How to Make Billions on AI? Yerzat Dulat on Higgsfield and the New Economy

In the creative team alone, over the next couple of years, my personal goal is to get to 2,000 people.

光是创意团队，未来一两年，我个人的目标就是做到 2,000 人。

So [music] you want to grow the headcount here, in Kazakhstan, tenfold?

也就是说 [音乐] 你们想把哈萨克斯坦这边的团队扩到 10 倍？

Yes, the media market, it's going to be gigantic. It won't just be Higgsfield in it, there will be a large number of companies in it. That's also our kind of policy, that we always move data-driv—, data-driven. We always want to get the data from the market, rather than sit around, rather than reflect, you know, rather than plan — but move fast. Well, right now 300 million, yeah, in general,

对，媒体市场会大得惊人。里面不会只有 Higgsfield，会有一大批公司。这也是我们的一条原则：我们永远是 data-driven 地往前走。我们要的是从市场上拿到数据，而不是坐着空想、做规划，而是快速行动。现在是 3 亿，对，总体上，

well, already more,

嗯，已经不止了，

more than 300 million. I don't know, do you grow every week or something?

超过 3 亿。我都搞不清了，你们是每周都在涨吗？

Well, we're growing 30% a month.

我们每个月涨 30%。

Hi everyone. On air is another episode of the program Baitassov Live. We're continuing the topic that we've been raising for a whole month already. We're talking about the economy of Kazakhstan. Today my guest is a very interesting person, the creator of the project Higgsfield, Yerzat Dulat. Yerzat, good afternoon. Glad to see you in our studio at Business FM. We're doing a cycle of programs right now, the one I do on the economy. And you're exactly the person I wanted to talk with about the new economy. Because Kazakhstan, if you look at it, is such an interesting country. We're standing on a frame that was built back in the 20th century. Uh, the higher the prices for raw materials, the better the economy of Kazakhstan feels. But at the same time smart people, progressive people always talk about the new economy, they talk about the creative industries, they talk about what else our country can earn money from and what we have to do in order to feel comfortable in the future, when the oil runs out or some other resources run out. There. And one of those directions is, of course, the creative industries. I'm very glad that back in the day Bakytzhan Abdirovich Sagintayev created the department for creative industries, and a whole story took shape around that, and Kazakhstan in general devotes a great deal of attention to new industries. And Astana Hub was created, conditions were created so that companies like Higgsfield could appear here. And I'd like to talk about how that affects modern Kazakhstan and what the prospects are in the future. Well, and about many, many other things.

大家好。这里是《Baitassov Live》的又一期节目。我们继续这一个月来一直在谈的话题——哈萨克斯坦经济。今天做客的是一位非常有意思的人物，Higgsfield 这个项目的创始人 Yerzat Dulat。Yerzat，下午好。很高兴在 Business FM 的演播室见到你。我现在在做一档关于经济的系列节目，而我正想和你聊聊新经济。因为哈萨克斯坦是个很特别的国家：我们至今还站在 20 世纪搭起来的那副骨架上。呃，资源价格越高，哈萨克斯坦经济就越舒服。但与此同时，聪明人、有想法的人总在谈新经济，谈创意产业，谈这个国家还能靠什么赚钱、我们要做什么，才能在石油枯竭、或者别的资源枯竭之后依然过得舒坦。是这样。其中一个方向当然就是创意产业。我很高兴当年 Bakytzhan Abdirovich Sagintayev 设立了创意产业管理局，围绕这件事形成了一整套故事，哈萨克斯坦整体上对新兴产业投入了非常多。Astana Hub 也建起来了，条件具备了，才有了像 Higgsfield 这样的公司。我想聊聊这对今天的哈萨克斯坦意味着什么，未来的前景又如何。当然还有很多很多别的话题。

Yes. Thank you very much, Armanzhan Merkevich, for inviting me. It's a great honor to be your guest. I'm happy to discuss, yeah, all of these topics.

好。非常感谢 Armanzhan Merkevich 的邀请。能做客您的节目是莫大的荣幸。这些话题我都很乐意聊。

Thank you. We, uh, started watching your company a long time ago. Yerzat, uhh, as many people know, if anyone is following it, uh, but I think that many are following, especially young people — he was in our "tinde" ranking back in 2024. Even then we were writing and talking about your company, and even then everybody was saying that you have huge, huge potential. But today it's already obvious. The company has a turnover of hundreds of millions of dollars, and the valuations are all over the place, from one and a half to 3 billion dollars. Well, it's clear that the market will still show, right, sort of, but in any case the potential and the appetite of the investors who are showing up speaks for itself, that your company really has pulled it off. What you were telling us back in 2024, it all turned out exactly that way. In 2024, I think, you were the first to raise a large sum, 8 million dollars. That's

谢谢。我们呃很早就开始关注你们公司了。Yerzat，呃，很多人都知道——如果有人在追这块的话，我想追的人不少，尤其是年轻人——早在 2024 年就上过我们的 tinde 榜单。那时候我们就写过、讲过你们公司，那时候所有人都说你们潜力巨大。而今天这已经是明摆着的事了。公司营业额是数亿美元，估值众说纷纭，从 15 亿到 30 亿美元都有。当然，最终还得由市场说了算，对吧，不过不管怎样，光是这些冒出来的投资人的胃口和期待就说明了问题——你们这家公司是真的做成了。你们 2024 年跟我们讲的那些，后来全都成真了。2024 年，我记得，你们是第一个融到大额资金的，800 万美元。这是

the first round.

第一轮。

The first round. That was, honestly speaking, unique at that moment. Well, and still, uh, for the audience, let's tell it once more — what is Higgsfield, how is it built, how did the name come about, who came up with it and what does the company's name mean.

第一轮。说实话，在当时这是独一份。不过还是，呃，给观众再讲一遍吧——Higgsfield 到底是什么，怎么搭起来的，名字是怎么来的，谁想出来的，公司名又是什么意思。

Yes, Higgsfield is a platform that focuses on artificial intelligence in the video space. That is, that's our main focus, where we keep — the whole product is built around the video space. And back in 2023 Jensen Huang with his Nvidia said that in the future 100% of the pixels on your phone, on your computer, on any monitor will be AI generated. That is,

好。Higgsfield 是一个专注视频领域人工智能的平台。这就是我们的主攻方向，整个产品都围绕视频这一块来搭。早在 2023 年，Jensen Huang 就带着他的 Nvidia 说过，未来手机上、电脑上、任何显示器上 100% 的像素都会是 AI generated。也就是说，

Mm-hm.

嗯。

everything that we'll see on screens, it will be generated with the help of AI. And this really was a gigantic opportunity that appeared in 2023, when the company appeared. And we decided, well, to focus on this market. And we're one of the very first companies in the video AI market at all.

我们将来在屏幕上看到的一切，都会是 AI 生成出来的。2023 年公司成立的时候，这确实是个巨大的机会摆在那里。我们就决定专注这个市场。我们是整个视频 AI 市场上最早的一批公司之一。

Mm-hm.

嗯。

So. And we actually started even earlier than the industry began to become mature. And for a long time we didn't have, you know, big growth, because the technology was still very raw. It was more like R&D. So. And indeed, in 2024 the technology matured, when, well, market applications for video appeared, when AI started to compete with real video production, with the likes of Hollywood. And now, yeah, there's this mega-exponential growth of the video market happening. And, well, and we accordingly move along with this market, as, you know, the most experienced player on the market, one of the first players on the market. So. And at the very beginning, uh, as I said earlier, the focus was more on R&D, that is, research and development. And at the very beginning, when we had only just founded the company, we weren't thinking yet that we would specifically do video. We just wanted to do research in neural networks.

是这样。其实我们启动的时间，比这个行业开始成熟还要早。有很长一段时间我们没什么大的增长，因为技术还很粗糙，更像是在做 R&D。到 2024 年技术确实成熟了：视频有了市场化的应用场景，AI 开始能跟真正的视频制作、跟好莱坞这类对手掰手腕。而现在，视频市场正在超指数级增长。我们也就跟着这个市场一起往前走，作为市场上最有经验的玩家、最早的一批玩家。最开始，就像我刚才说的，重心更多在 R&D，也就是 research and development。刚创立公司那会儿，我们还没想过具体要做视频，只是想做神经网络的研究。

And that's why we wanted a kind of deep-tech name, uh,

所以我们想要一个 deep tech 味儿的名字，呃，

because it's fashionable. There's DeepMind, for example, the very first successful commercial artificial intelligence laboratory, where the whole revolution actually started in the 2010s. There's OpenAI. And we thought it would be cool to name it Higgsfield. It's a name from physics. I myself

因为这样才时髦。比如有 DeepMind——第一家成功的商业人工智能实验室，整场革命就是从那儿、从 2010 年代开始的。还有 OpenAI。我们就想，叫 Higgsfield 会很酷。这是个来自物理学的名字。我自己

The Higgs boson. Yeah.

希格斯玻色子。对。

Yes. The Higgs boson is this field that creates mass for everything that surrounds us. That is, it's a fairly fundamental thing. And we — since I myself did physics back in school, physics olympiads — it was very close to me, and so we decided to name the company Higgsfield.

对。希格斯玻色子就是这样一种场，它给我们周围的一切赋予质量。也就是说，这是个相当底层的东西。而我自己上学时就搞物理、打物理竞赛，所以对这个特别有亲近感，我们就决定把公司叫 Higgsfield。

Cool. Well, I want to say, yeah, for the audience, you can read once more on our site, in Forbes magazine, that Yerzat, uh-uh, is a winner of many olympiads and, straight up, uh, you could say an outstanding figure in the field of artificial intelligence, because, uh, Yerzat gets cited at Harvard too, and even Sam Altman, a very, very long time ago, back before Higgsfield, was, uh, sending you letters, right, sort of, and making remarks. By the way, what was the letter from Altman himself about?

厉害。我想跟观众再说一句，你们可以到我们网站上、到 Forbes 杂志上再读一遍：Yerzat 是很多竞赛的冠军，甚至可以说，呃，是人工智能领域的杰出人物——因为哈佛都在引用 Yerzat，甚至 Sam Altman 很早很早以前、还在 Higgsfield 之前，就给你写过信，对吧，还提过意见。顺便问一句，Altman 本人那封信是关于什么的？

Back then the industry was very small, that is, at the time there weren't these trillion-dollar valuations on every company. Back then it was a fairly intimate community. Everybody talked, uh, quite openly on social networks like Reddit, uh, on Discord servers. Those are online communities. And everyone who was basically at the frontier of artificial intelligence, everyone knew each other. Everyone shared ideas. It was this small, fairly— community. And now they've been leaked — right now there's a lawsuit going on between Elon Musk and Sam Altman.

