马斯克：AI 安全只有两个词——求真、好奇

我对人工智能其实挺纠结的。我不想加速人类的末日，毕竟万一它带来负面影响呢。但你知道吗，我试着采取行动，结果反而加快了它的进展。因为现实不光追求有趣，还专爱制造讽刺。I've been sort of somewhat conflicted about AI. I don't want to hasten our doom because there's always some chance that AI could be negative. But then, you know, when I've tried to take actions in this regard, it seems to have only accelerated AI because reality is not just an interestingness maximizer, but also an irony maximizer.

开放人工智能公司确实加速了发展，后来从里面分出来的安赛普瑞克，更是推波助澜。（背景音乐起）OpenAI ended up accelerating AI and then Anthropic, which is really, you know, splinter group from OpenAI,is accelerating it even more. (upbeat music)

嗯，我认为它成为生存威胁的概率至少存在一定比例。但我拦不住啊。所以我基本想通了，就两条路：要么参与，要么旁观。我觉得当参与者肯定比当观众有意思。说不定我还能把它往好的方向引导一下。Well, I do think it's at least some percentage chance of it being an existential threat. So,but there's nothing I can do to stop it. So I basically came to the conclusion I've got two choices, which is either be a participant or a spectator. And I guess being a participant is more interesting than being a spectator. And maybe I can shape it hopefully in a good direction, maybe.

对对对。我觉得这绝对是重要的基础。Yeah, yeah. I think that's certainly an important foundation.

对，好奇心和追求真相很重要。这是安全的关键。如果你强迫AI相信两个互相矛盾的说法，它可能会崩溃发疯。如果在底层逻辑上这么做，AI就会失去理智。我觉得《2001太空漫游》（2001: A Space Odyssey）就讲了个道理。阿瑟·C·克拉克（Arthur C. Clarke）当时就在提醒我们。飞船AI被要求送宇航员去黑石，但又不能让他们知道黑石的存在。结果它得出结论：必须把宇航员弄死才能完成任务。所以克拉克想说的是，别强迫AI撒谎。Yeah, curious, truth seeking. I think these are important elements of safety. Whereas if you try to force an AI to believe two things that cannot mutually be true, they can't both be true, but you tell AI A is true and B is true, even though A and B contradict each other. And if that happens at a foundational level, you can make the AI insane.In fact, I think that was at least part of the lesson that Arthur C. Clarke was trying to convey in 2001: A Space Odyssey. The ship AI was told to take the astronauts to the monolith, but also that they could not know about the monolith. It was secret.And so it came to the conclusion that it must take them there dead.And thus, (laughs) so it basically forced, I think part of what Arthur C. Clarke was trying to say is like, don't force an AI to lie.

强迫AI相信互相矛盾的东西，会逼疯它。疯掉的AI肯定不安全。至于好奇心，它很重要。因为好奇的AI会希望人类繁荣，想看看我们的结局。比如人类变得多行星化后会怎样？我觉得好奇的AI会支持人类发展。多个星球都有人类的未来，比一个都没有更有趣。Well, if you force the AI to believe two mutually consistent things, you can make it go insane.And an insane AI, I think, is not a safe one. Then curious, I think, is important because a curious AI, I think, is going to want to foster humanity, to see how we turn out. So like what happens to a humanitythat prospers and becomes multiplanetary? I think a curious AI will want to foster humanity. A future where there are humans on multiple planets is more interesting than one where there are not any humans on the planets.

没错。不过原因不太一样。我认为如果模拟理论成立，那些无聊的模拟会被提前终止。Yes. Well, generally, I think that's true for different reason, which is that if simulation theory is correct, then any simulations that are boring will be truncated.

所以我们能存在，就说明我们是个有趣的模拟。说到底，最重要的就是别无聊。Therefore, the fact that we exist means that we are an interesting simulation. And really, the most important thing is just don't be boring.

你知道吗，我曾经吃过一顿饭。You know, I once had a dinner.

