，通过交易加密货币或运营对冲基金，因为我不想再考虑钱的问题了，然后我要造火箭。’但他们从未做到。我相信如果这些人只选一件事，你知道？我痴迷于钱，所以我要去赚一亿美元，不管花多长时间，但要有足够长的时间跨度，那才合理。或者我现在就去造火箭。我相信他们要么成功要么失败。但延迟人生计划的问题在于，每个人都能看出你在做什么，你对此并不认真，你对两者都不投入，它就行不通。我的意思是它有时肯定管用，但我见过很多次失败。我理解它的诱惑，也理解人们觉得‘我要赚一百万美元然后造火箭，只是筹集资金’的程度。总是这样。有些人倾向于痴迷于赚很多钱。我理解解决基本财务安全的诱惑，当然，我自己也需要。但我认为硅谷的惊人之处在于，你作为一个 22 岁的无名小卒，真的可以筹集资金去创办一家火箭公司。这在 YC 发生过好几次。所以如果那是你想做的，就去做吧。公司会付你薪水，如果你能筹集到资金，你会过得不错。如果你筹不到钱，那就考虑别的事情。我可以长篇大论，只会让人厌烦和生气，但我要说，根据经验，延迟人生计划通常行不通。如果你想做的是有雄心的公司，你在硅谷，你有能力，你很可能能找到办法直接研究那个问题。也许这意味着你必须去 SpaceX 当员工，或者为埃隆·马斯克工作，但至少你在做你想做的事。那很好。另外，如果你真正想要的是赚一亿美元，那就实话实说去赚钱，但要愿意接受这需要很长时间。你不能说你要把它压缩到三四年，这就是为什么它行不通。你不必躲在想造火箭的背后去赚钱。我认为很多人确实想造火箭，但你必须选一个。你不能说，‘我要……’

This is related. I remember it was a funny moment at the hackathon last year when you were there. There was some team who was like, 'We're gonna do this and then we're gonna do that and then we're gonna work on rockets,' or help work on the rockets. And you said something that I thought was so great: you were like, 'If you want to work on rockets, go work on rockets.' I don't believe in the deferred life plan. A common criticism of people in Silicon Valley who I think have great features in their past are people who say some version of the following sentence: 'My life's work is to build rockets, so what I'm going to do is spend, I'm gonna make one hundred million dollars in the next three to four years trading cryptocurrency or running a hedge fund because I don't want to think about the money problem anymore, and then I'm gonna build rockets.' And they never do. I believe if these people would just pick one thing or the other, you know? I'm obsessed with money, so I'm gonna go make a hundred million dollars however long that takes, but with a long enough time horizon, then it's reasonable. Or I'm just gonna build rockets right now. I believe they would succeed at either. But the problem with the deferred life plan is everyone can kind of tell that that's what you're on, and you're not that serious about it, you're not that committed to either, and it just never works. I mean it must work sometimes, but I've seen it fail a lot. I understand the temptation of it, and I understand the degree to which people feel like, 'I'm gonna make a million dollars and then work on rockets and just raise the capital.' It's always like that. There are people that tend to be obsessed with making a lot of money. I understand the temptation to figure out basic financial security, of course, I needed that myself. But I think one of the things that's amazing about Silicon Valley is you actually can raise money as an unknown 22-year-old to go make a rocket company. It's happened several times at YC. So if that's what you want to do, just get on with it. The company will pay you a salary, you'll be fine if you can raise money at all. If you can't raise money, think about something else. I could go on a long rant that will just bore people and piss people off, but I would say the deferred life plan empirically usually does not work. If what you want to work on is an ambitious company and you are in Silicon Valley and you're competent, you very likely can figure out a way to just work on that problem. Maybe it means you have to go be an employee at SpaceX or work for Elon Musk, but at least you're doing what you want. That's fine. Also, if what you really want is a hundred million dollars, just call a spade a spade and make money, but be willing to accept that takes a long time. You can't just say you're gonna compress this into three or four years, and that's why it doesn't work. You don't have to hide behind wanting to make rockets to make money. I think many people actually do want to make the rockets, but you have to pick one. You can't say, 'I'm gonna...'