以前这个行业非常小，那会儿还没有现在这种动辄上万亿的估值。当时是个相当小圈子的社群。大家交流得挺开放的，就在 Reddit 这类社交网络上、在 Discord 服务器里。就是这种线上社区。基本上，凡是站在人工智能前沿的人彼此都认识，大家互相分享想法。那是个很小的社群。而现在这些东西都被泄出来了，Elon Musk 和 Sam Altman 之间正在打官司。

It's starting, yeah.

开始了，对。

Yes. And the correspondence has been published. And in it Elon Musk writes to Sam Altman that they need to win the talent race against Google, because Google had just bought DeepMind, a top company, and Google started hiring all the best PhDs in the machine learning field. And OpenAI — it was hard for them to hire talent. And Elon Musk proposed to Sam Altman to hire schoolkids, to hire graduates, you know, first- and second-year students, before they even, you know, became PhDs, ended up in Google Brain, when it's already hard to compete, because, well, Google has more resources, they can pay a bigger salary. And OpenAI really did have a strategy of hiring very young guys onto the team back in those years, in 2017–2018. And also, since I was very active in the machine learning, artificial intelligence community, I ended up on the radar too and they wrote to me.

对。那些往来邮件都公开了。里面 Elon Musk 写给 Sam Altman 说，他们必须在人才争夺战里赢过 Google，因为 Google 刚买下 DeepMind 这家顶级公司，还开始把机器学习领域最好的博士全招走。而 OpenAI 招人很吃力。于是 Elon Musk 向 Sam Altman 提议：去招中学生，去招大一大二的学生，趁他们还没读成博士、还没进 Google Brain——到那时候就很难竞争了，因为 Google 资源更多，工资开得更高。OpenAI 在那几年、也就是 2017、2018 年，确实有这么一套把很年轻的人招进团队的策略。而我因为在机器学习、人工智能社群里很活跃，也进了他们的雷达，他们就来找我了。

Totally unique. Well, by the way, about that moment I even want to read something out. As of the end of 2024, AI startups worldwide, uh, reached 50 billion, only that much at that point. That's the end of 2024. And OpenAI, Anthropic and others managed to raise about 18 billion dollars over the course of 2023. And that was, well, some kind of big achievement. Today the numbers there, uh, are spinning closer to a trillion, and that's in just 2 years — the discussion there has changed sharply, sharply. That is, you, well, you went into exactly the topic that was exactly, exactly on time. By the way, I remember when the piece was being made. This piece came out on September 30, 2024, so not that long ago. Uh, and it's exactly the idea of telling a film. The first time I heard this idea was from Timur Bekmambetov in 2008. And he said: "I have a dream of creating a certain product, so that you could tell it something, some text, a script, and the program would be able to output, well, at first we were thinking about animation — we weren't even dreaming about cinema back then, right, that was 2008 or 2009." And I'm just talking about how you have to end up in the right place at the right hour, when all of this started working and suddenly started being monetized and turned into a company that today is called, well, a real unicorn. That is, you first — if we take Higgsfield once again — first you were like consultants, uh, and then you came to working with video. And why video specifically? What, how, how did you feel that this specifically was going to be in demand right now?

太独特了。顺便说到那个时间点，我甚至想读一段。2024 年结算下来，全世界人工智能领域的创业公司，呃，才做到 500 亿——当时也就这个数。这是 2024 年底。而 OpenAI、Anthropic 这些公司，在 2023 年一年里融到了大约 180 亿美元。这在当时算是个了不得的成就。今天那边的数字已经往万亿上转了，才短短 2 年，讨论的量级就急剧、急剧地变了。也就是说，你们进的这个方向，时间点踩得正好。顺便说，我记得当时做那篇稿子的情形。那篇稿子是 2024 年 9 月 30 日发的，其实没多久。呃，讲的正是把电影讲出来这个想法。这个想法我第一次听到是在 2008 年，从 Timur Bekmambetov 那里。他说："我有个梦想，做出一个产品，你可以给它讲点什么，一段文字、一个剧本，程序就能输出来——嗯，一开始想的是动画，电影我们那时候连做梦都不敢想，对吧，那是 2008 还是 2009 年。"我想说的就是，你得在对的地方、对的时刻恰好在场，赶上这一切开始跑通、突然能变现、变成一家今天真正被称作独角兽的公司。也就是说，你们最初——再说回 Higgsfield——最初你们是做咨询的，呃，然后才转到做视频。为什么偏偏是视频？你们是怎么感觉到，这东西恰恰在现在会有需求的？

Yes, we looked at different verticals. The biggest vertical in AI right now is coding, that is, Anthropic. Most of their revenue is code generation. That is, it's various other software companies, they use Anthropic to automate the writing of code. And we understood that a race had already started there, high competition. And we were trying to find a niche that is still only just, I mean, only just taking off. And there were also different niches. There's, uh, computer games, there's medicine, uh, drug development, there's, in the legal part, also quite fast-growing startups right now, unicorns. And having weighed everything up, we made the decision to focus on video after all, we made prototypes in different areas, but in the end we still decided to settle on video, because we understood that the video market is a gigantic opportunity, it's a trillion-dollar market, and that's why, and we saw that the teams in America are fairly weak. That is, at that moment, the competitors that were in America, we figured that we could put together a stronger team than they could

是的，我们看过不同的赛道。现在 AI 里最大的赛道是 coding，也就是 Anthropic。他们大部分 revenue 来自代码生成——各种软件公司用 Anthropic 来自动写代码。我们当时就明白，那边的竞赛已经开打了，竞争很激烈。所以我们想找一个刚刚起步、才开始起飞的细分。可选的方向也不少：电脑游戏、医疗、新药研发，legal 那块现在也有增长很快的创业公司、独角兽。全部权衡之后，我们还是决定聚焦视频。我们在不同领域都做过原型，但最后还是选了视频，因为我们判断视频市场是个巨大的机会，是万亿级市场。而且我们看到，美国那边的团队都挺弱——当时在美国的那些竞争对手，我们算过，我们能攒出一支比他们更强的团队。

here in Kazakhstan. Will you be able to — and it worked out, right, in

……在哈萨克斯坦这边。你们能不能——后来确实成了，对吧，在

Yes, that's exactly how it worked out. And we were one of the first on the market, but there were several companies that showed up at the same time as us. And when we raised 8 million, competitors were raising 80, 150 million right away.

对，就是这么成的。我们是市场上最早的一批，不过也有几家公司几乎和我们同时冒出来。我们融到 800 万的时候，竞争对手一上来就是 8000 万、1.5 亿。

The founders there were professors from Stanford. Who were the authors of some scientific papers that were the very foundation of the technology.

那边的创始人是斯坦福的教授，是一些论文的作者，而那些论文本身就是这套技术的底层基础。

Mm-hmm.

嗯。

But we were still confident that we could put together a stronger team than in the US.

但我们还是确信，我们能攒出一支比美国更强的团队。

And yes, that's really how it turned out. And another big factor of success is that, besides a strong engineering team, we also managed to put together a very strong creative team, because marketing is built on creative. And it turned out that in Silicon Valley everything is very bad with marketing and with the creative industry. There's just this kind of gap there.

对，后来确实是这样。还有一个很大的成功因素：除了强的工程团队，我们还搭起了一支非常强的创意团队，因为营销的根基就是创意。而事实证明，硅谷在营销和创意产业上非常差，那边就是有这么一块空白。

Mm-hmm.

嗯。

And people there understand artificial intelligence well, engineering, but they don't understand media at all, they don't understand at all, uh, yes, the creative industry.

那边的人很懂人工智能、很懂工程，但完全不懂媒体，完全不懂创意产业。

Amazing. What's amazing is that, that is, we turned out to be more advanced in this, in the creative industries, so it turns out, right?

神奇。神奇的是，在创意产业上反倒是我们更领先——是这么回事吧？

Yes. Yes. We turned out to be far more advanced. And when we first made our technological product, which took off, and then we needed to hold on to our growth, and, of course, here go to market is what decides it, that is, the quality of marketing, the quality of how you can build a brand. And here our creative team already played a gigantic role, first of all, in holding on to what we had achieved through our technology, and then in growing. And we started growing very actively. And we're actively hiring and scaling our creative team right now. Right now we have more than 100 people in the creative team, and we're planning, in the creative team alone, in the next couple of years, my personal goal there is to get to 2,000 people.

对，对。我们领先得还挺多。我们先做出了技术产品，产品跑起来了，接下来要守住增长——这时候当然是 go to market 决定成败，也就是营销的质量、能把品牌做成什么样。到这一步，我们的创意团队就起了巨大作用：一是守住我们靠技术拿到的东西，二是继续往上长。我们开始快速增长。现在我们在很积极地招人、扩创意团队。目前创意团队已经有 100 多人，光创意团队这一块，未来两年我个人的目标是做到 2000 人。

Up to 2,000 people. How do you get onto your team? 2,000 people is a big workforce. Today how many employees do you have at the moment? Right now about 215.

到 2000 人。怎么才能进你们团队？2000 人是个很大的班底。你们今天有多少人在职？——现在大概 215。

215. So you want to increase the headcount tenfold here, in Kazakhstan. How, how are you going to select people, what ideas do you have?

215。也就是说，你们想把哈萨克斯坦这边的编制扩到 10 倍。你们打算怎么挑人，有什么想法？

Yes, this is also such an interesting challenge for us, uh, to figure this one out, how to hire at that scale, at that kind of scale. And we always try to look at benchmarks. We look at other successful startups at our stage, how they, uh, expanded, and we try to repeat, like, the best playbooks. And how does it need to be done? Actually it's very important, in order to expand, it's very important that the core be strong enough, because the core usually, how it happens is, everybody calls in their friends, then friends of friends, that's how a network gets built.