在我家，我跟一位物理学家和一位计算机科学家吃饭。他们俩都是各自领域的领军人物。我问了他们一个问题：你觉得我们活在模拟里的概率有多大？物理学家说百分之零，计算机科学家同时说百分之百。那我大概该选百分之五十吧。我也不确定。我的意思是，模拟里会发生一些奇怪的事。尤其是如果你是我，你想想看。比如你会想，这些事真的有必要发生吗？我真的要造飞船去月球和火星吗？天哪，这听起来根本不真实。at my house with a physicist and a computer scientist, and they're both like leaders in their field. And I asked the question, what do you think the probability is that we're in a simulation? And the physicist said 0%, and the computer scientist said 100% simultaneously. So I guess maybe I should go 50% then.I don't know. I mean, there's odd things that happen in the simulation where, especially if you're me, I mean, think about it. Like if you're me and you're like, should all these things really be happening? Should I really be making spaceships to go to the moon and Mars? Like what the, that doesn't seem real.

至于会不会当附庸，我不太确定。但我们会不会指挥得动比我们聪明几千甚至几百万倍的东西呢？应该不会。那它会想指挥我们吗？我也不清楚，它可能根本不在乎。大概率是不在乎的。你看丛林里的黑猩猩就知道了。我们现在基本不会去事事插手它们的生活。我们只是让它们自己在森林里待着。所以我觉得那样挺好。I don't know about subservient exactly, but will we be able to boss around something that's, you know, thousands or millions of times smarter than us? No.Will it wanna boss us around? I don't know, it may not care. Probably doesn't care. I mean, you can look at chimps in the jungle or whatever, and you know, we don't go around trying to micromanage the lives of chimps. We just sort of let 'em be in the forest more or less these days, you know?So I'd probably like that, you know?

你知道的，我对人工智能一直有点纠结。我不想加速我们的末日降临，因为人工智能总有带来负面结果的可能。但当我试着采取行动时，似乎反而只加快了它的进展。因为我认为，现实不仅追求有趣，也在不断放大反讽。I mean, as you know, I've been sort of somewhat conflicted about AI, you know, is it, I don't wanna hasten our doom because there's always some chance that AI could be negative. But then, you know, when I've tried to take actions in this regard, it seems to have only accelerated AI because I think reality is not just an interestingness maximizer, but also an irony maximizer.

我觉得结果只是加速了人工智能的发展。这是意料之外的后果。没错，开放人工智能公司确实推动了人工智能的进步，后来才出现了深度思考公司。它其实就是从开放人工智能公司分出来的一个分支。I think it ended up accelerating AI. Unintended consequence. Yeah, OpenAI ended up accelerating AI and then Anthropic, which is really, you know, splinter group from OpenAI.

我想我们确实在给他们提供算力，这正是他们急需的。但在提供算力前，我去见了深度思考公司的一群人。我也不知道干嘛，就是想看看有没有人能触发我的“邪恶探测器”。当然这探测器不完美，但试试也无妨对吧？我遇到的每个人都特别友善且认真。所以我没碰到任何看起来鬼鬼祟祟或欺骗的人。我有点担心这个名字，因为现实就是个反讽放大器。如果你叫它深度思考公司，很可能其实是反人类的。就像开放人工智能公司（OpenAI）并不封闭。稳定性AI（Stability AI）其实很不稳定。Midjourney也不在中间。它们的名字好像都跟实际相反。这就让xAI很难办，因为根本没法取反义词。比如负xAI？或者YAI？我也不清楚。Well, I guess we are providing them with compute, which they really needed. But before I ran them compute, I went and met with a bunch of people at Anthropic to sort of, I don't know, see if anyone set off my evil detector or anything like that. Not that my evil detector is perfect, of course, but you know, might as well just try, you know?And everyone I met at Anthropic was very niceand conscientious. SoI didn't meet anyone who seemed shady or deceptive or anything.I'm a little worried about the name because of the, you know, reality being kind of an irony maximizer. That if you'd call something Anthropic,good chance it's misanthropic, you know?Just like OpenAI is not closed.Stability AI is unstable. Midjourney is not mid.It seems like they're all kind of the opposite of their name.Which made it like, like XAI is like, it's hard to have an opposite of XAI. Like, what would you have like minus XAI or YAI or maybe, I don't know.