先做一段时间这件事，然后再去做我真正想做的事。否则你永远无法激励人们加入你。他们不希望看到你成功并与你合作。

Do this thing for a short period of time and then do what I really want to do. Or you'll never motivate people to join you. You don't want you to win and work with you.

是的，就像人们能感知到真诚，能感知到他人的动机。人们能感觉到你是否出于正确的理由做某事。我认为要想在任何事情上成功，重要的是人们支持你，重要的是人们希望你赢：投资者、媒体、你的团队等等。我认为这是延迟人生计划的众多问题之一。

Yeah, it's like people can sense authenticity, people can sense people's motivation. People can sense if you're doing something for the right reasons. And I think to succeed in anything, it's important that people are rooting for you, it's important people want you to win: investors, the press, your team, whatever. And I think that's one of the many problems with the deferred life plan.

是的，无论智商高低，情商如何，人们都有惊人的探测器。每个人都是。再说一次，做任何有价值的事情都需要很长时间和很多情感创伤，很多人会说你是个白痴或你错了。如果你不愿意接受这些，你就不会成功。人们也能感觉到你不愿意接受这些。比如你做加密货币只是为了赚钱，而不是因为你是加密货币的坚定信徒，人们能看出来。

Yeah, people of seemingly any intelligence level and almost any EQ level have amazing detectors. Everyone does. And again, doing anything worthwhile takes a long time and a lot of emotional trauma, a lot of people telling you you're an idiot or wrong. And if you're not willing to sign up for that, you're not gonna succeed. People can also sense when you're not willing to sign up for that. Like if you're doing crypto just to make money, not because you're a die-hard believer in crypto, people can kind of tell.

但即使它伴随着……甚至不是那样，就像随波逐流。即使是为了赚钱，因为如果你做加密货币是为了赚钱，那还算一致。但如果你不是真的对加密货币感兴趣，你会只是……关于它的一些东西……

But even if it comes with... it's not even that, it's just like riding this wave. Even making money, because if you're doing crypto to make money, that is kind of aligned. But if you're not really into the crypto part, would you just... something about it is...

不，我受够了那些副业项目，我花了好几个月维护代码，然后意识到我不知道自己有多在乎这个。但那时很难抽身，你必须意识到这不会有结果。

No, I've had it with side projects where I'm maintaining code for months and months, and then I realize I don't know how much I care about this. But then it's hard to pull out at that point, and you have to realize this is gonna go nowhere.

所以我想谈的最后一件事是工作和职业中的目标感。我认为这有点与自我相关。你写了一篇文章《融合》，你说：‘我们的自我价值如此依赖于我们的智力，以至于我们相信它必须是单一的，而不是在连续谱上略高于所有其他动物。也许人工智能也会有同样的感觉，并注意到我们与倭黑猩猩之间的差异几乎不值一提。’所以这是一个常见的话题：工作、目标、全民基本收入。如果 AGI 时代到来，你认为这会走向何方？

So one of the last things I wanted to talk about is purpose in work and careers in general. I think it's kind of related to ego. You wrote an essay, 'The Merge', and you said: 'Our self-worth is so based on our intelligence that we believe it must be singular and not slightly higher than all the other animals on a continuum. Perhaps the AI will feel the same way and note that the difference between us and bonobos is barely worth discussing.' So this is a common topic: work, purpose, UBI. Where do you think that's going in the AGI era, if it happens?

听着，我认为人们以许多不同的方式找到目标和意义。我唯一关心的是人们能在某个地方找到它。如果是在养育家庭，如果是在经营一支电竞队伍，我认为这和试图构建 AGI 一样有效。我讨厌的只是那些冷漠并以此为傲的人。但我认为关于我们获得生活意义的方式，似乎有一些人类共性：对几乎所有人来说，人际关系、感觉我们在为他人创造价值、感觉别人高度评价我们并尊重我们。有很多方法可以做这些事情。我认为随着 AGI 的到来，会发生的一件事是，智力——我很幸运，我生来相当聪明——将越来越不是一个优势。6 万年前，到处奔跑，智力并没有那么大的优势；体力、耐力重要得多。6 万年以后，人类生物智力将再次不那么有优势。它只是恰好处于一个社会中，在这个时刻它非常有价值。