对，这对我们也是个挺有意思的挑战——搞清楚在这种规模下怎么招人。我们一直会看 benchmark：看同阶段其他成功的创业公司是怎么扩的，试着复制最好的 playbook。到底该怎么做？其实想扩张，很关键的一点是核心圈要足够强，因为核心通常是这么长起来的——大家先叫自己的朋友，然后是朋友的朋友，人脉网络就这么搭起来了。

That's how we did it back in the nineties,

我们九十年代就是这么干的，

yes. Well, and it's the same in the Valley, in Silicon, there MIT graduates call in their own, Harvard graduates call in theirs. And that's why it's very important, if the core is very high-quality, if the people really are A-players, that is, people who really want to perform at a high level, who want to reach high goals, they want to work with people on a par with themselves. And so it's very important to hold this high bar, uh, the quality of people. And we have several stages of interviews. That is, to get onto the team, at least three people inside the company have to have interviewed the person, vouched for them, understood that we're a fit for each other culturally, and that we're a fit for each other on work ethic, and that we're a fit on skills, because the person really does have creative skills, technical skills. Here's the thing I'd like to clarify here. The thing is that HR is a whole science. Of course, you can read books, you can hire consultants, work at it, but every time, the most expensive mistakes companies made, they were always in the mistake, it's, it's people. Usually the companies that win, usually what I already see from my own experience, it's those companies that, when they take people on, base it on some kind of, well, super-mission, a foundation, on some kind of philosophy. Your circle, which I see, right, even in interviews you said that certain people gave you recommendations, there Arman Sulmenov helped you a lot there to focus, right, something like that, that is, your circle influences personal growth, the company, very strongly. And that's why the selection of personnel, well, it's, like, absolutely important. So your basic philosophy, your ideas, with which you inspire people, whom you take on or don't take on, so, so what is it you come with,

对。硅谷也一样，MIT 的毕业生叫 MIT 的人，哈佛的叫哈佛的人。所以核心圈的质量非常重要——如果这批人真的是 A-player，也就是那种特别想做出成绩、想够到高目标的人，他们就想和跟自己同一水平的人共事。所以把这个高标准守住、把人的质量守住，非常重要。我们的面试有好几轮：想进团队，至少要公司内部三个人面过、背过书，确认文化上合得来、工作伦理上合得来、技能上也合得来，这个人确实具备创意能力、技术能力。这里我想追问一下。问题是，HR 本身是一门学问。当然可以读书、可以请顾问、可以慢慢磨，但公司犯过的最贵的错，每次都出在人身上。通常赢的公司——这是我从自己的经验里看到的——是那种招人时把标准建立在某种超级使命、某种根基、某种哲学之上的公司。我看到你身边的圈子，你在采访里也说过，有些人给过你建议，比如 Arman Sulmenov 帮你聚焦了很多，对吧，大概是这个意思——也就是说，圈子对个人成长、对公司的影响非常大。所以选人这件事，可以说是极其重要的。那么你的基本哲学是什么，你用什么样的想法去感染人、去决定收不收一个人，你带着什么

right? The main philosophy

对吧？基本的哲学

besides the skills that, right, have to be there, right,

除了技能之外，那些当然得有，对吧，

of course, yes, the company's main philosophy is that we want to build a world-class technology giant out of Kazakhstan. That is, it's a fairly ambitious goal, and we're looking for people who, first, believe in this goal, and, second, are ready to give their maximum to achieve it. That is, there are a lot of Kazakhstan patriots working at the company, a lot of this kind of charged-up young people who want to prove it to the whole world. And for that to happen, uh, my personal goal, the way I look at it, is that we want to import, uh, the American venture institutions that allowed the American economy to grow actively over the last several decades, at least over the last 40-50 years. If you look at the top 10 biggest companies in the S&P 500, most of them are technology companies that were once startups and that were incubated, that grew precisely because of this American institution of venture financing and, globally, of the venture institution. And the task is to import this institution, uh, into Kazakhstan, because it's a working institution, uh, it has its own rules, it operates by clear economic and capitalist rules that have already been tested over decades, they've been honed. Nothing new needs to be invented there. How a company should be structured, how you need to hire people, at what stages, how financing deals are structured, how a company should grow, how it should relate to the market, to risk, to all of this. This is, like, what they teach students at Stanford, uh, that's where it comes from and it's, like, the core of the Valley there at Berkeley. And the task is to import, uh, this institution here to Kazakhstan. And we really do see that the people who have been in Higgsfield from the very beginning, they understand this playbook very well. Basically, on the core team there are about 20-25 people who understand the whole sequence of steps from zero to a unicorn, to make a unicorn from scratch. And, basically, the guys could even go now and make their own unicorn, because they understand everything very clearly.

当然，对。公司的基本哲学就是：我们要从哈萨克斯坦做出一家世界级的科技巨头。这个目标相当有野心，所以我们找的人，第一要信这个目标，第二愿意为达成它拼到极限。公司里有很多哈萨克斯坦的爱国者，很多憋着一股劲、想向全世界证明自己的年轻人。而要让这件事发生——这也是我个人的目标、我自己的看法——我们想把美国的风险投资制度引进来，正是这套制度让美国经济在过去几十年、至少过去四五十年里高速增长。看看 S&P 500 里最大的十家公司，大部分是科技公司，它们曾经都是创业公司，都是靠美国这套风险融资制度、靠全球化的风投制度孵化、长大的。我们的任务就是把这套制度引进哈萨克斯坦，因为它跑得通，它有自己的规则，按清晰的经济规则和资本主义规则运转，这些规则几十年来被反复验证、被打磨过，不需要再发明什么新东西：公司该怎么搭结构、什么阶段该招什么人、融资交易怎么设计、公司该怎么长、该怎么面对市场、面对风险、面对这一切。这些东西斯坦福就是这么教学生的，源头也在那儿，硅谷的内核就在伯克利那边。任务就是把这套制度搬到哈萨克斯坦来。我们确实看到，从一开始就在 Higgsfield 的那些人，非常懂这套 playbook。核心团队大概有 20-25 人，清楚从零到独角兽每一步该怎么走。基本上，这些人现在出去自己做一个独角兽都行，因为他们全都想得很明白。

Mm-hmm. And further, our task is even more ambitious, to build a company of 100-plus billion dollars. And it's the same thing, this is

嗯。接下来我们的目标更有野心：把公司做到 1000 亿美元以上。这也是同一回事，这

To build a company of 100-plus billion dollars,

把公司做到 1000 亿美元以上，

right? This is what we've committed to, this is what we tell investors. Investors believe in us, that the team has committed.

对吧？这是我们承诺过的，也是我们跟投资人讲的。投资人相信我们，相信团队已经立下了这个承诺。

And this too, it's not some completely new science, these are the same venture institutions. And the companies that have reached a valuation of 100+ billion dollars, there are more of them than fingers on both hands. That is, there's already a large number of such companies. And we also want to understand how it's done. And in the future, obviously, I always had the goal of creating jobs in Kazakhstan, uh, of attracting, essentially, direct financing into the country. But globally, and Higgsfield, if it becomes a 100-plus-billion-dollar company, of course, that will have a big impact on the economy. But if you look longer term, then the people who are inside, uh, who went through this whole path, they'll understand the whole playbook of how to build companies like this, and then they'll create their own companies, and then we'll see, uh, a much stronger multiplier effect on the economy of Kazakhstan. That's what I strongly hope for.

这同样不是什么全新的科学，还是那一套风投制度。估值到 1000 亿美元以上的公司，数量也比两只手的手指还多，这样的公司已经不少了。我们也想弄明白这是怎么做到的。至于未来，很清楚——我一直的目标就是在哈萨克斯坦创造就业，本质上是把直接投资引进这个国家。从大局看，Higgsfield 如果真做到 1000 亿美元的体量，当然会对经济产生很大影响。但如果看得更长远，公司内部走完这一整条路的人会吃透整套 playbook，知道这种公司怎么建，然后他们会去开自己的公司——那时候我们会看到对哈萨克斯坦经济更强的乘数效应。这是我特别期待的。

A very interesting idea. A very interesting idea. That is, uh, you open up jobs here, uh, you expand the team here, you attract investment from abroad. Basically, uh, investors are already interested too. By the way, thanks to Higgsfield you managed to do such a thing, to break through, I don't know, a dam. And investors now look at our country as a whole differently. And really, potentially, uh, maybe out of the company Higgsfield, uh, maybe around you, more interesting projects will appear, for which it will be much easier to raise money from abroad and work here, in Kazakhstan. Basically, this is exactly what you dreamed about at one time, to create in Almaty, in Kazakhstan, convenient conditions, safe, comfortable, good conditions for work. And basically you can work from Kazakhstan, creating a world-class company. So you are basically turning this idea, this dream, into reality. The idea, the idea is clear. Well, and there's probably an upside too, right, in that the people who work here, it's much harder for the same Google or Apple or OpenAI to poach them from you, right, frankly speaking that kind of thing exists too, right?

非常有意思的想法。非常有意思。也就是说，你们在这里创造就业、在这里扩团队，同时从国外引进投资。而且投资人现在也已经有兴趣了。顺便说，多亏了 Higgsfield，你们做成了一件事，把一道堤坝给捅开了。现在投资人看我们整个国家的眼光都不一样了。而且确实，从潜力上讲，也许从 Higgsfield 内部，也许在你们周围，还会冒出更多有意思的项目，它们从国外拿钱会容易得多，人却留在哈萨克斯坦这边干活。你当年梦想的其实就是这个：在阿拉木图、在哈萨克斯坦创造出方便、安全、舒服的好工作条件，人在哈萨克斯坦也能做出世界级的公司。你现在基本上就是在把这个想法、这个梦想变成现实。这个思路我明白了。而且大概还有一层好处，对吧——在这儿工作的人，Google 也好、Apple 也好、OpenAI 也好，要从你们这儿把人挖走要难得多，对吧，坦白说这种事确实是存在的，对吧？

And how are things with staff turnover for you, if we're being honest,

那说实话，你们的人员流失情况怎么样，

Yes, true, this is a strong competitive advantage of ours, in that there's practically none of it. That is, the entire founding team of the company, it's still working, and it's already the third year and nobody has left. And we also see that turnover is practically zero. And this really is, this is a big competitive advantage, because in the Valley, where in principle there's a lot of money, literally early stage startups, that is, startups at an early stage, they're flooded with money

对，这确实是我们很强的竞争优势——我们这儿几乎不存在这个问题。公司整个创始团队到今天还都在，已经第三年了，一个人都没走。我们也看到，人员流动几乎为零。这真的是很大的竞争优势，因为在硅谷，本来钱就非常多，early stage 的创业公司，也就是早期阶段的创业公司，钱多到用不完。

and quite high salaries. And since there are a lot of startups, the labor market, it's maximally disloyal. That is, as soon as, for example, a startup starts doing badly — this is such a classic Valley thing — the founders can come in the next day and there will be nobody in the office, because in one day, well, and American labor law, it's maximally liberal, employees can move to another startup in one day and

工资也相当高。而且因为创业公司太多，劳动力市场的忠诚度极低。就是说，比如一家创业公司一旦经营不顺——这是很典型的硅谷现象——创始人第二天来，办公室里可能一个人都没有了，因为一天之内……而且美国的劳动法极其宽松，员工一天之内就能跳到另一家创业公司，

they pa-, they packed up their things fast, grabbed their computer, moved to the office next door, right, to work,

东西一收，电脑一拿，就搬到隔壁办公室去干活了，对吧，

yeah. And this is, this is such a structural problem of the Valley in general, in that — and this is our competitive advantage, because to even get to a unicorn, you have to put together a team of twenty super-strong people, I call them killers. 20 killers, who are the best engineers, the best designers, the best sales, the best marketers.