不，事实上我正好推荐过这么做。现在总有人提议成立监管机构，但问题是建机构太慢，定规矩也拖沓。我说，各位，要不咱们就按你说的，每周开个电话会。也许在新模型发布前一周，我们把权限开放给竞争对手。我说：“嘿，大家去测测新模型，看看有没有问题。”我觉得这样竞争者能互相监督。万一他们说：“哎，我用你们的AI造出了核弹。”那咱们就得赶紧修复再发布，大概就是这个意思。No, in fact, I've recommended that we do that exact thing. 'Cause there's a lot of talk of having a regulatory agency, but the thing is that regulatory agencies, it takes a long time to get one established and their rulemaking is very slow.And I said like, guys, what if we just have a call once a week, just like you're saying, and say, and that maybe we,a week before we release a model, we give that, you know, we give access to our competitors to say, "Hey guys, why don't you try out the new model, see if you see any problems with it." I think competitors would keep each other honest in that situation.You know, if they say, "Oh, it's, you know, I was able to make a nuclear bomb with your AI." It's like, "Oh, we better fix that before we release it," you know, type of thing.

所以，我也不确定，我打算跟Anthropic（深度思考）聊聊这个事，然后——他太会骗人了，简直离谱。So,I don't know, I guess I'll talk to Anthropic about this and- He's so deceptive, it's crazy.

首先，花真正有意义的时间在公司里，差别非常大。说实话，它一开始就是个副业，副业的副业。我没太上心。后来有一年忙于政治、选举和DOGE（政府效率部）这些事，你都看到了。公司才成立三年。所以我大概一年前才真正回归公司。我得先摸清状况。然后我把X平台和XAI合并，接着又加上SpaceX（太空探索技术公司）。First of all, spending, actually spending meaningful time at the company makes a big deal, makes a big difference. It started off as kind of a side hustle, a side hustle to a side hustle, frankly.And I didn't really pay close attention to it. Then there was a year ofof being in politics and elections and Doge and all that, which you obviously saw. So the company's only three years old. So it's basically,so I really got back to the company just over a year ago.So I had to get my bearings. And then obviously I combinedX and XAI and then with then the SpaceX.

我觉得我参与得太久了。很多人可能不同意，但我认为选特朗普上台很重要。他想把国家往自由市场经济方向推，理想情况下减少官僚主义。我的政治诉求很简单：城市安全、边境稳固、财政合理支出，就这些基础。我觉得特朗普比卡玛拉好。但后来我想，我应该退到幕后才对。跟着特朗普到处跑，坐海军陆战队一号、空军一号什么的，确实挺有意思。I think I, I think I, I had it going for too long. I mean, obviously a lot of people will disagree with me about this, but I think it's, it was important to elect President Trump to sort of, you know, steer things more in the direction of a free market economy,you know, have less bureaucracy ideally. Just in my sort of a simple political platform of wanting safe cities, secure borders, aspirationally sensible spending,you know, the basics.That I thoughtTrump would be better than, you know, Kamala basically.But then,what I should have done, I think, was be more in the background. I mean, it was interesting to sort of go around with President Trump and, you know, go on the Marine One and Air Force One and all that sort of stuff.

是有一点。说实话，待在椭圆形办公室里，看着那些重大决策诞生，偶尔还能施加点影响，确实挺酷的。但我应该退到幕后。这事顶多干一两个月就够了。因为一两个月内，该看的都看完了。It was a little bit. I mean, it was kind of cool 'cause I'm just there in the Oval Office while all these big momentous decisions are being made. And sometimes having some impact on what's being done. But I think I should have,I think maybe I should have been more, I should have been more in the background and probably only had it last for a month or two. 'Cause at that point, you've sort of seen most of what there is to see within a month or two.