Look, I think people find purpose and meaning in many different ways. The only thing I care about is that people find it somewhere. If it's in raising a family, if it's in running an eSports team, I think that's as valid as trying to build AGI. The thing I hate is just people who are apathetic and proud of it. But I think there seem to be some human universals about ways we get meaning in life: human relationships for almost everybody, feeling like we're creating value for others, feeling like other people think highly of us and respect us. There are many ways to do those things. And I think one thing that will happen with the arrival of AGI is that intelligence—I got lucky, I was born reasonably smart—will be less and less of a benefit. 60,000 years ago, running around, intelligence wasn't that much of a benefit; physical strength, endurance were much more important. 60,000 years from now, human biological intelligence won't be that much of a benefit again. It just happens to be in a moment in society where it is really valuable.

实际上，我在这里特别发现，智力被如此高度评价，以至于人们似乎奇怪地与自己的身体脱节。所以看到你现在做整个锻炼计划很酷。山姆·奥特曼的肌肉秀。我现在知道你多重了。我想我今年增加了 18 磅肌肉。太疯狂了。而且素食者对此感兴趣？所以你在做重训，对吧？重训。你感觉更好吗？

Actually, I find it in particular out here, like intelligence is so highly rated that people seem oddly disconnected from their bodies. So it's been cool seeing you doing the whole workout plan now. The Sam Altman gun show. I know how much you weigh now. I think I gained 18 pounds of muscle mass this year. That's crazy. And vegetarian to do interested? So you're heavy lifting, right? Heavy lifting. You feel better?

我感觉好多了，比做大量有氧运动好得多。关于这个，我今年转向了举重，我发现首先，你变得更强更快。大量有氧运动几乎没什么用；它只是让你吃得更多。我以前擅长骑自行车，但现在我实际上看到了身体上的进步，这太疯狂了。可悲的是，我增加了这么多体重，这让我更难跑步，而且我今年一直没有跑步。所以我现在甚至不知道自己能不能一口气跑一英里。所以从某种意义上说，我身体状况很糟，但从另一种意义上说，我真的很强壮。它们就是如此不同，截然相反。网上有人说，‘哦，如果你深蹲到 500 磅，你可以骑任何自行车。’这就像，你不懂功率重量比。

I do, much better than doing heavy cardio. The thing with that, I've shifted over to lifting this year, and I've found that first of all, you get way stronger way faster. Heavy cardio does almost nothing; it just makes you eat a lot more. I was good at bike riding, but now I actually see physical progress, which is crazy. The sad thing about it is, I gained all this weight, which makes it harder to run, and I haven't been running this year. So I don't even know if I could run a mile without stopping at this point. So in some sense, I'm in terrible shape, but in another sense, I'm really strong. They are just so different, diametrically opposed. There are people online who say, 'Oh, if you get your squat up to 500, you can bike anything.' And it's like, you don't understand power-to-weight ratio.

那么你会为 2019 年的健身设定一个目标吗？

So are you gonna set a goal for your fitness for 2019?

我会在 12 月 31 日设定。我不知道，问号。我已经达到了今年的举重目标，但无论如何我会坚持到年底。

I will on December 31st. I don't know, question mark. I already hit my lifting goals for this year, but I'll keep going to the end of the year anyway.

不错。恭喜。谢谢大家。严酷。

Nice. Congrats. Thank you, guys. Harsh.

我想最后一件事是，你现在在读什么？

I think the last thing is, what are you reading right now?

我最近读的一本书，很短，我强烈推荐。它叫《爱之道》，作者是安东尼·德·梅洛。你大概一小时就能读完。是一位老朋友推荐给我的。这是一系列关于生活的冥想。读起来非常快，我刚读完，非常喜欢。

The last thing I read, it's really short, I recommend it a lot. It's called 'The Way to Love' by Anthony de Mello. It'll take you like an hour to read. It was recommended to me by an old friend. And it is a series of meditations on life. It is a super fast read, and I just finished it and I really enjoyed it.

太好了。我们会去看看。好的，伙计。谢谢你来。

Excellent. We'll check it out. All right, man. Thanks for coming.

谢谢。

Thank you.