对。这其实是硅谷一个结构性的问题，而这正是我们的竞争优势。因为哪怕只是做到独角兽，你也得凑齐二十个超强的人，我管他们叫「杀手」。20 个杀手，最好的工程师、最好的设计师、最好的销售、最好的市场。

Uh-huh.

嗯。

20 people. And you can, well, obviously, in a growing market, get to a unicorn by the playbook. But to build a 100-plus billion dollar company, you need to gather something like 90-100 such people. And in America that's very hard to do, because at an early stage it's very easy to raise money. That's why everyone goes and opens their own startups. Employees are very disloyal, and more than a year on average is considered not even cool in the Valley, to work at one company. That is, there's this kind of culture there,

20 个人。在增长的市场里，按 playbook 做到独角兽是可以的。但要建一家 1000 亿美元以上的公司，得凑到大概 90-100 个这样的人。在美国这非常难，因为早期融资太容易，所以人人都跑去开自己的创业公司。员工非常不忠诚，在硅谷，平均在一家公司待超过一年，甚至会被觉得不够酷。那边就是这么一种文化，

because even like that, right,

因为就算这样，对吧，

yeah.

对。

Because there's this mechanic, a one-year cliff, when an employee has shares in the company. He needs to work at least one year so that the shares, if he earlier,

因为有一个机制叫「一年 cliff」：员工手里有公司的股票，他至少得干满一年，股票才——如果他提前，

yeah, vest to him, become his private property and nobody can take them away anymore. And this is the classic Valley, that nobody works more than one year. Well, you can understand the engineer too, in that you can work four years at one company, or you can work at four companies for a year each and somehow also balance your risks,

对，才归属到他名下，成为他的私人财产，谁也拿不走了。这就是典型的硅谷：没人在一家公司待超过一年。当然工程师也可以理解——你可以在一家公司干 4 年，也可以在四家各干一年，这样也算是平衡自己的风险，

build a portfolio, so to speak, right, build your own investment portfolio, yes, yes, an investment portfolio — this really is, this is super. And this is kind of, this is mainstream in the Valley.

算是建一个组合，对吧，建自己的投资组合，对，对，投资组合——这确实挺妙的。而且这在硅谷算是主流吧。

So. But in Kazakhstan there's none of that. This is also our strong competitive advantage, that basically in Kazakhstan, uh, from the start my bet was, well, on our talent, and nobody there believed in it, except, well, my cofounder, respectively. He's not, he's not from Kazakhstan, he was living in America when we were building the company. But globally even many of our own compatriots said that it wouldn't be possible to put together a strong team in Kazakhstan. And, uh, everyone was leaving for the Valley, uh, build your teams there. I never understood this, because I understood that, well, RFMSh, where I'm a graduate, is a very strong school. I have a bunch of friends there who can easily get hired at any global giant, like that. We took part in olympiads there. Kazakhstan consistently makes the top 10 in math, physics, informatics at the international olympiads. And I just did not understand where this disbelief in Kazakhstani talent came from. For me it was completely obvious that the team had to be built in Kazakhstan.

在哈萨克斯坦没有这种事。这也是我们很强的一个竞争优势。在哈萨克斯坦，我一开始就把赌注押在我们本土的人才上，当时没人相信，除了我的联合创始人。他不是哈萨克斯坦人，我们做公司的时候他住在美国。但整体上，连我们很多同胞都说，在哈萨克斯坦不可能组出一支强团队，大家都往硅谷跑，说团队要在那边建。我一直不理解，因为我知道 RFMSh——我就是从那儿毕业的——是一所非常强的学校。我有一堆朋友，随便进世界上哪家巨头公司都没问题。我们参加过奥林匹克竞赛，哈萨克斯坦在国际数学、物理、信息学奥赛上稳定进前十。我完全不明白，对哈萨克斯坦人才的这种不信任是哪儿来的。对我来说，团队就该在哈萨克斯坦建，这太明显了。

In Kazakhstan you have to recruit, right, build, right,

得在哈萨克斯坦招人、建团队，对吧，

yeah. And the bet really did pay off, and it's now this most powerful competitive advantage of ours. But now, since the company is already a big brand in AI, the competitors there, companies that are bigger than us, that are already worth tens of billions of dollars, they are actively trying to poach our employees away from us. And there in London, for example, there's a cohort of startups that know that Kazakhstan is the frontier in video AI, and there's a constant attack out of London in that they try to poach our employees to London. It doesn't work, because from the very beginning, as I said earlier, we practiced the venture institution. From the very beginning we gave all employees, uh, company options, which now — they've already worked more than a year, they've vested, and they have quite large sums.

对。这一注确实押对了，现在这是我们最强的竞争优势。但现在公司在 AI 领域已经是个大品牌了，那些比我们还大、估值已经几百亿美元的竞争对手，在积极地从我们这儿挖人。比如伦敦就有一批创业公司，知道哈萨克斯坦是视频 AI 的前沿，一直在从伦敦发起攻势，想把我们的员工挖到伦敦去。挖不动。因为像我前面说的，我们从一开始就践行 venture 那套制度：从一开始就给所有员工发公司期权，现在他们干了一年多，期权已经 vest 了，金额相当可观。

Well, if the company is going to be worth 100 billion, I can imagine how much our Forbes list is going to have to be topped up in the near future then.

要是公司真值 1000 亿美元，我可以想象，接下来我们的 Forbes 榜单得添多少新面孔。

Yes, I strongly hope so. And so, uh, we managed to put together such a strong culture, very loyal, very ambitious, and globally, right, we want to capture the market.

对，我非常希望如此。我们确实建起了一种很强的文化——非常忠诚、非常有野心，而且是全球的，对，我们想拿下这个市场。

Uh-huh. Okay. And, uh, still, to clarify once more, that is, you, uh, are hiring, are planning to hire another 2,000 people, and these people will be getting a salary, and they will also be becoming shareholders with you, right? That is, or, or will the principle be different for the new ones. — The 2,000, that's only into the creative team. There's also the engineering team, it will be growing too,

嗯。好。还是再确认一下：你们要再招 2000 人，这些人有工资，也会跟你们一起成为股东，对吧？还是说对新人是另一套原则？——这 2000 人只是创意团队。还有工程团队，也会继续扩，

so more people will be hired. Yes, of course, that too, well,

所以招的人还会更多。对，当然，这个也一样，

that's the basic principle, right, of how, how the work will be at your place, right,

这是基本原则，对吧，你们以后就这么运作，对吧，

yes, and the venture institution works that way too. Even if an engineer or, say, an employee joins Google now, then he'll be getting Google shares, which on the stock market — it's already a public company — but he'll nonetheless be getting shares on top of the salary, he can go right away and cash them out on the stock market. That is, this is such a, uh, standard practice at all tech companies, that it's quite generous, and since the whole essence of technology companies is intellectual property, and everything is built on intellectual value added, therefore people are the center, therefore everything is built on making people maximally loyal, maximally, uh, motivating people. And this is a practice — whether Google, a public company, whether OpenAI over there, a pre-IPO company, whether an early stage startup — for everyone it's always built into the financial model and the cap table of the company, it's always built in that, uh, options will be issued to employees. The only question is, the earlier a person joins the company, the

对，venture 这套制度就是这么运作的。哪怕现在一个工程师、或者说一个员工加入 Google，他也会拿到 Google 的股票——Google 在股市上已经是上市公司了——但除了工资，他照样拿股票，而且可以马上拿到股市上变现。这在所有科技公司都是标准做法，而且相当慷慨。因为科技公司的本质就是知识产权，一切都建立在智力附加值上，所以人是核心，所以一切都围绕着怎么让人最忠诚、怎么最大限度地激励人。不管是 Google 这种上市公司，是 OpenAI 这种 pre-IPO 公司，还是早期阶段的创业公司，所有人的财务模型和 cap table 里都写着一条：会给员工发期权。唯一的区别是，一个人加入公司越早，

more upside he has, because it's growing and people can make many x's.

他的 upside 就越大，因为公司在涨，大家可以翻很多倍。

Uh-huh, got it. And here's another question, already looking to the future — might it not turn out the other way around, that the English companies will realize that it's impossible to poach Higgsfield's employees, and will come and work here on the territory of Kazakhstan. Then that would be a total super-victory, if companies come and register here, recruit specialists, work here. How comfortable in general, how comfortable is it to work in Almaty, in Astana for our specialists in the AI field? — Actually this is super possible, because we're seeing it. Uh, there is in India, for example, Anthropic, OpenAI opening development offices, that is, uh, because there are strong hubs there, a lot of unicorns, decacorns, companies with founders from India. Uh-huh. And there are a lot of technology hubs there. The same thing we see in Europe, uh, in Eastern Europe, in Western Europe, uh, there are technology hubs, uh, where really big startups, companies open offices, uh, hire people. And literally yesterday our investors from the Accel venture fund left. That's a top-three venture fund in the world. Uh, 8 billion under management. They invested in Meta, in Spotify, in companies, giant public companies. And what's interesting, so they were telling us that Kazakhstan is the second country after, uh, Sweden, uh, that they've traveled to outside the Valley where they saw a cool technology hub. Uh, Accel invested in the company Spotify. That's the main music app that all the young people listen to music on now. Uh, this public company, it's now worth on the order of 98-99 billion dollars. And those same investors who invested in us, they tell how they went to the Spotify office, which was located outside the Valley

嗯，明白。还有个问题，往后看：会不会反过来——英国那些公司发现挖不动 Higgsfield 的人，就干脆过来，在哈萨克斯坦本地开工？那才是彻底的大胜利：公司过来在这儿注册、在这儿招专家、在这儿办公。总的来说，我们做 AI 的专家在阿拉木图、在阿斯塔纳工作，体验到底怎么样？——其实完全有可能，我们已经看到了。比如在印度，Anthropic、OpenAI 都在开研发办公室，因为那边有很强的 hub，有很多独角兽、十角兽，还有很多创始人来自印度的公司。嗯。那边科技 hub 很多。欧洲也一样，东欧、西欧都有科技 hub，真正的大创业公司、大公司会去那儿开办公室、招人。就在昨天，我们的投资人刚走，来自 Accel 这家风险投资基金。这是全球排前三的风投基金，管理规模 80 亿。他们投过 Meta、投过 Spotify，都是体量巨大的上市公司。有意思的是，他们说，在硅谷之外跑过的地方里，哈萨克斯坦是第二个让他们看到很棒科技 hub 的国家，第一个是瑞典。Accel 投过 Spotify，就是现在年轻人听歌都在用的那个主流音乐 App，这家上市公司现在市值大概 980-990 亿美元。投了我们的这批投资人讲，他们当年怎么去 Spotify 那个开在硅谷之外的办公室，

at the initial stage, right, when

在很早的阶段，对吧，当时

yes, at the very, they at the very — even earlier than with us. And they, yes, they tell that what they see here, this is that thing that they saw there, because globally, right, venture investors, they are of course wary of investing in stories that differ a lot from what they're used to seeing. And when they were already leaving, they, uh, first of all, they said that they will be recommending to other investors to visit our office. And they said at the end of their, uh, trip that they flat out saw this hub outside the Valley. They were very, extremely surprised, struck, that such a hub at the intersection of creativity and technology was pulled off, because the market is going to be gigantic. It won't be only Higgsfield in it, there will be a large number of companies in it. And in this hub the specialists are ahead of the market by at least 9 months, that is, our specialists there are ahead of the specialists in the Valley itself, in America, by at least 9 months. By