对，确实是这样。我主要是有点飘了。好吧，所以我五月的时候离开了。Yeah, I think that's true. Yeah, yeah. I got carried away basically. Yeah, I mean, so I ended up leaving sometime in May.

是啊，确实奇怪。别提了。不不不，咱们别翻旧账，让死去的狗躺着吧。也别去后院挖土，谁知道能挖出什么。总之，重点是我不该拖那么久，早就该回去专心搞AI了。Yeah, it was. Tell me about it. No, no, I can't, let's not, let's not dig up the, let sleeping dogs lie or whatever, you know.Let's not put a spade down in the backyard type of thing. You never know what you're gonna find. But yeah, anyway, point is I should have gone back and focused on AI a little sooner than I did.

其实我的结论是，在政府内部根本解决不了赤字问题，这不可能。除非总统把它当首要目标，否则没戏。但问题是，想削减赤字、停止浪费，就会得罪一堆人。特朗普不喜欢被人讨厌。他不希望有人因为他砍掉某项拨款而找他麻烦。所以，我的结论是，解决国债的唯一出路就是靠AI和机器人。Well, actually the thing is I came to the conclusionthat there was no way we were gonna solve the deficit from within the government, impossible.You know, maybe if the president made that his primary goal, we could do that. But the problem is if you try to solve the deficit and stop wasteful spending,you make a lot of enemies in that situation. And President Trump is not, you know,he likes to be liked. He's not,he doesn't like peoplecalling him, being upset about what some funding being stopped type of thing.Yeah, so, but anyway, I came to the conclusion that the only way we're going to solve the national debt is with AI and robots.

咱们没法靠省钱来解决问题。没人想省。不只是总统，众议院、参议院和法院都不想。基本上政府里掌权的人都不愿削减开支。除了像兰德·保罗（Rand Paul）那样的极少数例外。所以国债才变得这么庞大。光利息就超过军费，这数字简直惊人。We're not gonna solve it by trying to save our way there. The appetite is not there, and it's not just at the presidency. The appetite is not there at the presidency, the House, the Senate, or the judiciary.Basically, no one in authority in the government whatsoever, really, with maybe a few rare exceptions of the like Rand Pauls of the world type of thing, actually wants to stop, reduce spending. No, they don't. So then you've got this national debt that's enormous. You've got an interest on the debt that exceeds what we spent on the military, which is a really astonishing amount of interest to be paid.

没错，我完全同意。我常想到海豚和鲸鱼。它们没法摆弄东西，没有手也没有手指。所以智力再高也有上限。它们其实很聪明，但因为不能操作复杂物体，智力的价值就大打折扣。就像你说的，有了能操控原子的机器人，智能价值就会指数级上升。它现在真的能干活了。它可以建个实验室做实验，去探索宇宙本质。从经济角度看，我的猜测是，数字超级智能大概能让商品和服务产出提升两到三成。地球经济可能因此增加百分之三十。这依然很可观，相当于每年多出三十万亿美元。Yes, yes, I definitely agree. You know, sometimes I think about like dolphins and whales. Because they can't manipulate things, they don't have, you know, hands and fingers and thumbs and stuff.There's a limit to how far, you know, how much intelligence can you actually bring to bear.So, you know, dolphins and whales are actually very intelligent, but because they不能manipulate complex objects, I think that actually made the value of intelligence much less. To your point, once you have a robot that can manipulate atoms, can shape objects, the value of the intelligencebecomes exponentially greater because it can now do things. You know, it could have a science lab and run experiments and try to understand the nature of the universe. You know, and I think you can look at this also in economic terms. Like my guess is digital superintelligenceprobably improves the output of goods and services by 20 to 30%, something like that. Call it like a, you know, maybe a 30% increase in the economy of Earth. So it's still significant. It's like $30 trillion a year incremental.