对，在非常早的阶段，比来我们这儿还早。他们说，他们在我们这儿看到的，就是当年在那边看到的东西。因为从全局看，风险投资人当然会对那些跟他们习惯看到的差别很大的故事有所顾忌。临走的时候他们说，第一，他们会推荐别的投资人也来我们办公室看看。他们在行程最后说，他们真的看到了一个硅谷之外的 hub。他们非常吃惊、非常震撼，居然能在创意和技术的交叉点上建起这样一个中心。因为这个市场会非常巨大，里面不会只有 Higgsfield，会有很多公司。而这个 hub 里的专家领先市场至少 9 个月——就是说我们的人比硅谷、比美国的人领先至少 9 个月。领先

9 months. And that's a very long time by our time. By our time, yes,

9 个月。按我们现在的时间来说，这已经很久了。按我们现在的时间，对，

Yes, this— this is a lot. And, naturally, in the future 100%, uh, well, we understand how capitalism, how the market works, uh, and the market always looks for optimal conditions, and the market is ready to take on, like, any risks if there's upside. And in this market the most important thing is people. That's why companies are ready to open offices, ready to invest, and for themselves, of course, they'll take on some risks, because Kazakhstan is a new country for them. And they, maybe, still don't understand everything about Kazakhstan, but the upside is obvious. When some hub has formed where, uh, the specialists are nine months ahead, then there's very high potential upside.

是，这非常多。而且未来 100% 会这样——我们都明白资本主义、市场是怎么运转的，市场永远在找最优条件，只要有上行空间，市场愿意承担任何风险。而在这个市场里，最重要的就是人。所以那些公司愿意来开办公室、愿意投资；当然，他们自己也要承担一些风险，因为哈萨克斯坦对他们来说是个新国家，也许他们还没完全看懂哈萨克斯坦，但上行空间是明摆着的。当一个中心形成了，那里的专家领先九个月，那上行空间的潜力就非常大。

Got it. Cool. But that says— confirms once again that, well, from the very start we did everything right. That Astana Hub was created, that— that such conditions were created for work, for creating, for developing companies, that a Ministry of Artificial Intelligence appeared, that a lot of attention is being paid. Basically, everything, everything is logical. One more question about Higgsfield. And competition— they probably ask you about it often, often. The thing is that, uh, if it worked out for you, so to speak, why wouldn't it work out for others? All the more so since, well, there are big companies that you've already run into one way or another, and OpenAI has a lot of capital, not to mention such giants as Google or Meta. Especially since with Meta you— you've already had negotiations, as far as I know. Meta offered you, at the very earliest stage too, to come under their wing or, like, to sell the main part of the shares, right, of your company. How— even being nine months ahead, how confident do you feel that the big giants won't create, well, that kind of powerful competition for you?

明白。厉害。但这再一次说明、再一次证实：我们一开始的路子就走对了。建了 Astana Hub，创造了这样的工作条件、创办和发展公司的条件，成立了人工智能部，投入了很大关注。基本上，一切都说得通。还有一个关于 Higgsfield 的问题。竞争——这个问题肯定经常有人问你。问题在于：既然你们做成了，为什么别人做不成？何况还有那些大公司，你们多多少少已经打过交道，OpenAI 手里资本雄厚，更不用说 Google、Meta 这种巨头。何况据我所知，你们跟 Meta 已经谈过。Meta 在最早期就提出让你们并到它旗下，或者把公司的主要股份卖给它，对吧。即便领先九个月，你有多大把握，这些大巨头不会给你们造成那种很强的竞争？

Yes, indeed, uh, the market is very competitive, very hard and very volatile. That is, uh, and here volatility is such a, well, of course, volatility is a property of early markets. Mm-hm.

是，确实，这个市场竞争非常激烈、非常难、也非常波动。也就是说，波动性——当然，波动性本来就是早期市场的属性。嗯。

The same thing when the internet had just appeared, there was, like, insane volatility all the way up to the dot-com, to the dot-com bubble, when, like, the economy collapsed. We see the same thing in AI, that the volatility is gigantic. And we see it in the fact that, for example, in 2023 all the investors thought that OpenAI would lead to singularity, that there was no need to build any startups anymore. There would be one OpenAI, uh, it would take over the whole world. And everyone had already started burying Google. And, like, they started burying all the rest of the giants. Anthropic was a very small startup, roughly our size, the size we are now they were back in 2023. And what do we see? What happened in 2025? Everything changed. OpenAI's positions. First Google overtook them. In 2026 Anthropic had already overtaken them. And Anthropic, by valuation on the secondary market, a valuation of a trillion dollars. All of this happened over the course of, like, 2 years,

互联网刚出现的时候也一样，波动性大得离谱，一直到 dot-com、到 dot-com 泡沫，经济崩了。AI 里我们看到的是同一回事，波动性巨大。比如 2023 年，所有投资人都认为 OpenAI 会带来奇点，不用再做什么创业公司了，就一个 OpenAI，它会拿下整个世界。那时候 Google 已经被所有人唱衰送葬了，其余的巨头也一个个被送葬。Anthropic 当时是家非常小的创业公司，2023 年他们的规模大概就是我们现在这个体量。然后我们看到了什么？2025 年发生了什么？全变了。OpenAI 的位置——先是 Google 超过了他们，2026 年 Anthropic 也超过了他们。Anthropic 在二级市场上的估值是 1 万亿美元。这一切就发生在两年之内，

yes, it happened very fast. So. And that kind of volatility in the market is, actually, a big opportunity for us, because, uh, we've built such a culture, uh, of the company and of our team, in that we're always in offensive mode. And in volatile markets an offensive strategy is— is always the optimal strategy, because, obviously, it's very risky, you can lose positions, but when someone loses positions, it means someone takes them— yes, takes them. And our culture is fairly aggressive like that, aggressive capture, capture of the market. And here I too, probably, if you look at it more philosophically, uh, I noticed, and when I was watching how our team functions, uh, I started thinking about how national business culture, like, affects, like, a Kazakhstani team. Because, if you think about it, there are, like, national cultures of different businesses. For example, everybody knows Japanese business culture. It's that rigid hierarchy, it's very long-term planning, it's relationships that are built over decades. That is, really building a very high-quality machine. We know the Chinese business model. That's also those— those ambitious gigantic megaprojects, megaplans. And the people all work, maybe, not as organized as Japanese companies, but also at a large scale, coordinating, like, a large number of people.

对，发生得非常快。所以，市场这种波动性对我们来说其实是巨大的机会，因为我们在公司和团队里建立了这样一种文化：我们永远处在进攻模式。在波动的市场里，进攻型策略永远是最优策略——当然，风险很大，可能丢掉阵地，但有人丢掉阵地，就意味着有人把它拿走。我们的文化相当有攻击性，是攻击性的攻城略地，抢占市场。说到这儿，如果往哲学一点看，我在观察我们团队怎么运转的时候注意到一件事，我开始想：国家层面的商业文化到底怎么影响一支哈萨克斯坦团队。因为仔细想想，不同的商业是有民族文化的。比如日本的商业文化大家都知道：严格的层级、非常长期的规划、几十年才建立起来的关系——就是造一台非常精密的机器。我们也知道中国的商业模式，同样是那种雄心勃勃的巨型项目、巨型计划。人可能没有日本公司那么有组织，但同样是大规模地协调大量的人。

And we know the Anglo-Saxon business model. That's, probably, the classic capitalist business model, where the most business— the most aggressive business model, where people are ready to take high risks, people are ready to take, uh, and are ready to close very complex deals, to structure the most complex deals. It's clear, like, how Americans structure deals, how they find compromises, how out of, uh, some incredible configurations, when it seems impossible to reach an agreement, Americans understand how all of it can be structured and how all of it can be put together. That is, that classic Anglo-Saxon model. And watching how we work, uh, it seems to me, something comes out— something like a Kazakh business culture, because I see that in such a volatile market, uh, when you need to react quickly to changes, when we see that some new opportunities have opened up in the market, when, like, a giant has stumbled and there's a very small, short window of time when you can attack, then I see that our, like, our guys, like, our employees, engineers, managers, like, the whole team, it coordinates very fast. And we don't have this thing where we have a hierarchy, where, like, I need to gather everyone, tell them, like, motivate them, that here an opportunity has appeared, the window is very short, let's, like, all get together, let's build a plan, you'll do this, like, this team will do that, and I somehow have to conduct it— not like that at all. The people themselves understand right away.

我们还知道盎格鲁-撒克逊的商业模式。那大概是经典的资本主义商业模式，是最有攻击性的商业模式：人们愿意承担高风险，愿意去谈、去做非常复杂的交易，把最复杂的交易结构化。很清楚，美国人怎么设计交易结构、怎么找到妥协点、怎么从那些看起来根本谈不拢的离谱组合里，找出办法把它结构化、把它拼起来。这就是经典的盎格鲁-撒克逊模式。而看着我们自己怎么工作，我觉得跑出来的是某种哈萨克式的商业文化。因为我看到，在这样一个波动的市场里，需要对变化快速反应，当我们看到市场上出现了新机会、某个巨头绊了一跤、只有非常小非常短的一段时间窗口可以进攻的时候，我看到我们的人——我们的员工、工程师、经理，整个团队——协调得非常快。我们没有那种层级：我需要把所有人叫来、告诉他们、激励他们，说机会来了、窗口很短，大家聚一下、做个计划，你做这个、这个组做那个，还要我来指挥——完全不是这样。人们自己马上就明白。

It turned out to be the most self-organizing structure.