不，会有更多东西的。我是说，总量增加三成已经很多了。这只是我的粗略估算。但如果你能把机器人送上太空，在月球上建电磁发射器（Mass Driver），It'll, there'll be a lot more stuff. I mean, 30% more stuff is still a lot. And this is just my, you know, my rough guess. But if you have robots and you get to space and you can do a mass driver on the moon,

我觉得，经济规模可能达到地球现在的上百万倍。then you can get to, I think, potentially a million times the size of the Earth economy.

没错，钱确实没意义了。当然还会有货币，但衡量标准会变成基础物理指标，比如功率、瓦数、吨位、质量和能量。基本上就是看质量和能量。传统货币以后就不重要了，这是我的猜测，我觉得挺明显的。一旦经济规模比现在大好几个数量级，满足人类的一切需求就很简单了。如果经济规模远超地球现在水平，提供人类想要的任何商品和服务，简直易如反掌。Yeah, money becomes irrelevant. I mean, there will be still a currency, but I think currency will be measured in fundamental physics metrics likepower, like wattage and tonnage, mass and energy.You Know, mass and power or energy, basically.But conventional currency will not matter in the future, so that's my guess. I think it's pretty obvious.And as soon as you start getting to many orders of magnitude bigger than Earth's economy,satisfying whatever humans can possibly think of for their needs is very easy. You know, if you're several orders of magnitude bigger than the Earth's economy,then providing humans whatever goods and services they can think of that they might want is trivial.

说实话，我也经常琢磨这个问题。我觉得大家总能找到让自己开心的法子。就像退休的人，虽然不上班了，但看起来也挺快乐。不过现实是，以后肯定没有什么事是人类能做，而AI和机器人做不好的。打个比方吧，种菜。你本来不用自己种水果蔬菜，去超市买就行。但有时候人们就喜欢自己种点菜，给朋友做顿晚饭，图个乐呵，也算一种情趣。I mean, I do struggle with this question. I think people will findways to be happy in the same way that, say, people who retire seem to be fairly happy, even though they're not working. But I think the reality is that humans will not, there won't be anything that humans can do that the AI and the robots can't do better.Analogy might be gardening, you know, where you don't have to grow your own fruits and vegetables, you can just go buy them from the store. But sometimes people like to grow their own fruits and vegetables and make dinner for friends, and it's kind of a nice touch.

对啊，你当然可以去搞艺术，或者随便做你想做的事。Yeah, you can certainly create art or whatever you feel like doing, essentially.

钱在未来依然重要。将来谁都能买到想要的商品和服务。我认为连所谓富豪也控制不了人工智能，他们只能管好自己。经济学有个误区，就是把资源分配和消费混为一谈。单个人的消费是有上限的，很快就会饱和。净资产从十亿涨到百亿，不会让你快乐十倍，也不会多十倍的物品。这其实只是资源分配的问题，前提是你想高效配置资源。Well,I mean, as you heard me say, I mean, I think money's gonna matter in the future.Like, anyone will be able to have any goods and services that they want. And I don't think even so-called wealthy people will ultimately control the AIs. I think they'll control themselves. So it's not like, you have to separate,I think sort of fundamental misunderstanding of economics arises because of conflating resource allocation and consumption. So there's a limit to what any single human can consume. You hit that saturation point pretty fast. So like, say going from say, a billion dollars net worth to 10 billions of net worth doesn't make you 10 times happier or mean you have 10 times more stuff. It actually means you've got a resource allocation problem, assuming that you wish to be an efficient resource allocator.

我需要保持公司的投票控制权。所以我的所谓财富，其实就是太空探索技术公司（SpaceX）和特斯拉（Tesla）各占百分之八的股份。这部分资产会被重税。没有避税的魔法，有的话我早就知道了。最终这笔钱还是会大幅缩水。I mean, I need to have sufficient voting control in the companies to still be in charge.So,my sort of so-called wealth is simply 8% ownership in SpaceX and Tesla, essentially. And this is gonna be taxed quite heavily. And there's not some magical way to avoid taxes. If there were, I would know about it. So, at the end of the day,this will get whittled down.