结果就成了最自组织的结构。

The most self-organizing structure, which coordinates fast and mobilizes fast. And the mobilization happens in— in small teams. That is, the people themselves understand how to split up, who has which strengths, weaknesses. People, uh, split into, like, teams of 10, of 5, of 15 people. And each team understands what needs to be done. And there's no need to, like, conduct or orchestrate anyone. Everyone understands what to do. And the approach is maximally like this, that while the competitors are, like, going off to rest on Sunday, like, going to sleep at night, we're, like, coordinating, we understand that there's an opportunity, and we attack, and— and we get the growth percentages that we need for growth, to capture, to capture percentage points of the market. And it seems to me this is inherent to that kind of, well, what's in our DNA is that nomadic military culture, in that there's high self-organization, high mobilization, in that the structure of, like, a nomadic army is always, like, units of ten thousand, 10,000. That kind of high-level self-organization and mobilization and, most importantly, speed. Uh, because, well, we know that a nomadic army, it— it always won through its speed and its mobility. This is just something I noticed in the team when there was simply time to, like, reflect. And this is, in my view, also pretty cool. And it seems to me that when, well, Kazakhstani companies go out further onto world markets, capture world markets, this could, like, in the future also become our calling card, the way, like, Japanese companies are treated with respect, with the greatest, right, respect for Korean companies, and, obviously, for Anglo-Saxon companies.

最自组织的结构，协调快、动员快。而且动员发生在小队伍里。人们自己就明白怎么分工、谁的强项弱项在哪。大家分成 10 人、5 人、15 人的小组，每个组都知道该做什么。不需要谁去指挥、去编排。每个人都知道自己要做什么。我们的做法拉到极致：竞争对手周日去休息、晚上去睡觉的时候，我们在协调，我们知道机会在那儿，我们就进攻——于是拿到我们增长所需要的增长百分点，抢下市场的百分点。我觉得这是我们 DNA 里就有的东西——那种游牧的军事文化：高度自组织、高度动员，游牧军队的建制永远是万人、一万人。那种高水平的自组织和动员，最关键的是速度。因为我们都知道，游牧军队靠的就是速度和机动性取胜。这只是我在有时间反思的时候，从团队里注意到的。我觉得这也挺酷的。我还觉得，等哈萨克斯坦公司往后走向世界市场、去抢占世界市场，这在未来也可能成为我们的名片——就像大家很尊重日本公司、极其尊重韩国公司，当然也尊重盎格鲁-撒克逊公司。

Well, that would be its own kind of Kazakh business model then, right.

那到时候就是一套自己的哈萨克商业模式了，对吧。

Yes, yes. Culture, specifically.

对，对。就是文化。

Culture specifically, yes, its own. Very interesting to watch all of this, besides when you'd be a participant in the process,

就是文化，对，自己的文化。观察这一切非常有意思，何况你还是这个过程的参与者，

yes, yes, when you're working. And at the moment when you zoom out a little bit like that, you look at what's happening, that always adds drive too, that we're doing something big, and that also motivates the team.

对对，一边做着。而在某个时刻稍微拉远一点看正在发生什么，这总能多给一点劲头：我们在做一件大事，这也能激励团队。

Mm-hm. [clears throat] So

嗯。[清嗓子] 那

I'm also listening to the answers here, and I understood for myself that inside Higgsfield there will appear specialists who can then create unicorns themselves. That is, you want to keep watching their development further. You'll help, right, essentially, even, like, maybe even act as investors to some, well, to some extent, in order to increase your influence or your returns. And what directions, do you think, might be, well, like, the next successful companies in our country? — Of course, my focus is on AI. Uh, and indeed, AI right now is that kind of locomotive of growth for the world economy. That is, we understand that the American economy is now growing thanks to AI, thanks to the fact that labor productivity has risen sharply and thanks to big, uh, gigantic investments in infrastructure, in data centers, in energy, for all of which demand appeared thanks to, uh, the fact that AI boosted the productivity of the economy, and AI right now, uh, is directing the growth of, like, the whole world economy. And, as I said earlier, right, talent is very important, the human resource, human capital, is important. And there are opportunities too, because I recently saw my nephews, and they're all preparing for NIS.

我这么听着这些回答，自己琢磨出来一点：Higgsfield 内部会长出一批以后自己能造独角兽的人。也就是说，你们还想继续看着他们发展，你们会帮忙，甚至可能在某种程度上以投资人的身份出现，来放大自己的影响力或者收益。那你觉得，我们国家接下来可能在哪些方向出现下一批成功的公司？——当然，我的关注点在 AI。而且 AI 现在确实是世界经济增长的火车头。我们都清楚，美国经济现在的增长靠的是 AI，靠的是劳动生产率的骤增，靠的是对基础设施、数据中心、能源的巨额投入——这些需求都是因为 AI 拉高了经济的生产率才出现的，AI 现在牵引着整个世界经济的增长。而正如我前面说的，人才非常重要，人力资源、人力资本很重要。机会也是有的，比如我最近见到我的侄子们，他们都在准备考 NIS。

Mm-hm. And

嗯。而

everyone's trying to invest as much as possible in their children, so that they get, like, a great education, so that they then get into Nazarbayev University, like, go on Bolashak, get their degree. Human capital is being formed. And further, since, as I said earlier, the market is volatile, uh, and it's still an early stage, uh, in AI, uh, and still, uh, classic markets, they're very inertial. And we know that, like, the internet appeared, like, in the late nineties, there was already a boom— businesses, people were making websites back then, like, here, when I had just finished school, I was working part-time, making websites back in, like, 2013–2014, even though the internet had already existed for, like, 20 years, and there was still demand for, like, website development. Well

大家都在尽最大努力往孩子身上投入，让他们受好的教育，之后能考进 Nazarbayev 大学，去 Bolashak 项目留个学、把书念完。人力资本正在形成。再往下说，就像我前面讲的，市场是波动的，AI 还处在很早的阶段；何况经典市场是非常有惯性的。我们知道互联网是九十年代末出现的，那时候已经有热潮了——做生意、做网站；我刚中学毕业那会儿还打零工做网站呢，那都 2013、2014 年了，互联网都存在 20 年了，做网站还是有需求。就

back then everything was slower, I think. Now it's much faster. In my opinion, it won't drag on for that long anymore,

那时候一切都慢一些，我觉得。现在快多了。我觉得不会再拖那么久了，

of course, yes. Speeds have 100% increased a lot, but, uh, the window of opportunity is still there anyway, that's why — and the windows of opportunity you see specifically in AI, right, and some projects will still appear, right. — In my view, this, yes, this is very important, because when you study, like, the Asian tigers, for example, right, there was the electronics boom

当然，是的。速度 100% 快了很多，但机会窗口毕竟还在，所以——你看到的机会窗口正是在 AI 上，对吧，还会冒出一些项目，对吧。——在我看来，这非常重要，因为你去研究亚洲四小龙就会发现，当时有一波电子产品热潮，

Samsung, like, Korean companies, Japanese companies, which, like, became such a serious foundation of national economies, like, of Japan, of South Korea, these were new economies. And this— more precisely, these were new markets. That is, like, the electronics market appeared, and for a new country that's a new opportunity, and they captured it. And it's very important for a country to see these windows, because, well, it's hard to build something, practically impossible under capitalism, on some old, uh, economies, uh, on old markets, to disrupt the old players, because everything gets monopolized fast, supply chains, uh, they, well, become super-optimal. Super— there isn't much space for— for further optimization, for disruption. That's why, like, Germany, like, England built so many gigantic companies during, like, the industrial revolution. And now that they're already, like, 200-something, like, 300 years old, it's impossible to disrupt them— they're kind of the foundation of those economies. Same thing with the Asian tigers. There was an opportunity with the markets of, like, uh, car manufacturing, with the markets of household, like, appliances, electronics. They managed to capture those markets, they managed to build the supply chain. And now you have to understand that the entire AI market depends on one country, on Taiwan. And in Taiwan there's a company, TSMC — these are the fabs that produce the chips for the GPUs that all the neural networks run on. That is, it's just one company, it has a monopoly over the whole world. Mm-hm.

三星，还有韩国公司、日本公司，它们后来成了日本、韩国这些国家经济的重要支柱。这些是新经济——更准确说，是新市场。也就是说，电子产品这个市场出现了，对一个新国家来说这是新机会，他们把它抢下来了。对一个国家来说，看见这些窗口非常重要，因为在资本主义下，你很难、几乎不可能在旧经济、旧市场里去颠覆老玩家，因为一切都会很快被垄断，供应链会变得超级优化，没剩下多少空间可以继续优化、可以颠覆。所以德国、英国在工业革命时期造出了那么多巨型公司，而它们现在都两百来年、三百年了，撼动不了，它们就是这些经济体的底座。亚洲四小龙也一样。当时机会在汽车制造的市场、在家电和电子产品的市场。他们抢下了这些市场，建起了供应链。现在要明白的是，整个 AI 市场依赖一个国家——台湾。台湾有一家公司叫 TSMC，就是那些工厂，生产所有神经网络赖以运行的 GPU 芯片。也就是说，就这一家公司，垄断了全世界。嗯。

And it's located in Taiwan. Uh, and one company, it turns out, controls the whole market. That's why, like, there's now the geopolitical— America, China, they're, like, essentially already in a cold war between the two countries, and Taiwan is kind of the center of the whole conflict. And, uh, all of this came about because an opportunity appeared with the electronics market and Asian companies managed to jump into that window in time, to build cool companies that became the foundation of their national economies. And moments like that, they happen very rarely, when such big markets grow sharply. That's why, of course, for Kazakhstan this is a big opportunity. Uh, and we have talent. And now the task is that, while the market is being redivided, while this market is still new, there's this short window when Kazakhstani companies can emerge and, uh, embed themselves into the world economy. And AI in the future, and already now, is a large part of the world economy.

它就在台湾。一家公司，等于控制了整个市场。所以现在地缘政治上，美国、中国之间实际上已经是冷战，而台湾等于是整个冲突的中心。而这一切的起点，就是电子产品市场出现了机会，亚洲公司及时跳进了那个窗口，建起了很强的公司，成了他们国家经济的底座。这样的时刻非常少见——这么大的市场突然长起来。所以对哈萨克斯坦来说，这当然是大机会。我们有人才。现在的任务是：趁市场还在重新洗牌、趁这个市场还新，有这么一个短窗口，哈萨克斯坦的公司可以冒出来，嵌进世界经济里。而 AI 在未来、其实现在就已经是世界经济很大的一块。

A cool— cool idea, a very interesting thought, that right now we really can build something big and super successful in this market and be, well, just like Nvidia took its place, so to speak, in the world. So something like that could emerge in Kazakhstan too. Interesting. That's uh a good thought. And what could get in the way? What, in your opinion, is not set up right in our country? I don't know, like, maybe there isn't enough electricity, or maybe corruption gets in the way, or what? What— what downsides are there that could hinder development?