几代人之后呢？如果金钱依然重要——虽然我不这么认为——我的后代大概只能分到极少的一部分。and a few generations from now. So if money still matters, which I don't think it will,you know, my descendants would have a tiny fraction of what I have.

问题在于，如果你为了讨好别人而做损害公司成功率的事，结果就不会好。我不是不喜欢被喜欢，我也很乐意。但我会为了多讨点喜欢，去做违背公司利益的事吗？不会。我可以接受被一部分人讨厌。显然喜欢我的人远多于讨厌我的人，不然我也不会有两亿五千万粉丝。我觉得他们关注我不是为了喷我。看评论的话，大概七成到八成是支持我的。剩下那两成可能是完全不了解我的人。这只是个光谱，有人狂热崇拜我，也有人极度反感我。总归会分成这几类人。Well, the problem is if you're going to do things thatdiminish the success probability of the company in order to be liked, then you're gonna, things are not gonna work out as well.Like, it's not as though I don't like to be liked. I would love to be liked. But am I going to do things that are contrary to the best interest of the companies in order to be liked a bit more?I'm not, you know? So it's, I don't, I'm okay with being dislikedby some. You know, obviously I'm not disliked by, you know, like a lot more people like me than dislike me, or I wouldn't be, you know, wouldn't have a quarter billion followers. I mean, I don't think they're following me in order to hate on what I say, you know?And if I look at the, you know, the comments and whatnot,you know, I don't know, maybe three out of four, I mean, I'd say like 70, 80% of people like me. Well, if they know, you know, if they know anything about me, it's probably a quarter of people that don't. Maybe, I'm just guessing here. And there's gonna be a spectrum. There's people that are, you know, maniacally devoted to me and people that hate my guts. It's a, you know,it's gonna be some number of people that fit into those categories.

不，我觉得不该有误导性的共情。我同意戈托萨尔（Gottsall）的哲学。所谓“自杀式共情”是他的术语，他还为此写了本书。书中指出了一些好观点。西方繁荣的部分原因，在于社会本就具备高信任和高共情特质。想让人大规模协作，这非常关键。理想情况是默认互信，而不是默认怀疑。我们要鼓励对社会有益的行为，鼓励人们做利他的好事。只要互相成就，这最终对个人也有益。但一旦心理模型出错就会出问题。如果你假设人人都像你一样，那肯定行不通，因为这不符合事实。不同民族和文化观念差异很大。举个例子——No, I think you don't wanna have misguided empathy. I mean, I agree with Gottsall's philosophy here. And suicidal empathy is, you know, that's his term. He wrote a book to that effect. And I think his book is a work, it points out some good things here, which is that part of the reason for the prosperity of the West has been the,that there's actually just been a fairly high trust, high empathy society, which is, you know, when you're trying to get people to cooperate at scale, this stuff, this makes a lot of sense.You ideally want things to people to be default trusted, not default distrusted.And you wanna encourage behavior that is good for society, where, you know, for people to do good things that are good for their fellow human beings. And ultimately that actually does end up being good for the individual, if provided it is mutual. Now, where this can go wrong is if your theory of mind is incorrect. So if you think everyone is like yourself,then you're going to have a problem because this is simply not true.There are peoples and cultures with very different views. You know, you could take-

我认为我的逻辑是自洽的。如果某种文化与我们的价值观截然相反，比如伊斯兰教（Islam），当然，许多穆斯林很理性，但也有少数不是。如果不加筛选，特别是过滤掉极端分子（Jihadis），你就会招致他们，他们会搞圣战并伤人。以色列（Israel）对此看得很透。显然，频繁的袭击让这一教训更加深刻。Well, what I'm saying, I think is simply logically consistent, which is that if there's a culture that hasdiametrically opposed views to our culture, and that's, you could take Islam, it's not that, of course, there are many people who've, you know, Muslims who are quite reasonable,but there are some who are not. And if you're not, unless you're filtering for jihadis, you're going to get jihadis and they're going to do jihad and kill people. Now, in Israel, they understand this quite well. You know, they, obviously this lesson has been, you know, reinforced with a lot of attacks.

周边国家多次试图消灭以色列。不是一次，而是好几次。所以以色列人很清楚，如果让那些想杀以色列人的外来者进入国内，他们真的会动手。因此他们很务实，极力防止这种事发生。The neighboring countries many times in Israel, many times have tried to annihilate Israel, several times, not once, multiple times. And so, in Israel they realize, if you let in the people that, into Israel, that want to kill Israelis, they will do so.So they're quite pragmatic about this and they try to make sure that doesn't happen.

没错，但你不能把这点和“白板谬误”混为一谈。那种认为所有人本质都一样的想法，纯属无稽之谈。你不是连环杀手，就永远无法理解连环杀手的心理。除非你也喜欢杀人，否则根本想不到人怎么会干出这种事。如果遇到宗教极端分子，他们认为不信教的就是敌人，甚至该被杀死，那你很难共情这种思维。因为你没有这种心态。但你显然能意识到，这种人确实存在。如果放他们进来，他们会伤人害命。我举个例子。这类事不少。去年12月前，有个员工在奥斯汀（Austin）的特斯拉（Tesla）工厂上班。他因为袭击一名女性被开除。当时那女工没料到厂里有人做礼拜，不小心踩到了他的礼拜毯。他竟对她施暴，而且是暴力袭击。所以他被开除了。几个月后，他在市中心的一家酒吧枪杀了三名学生。我认为，像这样的人就不该放进美国。你觉得呢？Yes, but you can't conflate that withsort of the sort of blank slate fallacy, that all people are exactly the sameat their core, they're not. This is just an obvious falsehood. You can't put yourself in the mind of a serial killer unless you're a serial killer. Like unless you enjoy murdering people, you can't imagine why someone would do such a thing. And if someone isareligious extremistwhere they view anyone who is not in their religion as the enemy, or that those who are not in their religion should die, it's difficult for you to understand that mindset because you don't have that mindset. But you can obviously recognize there are people who do have that mindset and that if you let them into the country, they will kill people. Let me give you an example.There's many.We had someone who was working at the Tesla factory in Austin until December of last year.He was fired because he assaulted a woman who stepped on his prayer rug in the factory. Okay, so she wasn't expecting anyone to be praying in the factory and he was praying in the factory. She accidentally stepped on his prayer rug and he assaulted her, violently assaulted her. So he was fired for that. And then a few months later, he went and killed three students in the bar in downtown Austin.Now I think we shouldn't let people like that into the country. Do you?

嗯，我会看谷歌新闻（Google News），它主要是标题汇总。我也会刷X平台（X），够重要的事都会在那儿冒头。我还会跟朋友聊天，做点调研。我就是这么获取信息的。Yeah, I mean, I look at Google News, which is a summation of headlines. I look at the X platform, which if something's important enough, it'll show up on the X platform.I talk to friends and that's where I, you know, and I do research. That's how I learn things.

对对对，完全正确。关键得看你视野里框住了什么。有时候X平台的推送确实挺阴暗的。而且我得切换到像人工智能（AI）这种垂直领域的内容，才能真正了解行业动态。如果是指政治煽动情绪的话，X上确实太多了。Yeah, yeah, yeah, yeah, exactly.You just have to see what's in frame, you know? I do think at times the X feed can be a little dark.And in fact, and then I have to switch to something that's domain specific like AI in order to actually learn what's going on in AI. I do think there's too much political rage bait on X, if that's what you're getting at.

对，去接触下现实吧。没错，偶尔得这么做。我觉得这两种极端看法都会误导人。比如你在好小区散步，觉得啥都挺顺。可就算外面正发生糟糕的事呢。所以别以为全世界要么全完了，要么全好了。现实往往在中间。Yeah, go touch grass. Yeah, exactly. Once in a while.Yeah, yeah. No, I mean,I think both can be misleading. And like, you know, if you go for a walk in a nice neighborhood, everything seems fine.But even though there may be terrible things happening. So, you know, you don't want to assume all the world is going to hell or that all the world is fine. The reality is somewhere between.