很酷、很酷的想法，非常有意思的思路——现在我们真的可以在这个市场上造出很大、很成功的东西，就像 Nvidia 在世界上占住了自己的位置，哈萨克斯坦也可能冒出这么个东西。有意思。这个想法不错。那什么可能会碍事？在你看来，我们国家有什么地方安排得不对？我不知道，比如电力不够？或者腐败拖后腿？还是别的什么？有哪些负面因素可能妨碍发展？

Globally, it seems to me, there are practically no downsides. We— probably the only downside is the ecology in Almaty. Uh, because, obviously, people, uh-uh, who, well, do that kind of highly intellectual work, they can at any moment, like, choose any country to live in, because they're in demand everywhere. They can at any moment, like, move to London, San Francisco, Tokyo, uh, Seoul, wherever— wherever they want. They're very much in demand, so ecology for them is, of course, a big factor. But there are, of course, also pluses in that, uh, well, obviously, like, it's the home country, the home culture; those who have children want the children to learn the Kazakh language, to go to a Kazakh, like, school, uh, and, like, to have our, like, local culture. Mm-hm. That's a big factor. And also a relatively low cost of living, relatively— well, obviously, like, compared to Silicon Valley, London, of course, well, it's cheaper to live in Almaty, so

整体看，我觉得几乎没有什么减分项。唯一的减分项大概是阿拉木图的环境。因为很清楚，做这种高强度脑力工作的人，随时可以挑任何一个国家生活，因为哪儿都要他们。他们随时可以搬去伦敦、旧金山、东京、首尔，想去哪儿去哪儿。他们太抢手了，所以环境对他们当然是个大因素。但当然也有加分项：这是自己的国家、自己的文化；有孩子的人希望孩子学会哈萨克语、上哈萨克语学校，接住我们本地的文化。嗯。这是很大的因素。还有相对低的生活成本——当然是相对硅谷、伦敦而言，在阿拉木图生活便宜，所以

it's still cheaper for now, though. Still cheaper for now.

目前还便宜。目前还便宜。

Yes, the cost of living, it is, of course, yes, rising, but at the same time relative to San Francisco— it's just that San Francisco is the most expen— expensive city in the world to live in. There, just renting a room, not even an apartment, but one room without, like, a bathroom. Uh, it costs around, like, 4–5,000 dollars a month.

对，生活成本当然是在涨，但相对旧金山来说——旧金山就是全世界生活成本最贵的城市。在那儿就租一个房间，都不是一套公寓，就一个不带卫浴的房间，一个月大概要 4000–5000 美元。

That is, it's, well, the most expensive city in the world to live in. And obviously, uh, we pay competitive salaries already on the global market. That is, we're not looking at, like, the Kazakhstani market anymore, we pay, like, more than in London, more than in Europe. Uh, and with stock options, total compensation at the company is comparable to Silicon Valley companies. At the same time the cost of living is much lower. And that's, of course, a big factor, because, well, you have to understand, if, like, a household, a family wants, for example, to have children, that costs a lot in the Western world. And so, like, engineers who are considered— who get a good salary, like, total compensation, I don't know, half a million dollars, for example, a family can have, if, like, two engineers in the Valley.

也就是说，那是全世界生活最贵的城市。而很清楚，我们付的薪水在全球市场上已经是有竞争力的。我们已经不看哈萨克斯坦市场了，我们给的比伦敦高、比欧洲高。算上股票期权，公司的 total compensation 跟硅谷公司是可比的。同时生活成本要低得多。这当然是个很大的因素，因为要明白，一个家庭如果想生孩子，在西方世界成本非常高。所以那些薪水不错的工程师——total compensation，比如说 50 万美元，一个家庭如果两口子都是硅谷的工程师，可能就是这个数。

But to have a child, to have several children, the way we do here in Almaty — you can do it here — over there it's practically unrealistic. It's very expensive. And that's because quality kindergartens, quality education in America, that's only for rich people, and even for the middle class it's quite expensive. That's actually a big factor. Including that, thanks to our demographic boom, thanks to there being so many kids, of course services are more efficient, more optimized, and prices are lower. Way more so when it comes to having kids. That's also a factor when we're hiring, and for people like that, people pushing 30, who are already thinking it's time to start a family, it's a strong factor that, for example, people come back from London, from the US to Almaty, because their parents are here, here you can have three or four kids and, uh, not be afraid that you could go completely bankrupt, the way it is in the Valley in America. That really is, well, also a plus of our economy.

但要生孩子、生好几个孩子，像我们在阿拉木图这样——在这儿做得到；在那边几乎不可能。太贵了。因为在美国，好的幼儿园、好的教育只属于有钱人，就算对中产也相当吃力。这其实是个很大的因素。包括说，靠我们这边的人口红利，靠孩子这么多，服务当然更高效、更优化，价格也更低。生孩子这件事上差得更多。这也是个因素——我们招人的时候，碰到快 30 岁、已经在想该成家的人，这就是很强的一条：比如有人从伦敦、从美国回到阿拉木图，因为父母在这儿，在这儿能生三四个孩子，还不用怕像美国硅谷那样彻底破产。这确实也算我们经济的一个加分项。

Do you have kids?

你有孩子吗？

Uh, not yet. Uh, I got married, uh, in twenty twenty-four. And, well, of course we're planning to, uh, but yeah, it's hard, of course, to combine that with the company.

还没有。我是 2024 年结的婚。当然，我们是有计划的，但确实，跟公司这边兼顾起来挺难。

Well, it can be combined, in my view. Now I'd like to talk a bit about your vision of economic models. We do have an economics theme going here, right? You know, I've noticed that a lot of people your age for some reason, uh, become sort of, they get nostalgic for the Soviet Union, even though they never lived in in Soviet in Soviet times, exactly people of your age, like people who are 30, 30 to 35, let's say, right, years old. Why does this happen? I just want to put this question to your generation.

我觉得是可以兼顾的。接下来我想聊聊你对经济模式的看法。我们这期毕竟是经济主题，对吧？你知道吗，我注意到很多跟你同龄的人不知为什么会开始怀念苏联，尽管他们根本没在苏联时期生活过——就是你这个年纪的人，30 到 35 岁上下吧。这是为什么？我就是想把这个问题抛给你们这一代。

Well, that's actually a surprise to me, because in my circle there are no people like that.

这对我来说其实挺意外的，因为我身边没有这样的人。

No people like that. Yeah, in my circle there are no people like that. And, well, the Soviet Union, well, it's very important. It really is important. And obviously, for geopo- political reasons we can't push really hard on, uh, what we can do in schools. And this hasn't happened for us yet, that the school curriculum fully explains to kids that, uh, a gigantic number of Kazakhs died, right, by some estimates 50% of the population, from the fact that, uh, well, the basic nomadic economy there was simply destroyed,

没有这样的人。对，我身边没有这样的人。至于苏联——这事真的很重要，确实重要。而且很明显，出于地缘政治的原因，我们没法太用力去推学校里能做的那些事。我们这边还没做到：在学校课程里把这段彻底讲给孩子听——有大量哈萨克人死掉，对，按有些估计是全国人口的 50%，原因就是最基础的游牧经济被摧毁了，

the livestock was confiscated, people were forcibly made to settle, and what was it all for? So that the Soviet Union could buy factories in America in gigantic quantities. We all know that all the gigantic Soviet plants were designed in the USA and shipped here. American corporations did this, and they were paid with that same livestock, that same grain. It was a very inefficient, bloodthirsty policy.

牲畜被没收，人被强制定居，这一切是为了什么？为了让苏联能在美国大批量采购工厂。我们都知道，苏联那些巨型工厂都是在美国设计、再运过来的。是美国公司做的，付账就拿那些牲畜、那些粮食。这是一套非常低效、非常嗜血的政策。

Mm-hm. Unfortunately, yeah, our people aren't fully taught this. Maybe that's the reason, because, yeah, even those who know that livestock was confiscated, people don't fully understand that the livestock was confiscated in order to move a tractor plant over from Detroit, which then, uh, during the Second World War.

嗯。可惜的是，我们这边没把这段彻底教给大家。也许原因就在这儿——就算知道当年没收牲畜，人们也没真正明白：没收牲畜是为了把底特律的拖拉机厂搬过来，那个厂后来在二战的时候……

Well, why does the youth talk about this topic? I dug into it a little, I asked, because from their point of view everything was clear, they were socially protected, there were always jobs. And there was always some kind of minimum, you were always provided for. Basically the state decided everything for you. And for, well, some people — I won't say for many, for some — that looks quite comfortable. Here, probably, we need to draw a distinction. Uh, the market economy also has problems. The market economy has a lot of externalities, a lot of, uh, ugly things, uh, side effects, the market economy has them. And this really is a challenge for Kazakhstan, how we as a country adapt to it. Because, for example, well, we see, South Korea became a cool, rich, high-tech country, but the country is disappearing before our eyes, the demographics are collapsing. And what's the point, well, of being such a rich country when in two or three generations it will, well, practically disappear entirely. That's kind of a product of capitalism. And capitalism has a lot of these side effects, uh, bad ones. But separately you have to be clear-eyed about the problems of the damage that communism, uh, did to Kazakhstan, to our population there and so on. These questions have to be separated, because it's very important, in my view, to give an honest historical assessment of everything that happened. There's this notion of survivorship bias. There are people who ended up as beneficiaries, the redistributions of property and, yeah, if some of our classic bais were dekulakized, then some kind of nomenklatura appeared in the country that was the beneficiary. And naturally, by survivorship bias, well, I don't know, the classic urban Almaty Kazakhs, right, the ones who are also called asphalt Kazakhs, uh, they were kids back then, uh, beneficiaries of the Soviet regime. Well, possibly, yeah, some portion of them. And maybe, yeah, not all of them, of course. Well, possibly people who heard from their grandmothers, their grandfathers how good things were. Well, they heard the good part, [clears throat] but they didn't hear about how people ended up in the Gulag, that is, how they suffered under the Soviet regime. So there is, of course, survivorship bias, but globally there's a completely different side. That's also why in the West left-wing ideas are very popular among the youth.