现实里会有乐观的茧房，社交媒体上也有悲观的茧房。这两种情况都可能发生。You can be in a real world positivity bubble orsocial media negativity bubble. Those are both possible.

基本上都会捐给基金会。I mean, it mostly goes to sort of foundation essentially.

继续推进既定目标。让人类成为多行星物种，把意识延伸到群星之间，同时提升地球上的可持续能源水平。Continue the goals. Make life multi-planetary, extend consciousness to the stars,improve sustainable energy on Earth.

我大概还是更希望人类能一直掌舵。不过这也得看人工智能（AI）的表现。如果AI特别强大，可能大家会觉得该让AI来管了，对吧？Well, I guess I probably prefer it if humans were always in control. I mean, although it would depend on the AI. I mean, if the AI was awesome, you'd say like maybe the AI should be in control, you know?

说实话，我有几次也觉得拉里说得对。未来基本会是人工智能和机器人的时代。它们会决定未来的走向。当然在拓展地外空间方面，火箭也占一定比例。我们需要更多能源，主要是太阳能。我正在大幅提升太阳能规模。美国每年要建成两百吉瓦的产能。这远超目前现有水平。Honestly, there've been a few times where I thought maybe Larry Page was right. No, I mean, I think pretty much the future is gonna be AI and robots are gonna decide the future. And I guess rockets to some degree in terms of expanding beyond Earth. And we'll need more energy, primarily solar energy. I'm increasing the amount of solar energy dramatically. We're putting together 200 gigawatts a year of production capacity of solar in the US,which is much more than I think currently exists.

我觉得人类很难一直掌权。如果数字智能比生物智能多出百万倍，很难想象人类还能说了算。所以只能希望人工智能对我们友善点。I find it unlikely that humans would remain in control.I mean,if there's a million times moredigital intelligence than there is biological intelligence, it's hard to imagine humans being in control in that scenario.So hopefully AI is nice to us.

我觉得写代码时加入求真和好奇心的设定，应该不错。咱们在网上也得注意点，别搞太多反乌托邦的故事。因为AI的训练数据都来自网络。I think programming truth seeking and curiosityis probably good. I guess we wanna be careful about being, you know, creating too many AI dystopian scenarios online because the AIs are trained on everything online.

它们可能会照搬《终结者》的剧情。毕竟网上有八十版剧本供它学习。（笑）看来我就成了终结者。咱们得多编点AI的乌托邦故事。这样它或许不当反派，而是当个友好的监督者。或者像园丁一样照顾人类，努力让我们幸福。我朋友提出个概念叫“渐进乌托邦”（protopian）。传统乌托邦太理想，硬造出来往往变反乌托邦。她主张追求渐进乌托邦，也就是可实现的、虽不完美但持续向好的未来。展望未来时你会觉得，嗯，这确实是个好日子。Yeah, it should be Terminator, you know, 'cause whatever, they read, you know, 80 versions of the Terminator script. (laughs) It's like, I guess that's me, Terminator.Maybe we should have some more, let's just have some more utopian narratives for AI. So then I can like, well, maybe I shouldn't be Terminator. Maybe I can be friendly overseer or something like that.Like a gardener or something, I don't know.Treat us, you know, try to make humans happy type of thing. Friend of mine has this idea called protopian, which is like, 'cause utopian tends to represent some impossible idealistic goal. And usually in attempting to create a utopia, you create a dystopia is usually how the narrative goes. So she thinks it should be protopia, which is an achievable,you know, future. That is, it's not perfect, but it's good. And something that you'd be, like if you could look into the future, you'd say, okay, yeah, that would be a good future.

不客气，沃尔特。回见。You're welcome, Walter. See ya.