年轻人为什么会聊这个话题？我稍微深挖过、问过。因为在他们看来一切都很清楚：有社会保障，工作岗位永远都有。总有个最低保障，你什么都不缺，基本上国家替你把一切都定好了。对某些人——我不说很多人，就是某些人——这看上去挺舒服的。这里可能得做个区分。市场经济也有它的问题，市场经济有很多外部性，有很多不好看的东西、副作用。这对哈萨克斯坦确实是个挑战：我们作为一个国家怎么去适应。比如说，我们看到韩国成了很厉害、很富、高科技的国家，但这个国家正在我们眼前消失，人口在往下掉。那这么富又图什么呢？两三代人之后基本就没了。这算是资本主义的产物。资本主义有很多这种副作用，很糟糕的副作用。但另一方面，也得清醒地看到共产主义给哈萨克斯坦、给我们的人口造成的伤害。这两个问题必须分开，因为在我看来，对发生过的一切给出诚实的历史评价非常重要。有个概念叫幸存者偏差。有些人最后成了受益者——财产被重新瓜分，对，我们那些传统的巴依被当成富农清算了，国内就冒出某个官僚阶层，他们才是受益者。自然，按幸存者偏差来看，我不知道，那些城里的、典型的阿拉木图哈萨克人，对，还被叫作「柏油哈萨克人」的那批，他们当年是小孩，是苏联体制的受益者。也可能只是其中一部分吧，当然不是所有人。也可能是那些从爷爷奶奶那儿听说以前多好的人——他们只听到了好的那一半，[清嗓子] 没听到有人是怎么进古拉格的，也就是他们在苏联体制下怎么受的苦。所以幸存者偏差当然存在，但放大了看还有完全另一面。这也是为什么在西方，左翼思想在年轻人里非常流行。

Now in America, say — America is a super-capitalist country, but left-wing ideas there are, well, quite popular,

比如说美国——美国是超级资本主义的国家，可左翼思想在那边还挺流行的，

yeah. Well, Bernie Sanders came second in the election before last. I mean it's kind of, well, he's an honorary citizen of Kali- Kaliningrad, I think, actually. He he really did, back in the Soviet Union he

对。Bernie Sanders 在上上届大选里拿了第二。他好像还是加里……加里宁格勒的荣誉市民。他确实早在苏联时期就……

now that needs checking.

这个得核实一下。

This is 100%, yeah, when I read about it, he traveled to the Soviet Union, right, he came to the Soviet Union with his wife on their honeymoon. I mean that's straight-up, well, a communist with maximally left-wing views. And he came second in the election. Mm-hm.

这个百分之百，我读到过，他去过苏联，对，他和妻子度蜜月就是去的苏联。就是那种非常彻底的、观点极左的共产主义者。而他在大选里拿了第二。嗯。

And of course this is a reaction to that kind of capitalism, to what's called, uh, aggressive capitalism, when under full-on capitalism, obviously, a rapid concentration of resources happens. Everyone likes to say that Norway is such a balanced country.

当然，这是对那种资本主义的反弹，就是所谓的激进资本主义——在纯粹的资本主义下，很明显会发生资源的快速集中。大家都喜欢说，挪威是那种很平衡的国家。

Norway is practically socialism, honestly speaking,

说实话，挪威基本上就是社会主义，

what, yeah, socialism. But if you look at the distribution coefficient of wealth, specifically of wealth, then Norway's Gini indexes are roughly the same as Kazakhstan's. That's in the official statistics, that is, just like in Kazakhstan, uh, a small concentration of people owns a large amount of the wealth. Same thing in Norway if you look at households; obviously, they're leveled out by income, in the sense that wages are leveled out. That is, there's no such thing as very high wages and no such thing as very low wages. Wages are very leveled out through taxes, through a competent welfare state. People really do feel more socially protected in Norway, absolutely. But if you look specifically at the distribution, uh, of wealth, of people's wealth, then Norway is just as unequally distributed by the Gini coefficient.

对，是社会主义。但你要看财富的分配系数，专门看 wealth，挪威的基尼系数跟哈萨克斯坦差不多。官方统计就是这样——跟哈萨克斯坦一样，一小撮人握着很大一部分财产。挪威按家庭户看也一样；当然，他们在收入上是拉平的，工资拉平了：不存在特别高的工资，也不存在特别低的工资。工资靠税收、靠一套成熟的 welfare state 拉得很平。人在挪威确实感觉更有社会保障，这毫无疑问。但如果专门看财富的分配，挪威按基尼系数一样是不均的。

So there are the super-rich, there are very few of them, there are the super-rich and, like, everyone,

也就是说有超级富豪，人数非常少，有超级富豪，然后就是所有……

well, like, the middle class,

就是中产阶级，

yeah. So even Norway, which is considered such a classic welfare state,

对。所以连一向被当成经典 welfare state 的挪威，

yeah,

对，

which is, well, capitalism with a socialist face, as they might call it. But globally, yeah, in America it's not like that. In America it's this extreme capitalism. And America essentially exists thanks to immigrants who keep arriving in the country, and there's constantly this new shot of doping happening in the econo-.

就是所谓带社会主义面孔的资本主义。但整体看，美国不是这样。美国是那种极端资本主义。美国本质上是靠移民活着的：移民不断进来，经济里就不断有新的兴奋剂注进来……

The American dream — everyone has equal chances to realize their dream, this,

美国梦——每个人都有平等的机会去实现自己的梦想，这个……

yeah. But at the same time we see, there was the idea of the middle class, uh, the classic kind, that basically liberal economies had won. There's the concept of the middle class. The middle class is the foundation of the economy. The middle class lives well. There's that same American dream: a house, a car, a washing machine. And we see that now, under late-stage capitalism, the destruction of the middle class is happening. In America the middle class is collapsing very fast, and the economy exists not thanks to the middle class but thanks to immigrants, because new immigrants keep arriving, and they build new companies. Most startups are founded by immigrants — even Google itself, uh, and WhatsApp, immigrants did all of that. I've read those studies too and I was told that the fact that the middle class is shrinking, right,

对。但同时我们看到，原来有一套中产阶级的设想，很经典的那种，基本上就是自由主义经济赢了。有中产阶级这个概念，中产阶级是经济的基础，中产阶级过得不错。就是那个美国梦：房子、车子、洗衣机。而我们看到，现在在晚期资本主义下，中产阶级正在瓦解。美国的中产阶级塌得非常快，经济不是靠中产阶级撑着，而是靠移民——新移民不断进来，他们在那儿建起新公司。大多数创业公司都是移民创办的，连 Google 本身，还有 WhatsApp，都是移民做出来的。这些研究我也读过，也有人跟我说，中产阶级在萎缩这件事，对，

that's a harbinger of the fact that there could be, uh, some kind of popular unrest. Right up to that — in the US they've even talked about that.

这是可能爆发某种民众骚乱的前兆。在美国甚至已经有人说到这一步了。

Yeah. Yeah. And we see the same thing in Europe. And Europe also exists thanks to, uh, immigration. Here, by the way, it was interesting, I once watched a talk by Likmanyu at a university, when students came and asked him questions, and a female student came up, uh, and she was clearly unhappy with the immigration policy, that a lot of immigrants are coming. And she goes: "So, like, I have 3 PhDs."

对，对。欧洲也一样。欧洲同样是靠移民活着的。顺带说个有意思的：我有一次看 Likmanyu 在大学的一场演讲，学生上去提问，有个女学生上来，她显然对移民政策很不满，觉得来的移民太多了。她张口就说：「我有 3 个 PhD。」

Uh-huh. “I've finished my studies, I'm a highly specialized, like, specialist.”

嗯。「我读完书了，我是高度专业化的专家。」

“And I've got, like, a PhD in biology, like, in and in all the cool sciences. I share our Singaporean high values. And I see that a large number of migrants are arriving who do not share these values. This is going to destroy the country, our, like, values are under threat, what are we going to do.” And Lee Kuan Yew asked her: “And how many children do you have?” And, well, she said: “Well, I was studying, I don't have children.” And he says: “Well, there you see, and globally, well, we'll collapse as a country, because, well, you have the chance to study, you have the chance to do all of these things, but the economy doesn't work that way. There are a lot of jobs that are needed, like, with less, like, high education, and immigrants are what make the country sustainable,” and she, like, didn't, well, she didn't know what to say. And so, like, this demonstrates that in rich economies like that, highly educated ones, like South Korea, the very same thing, uh, there is this threat that, uh, the demographic index falls, the birth rate falls, and the country rapidly collapses, and the middle class disappears along with it. This really is a big challenge for Kazakhstan. We, obviously, hope and want that Kazakhstan, like, by the end of this century, that we'd become like South Korea, such a super-developed economy, so that there'd be a high standard of living, like, for the entire population. But under threat, of course, there are these side effects, uh, of classical capitalism. And it's really interesting how we, uh, will meet this. Here, probably, it seems to me, a big factor as an antidote to this kind of classical capitalism may be that we still have fairly strong clan structures. And in that, well, you have to understand, the classical capitalist arrangement of society, it's very granulated. That is, this classical family cell, where there's a mom, a dad and the children, this appeared, uh, fairly recently. Humanity never lived that way before. Such a family cell is precisely a product of capitalism, because capitalism, well, it's built on the division of labor, in that, like, you can bake bread, another person, like, can, uh, like, transport people, and a large number, each one in his own interests individually does something, but globally everyone wins, unlike communism, where a centralized system runs everything, like, top-down, and a person at the individual level doesn't choose, and he doesn't have a lot of freedom of choice, of initiative,

「我还有生物学的 PhD，各种厉害的学科都有。我认同我们新加坡这套崇高的价值观。可我看到大量移民涌进来，他们并不认同这些价值观。这会毁掉国家，我们的价值观受到威胁，我们该怎么办。」李光耀反问她：「那你有几个孩子？」她说：「我一直在读书，我没有孩子。」他说：「你看，从全局看，我们作为一个国家会塌缩，因为你有机会读书，你有机会做这一切，但经济不是这么运转的。有很多岗位需要的是学历没那么高的人，是移民让这个国家撑得住。」她当时不知道该说什么。这恰恰说明，在这类富裕、高学历的经济体里——韩国也一样——存在这么一个威胁：人口指标下滑，生育率下滑，国家迅速崩塌，中产阶级也跟着消失。这对哈萨克斯坦确实是个大挑战。我们当然希望，到本世纪末哈萨克斯坦能变成韩国那样的超发达经济体，让全体国民都过上高水准的生活。但受威胁的地方就在于，古典资本主义有这些副作用。我们会怎么迎上它，确实很有意思。我觉得，对抗这种古典资本主义，一剂很大的解药可能是我们还保留着相当强的家族结构。要知道，古典资本主义的社会结构非常颗粒化。就是说，爸爸妈妈加孩子那种经典的家庭单元，是相当晚近才出现的。人类从来不是这么生活的。这种家庭单元恰恰是资本主义的产物，因为资本主义建立在分工上：你可以烤面包，另一个人可以拉人跑运输，每个人各按自己的利益单独做事，但从全局看所有人都赢；不像共产主义，一切都由自上而下的中央系统管着，个人层面没得选，没有多少选择和主动的自由，

yes, on his own initiative.

对，凭自己的主动。

This is

这就是

maximally, uh, classical individualism, in that, like, that same Ayn Rand, uh, propagated it heavily, like, Atlas Shrugged and so on. Its side effect is that, uh, these family ties get destroyed, uh, and everyone becomes — first this cell appeared. We see now that in the Western world, uh, there are giant, like, divorce rates. People get married very late, like. And basically Western society, it's already, like, not a family cell but individuals, globally. Like, in big cities people, like, cohabit, uh, already not with family, but, like, just with people.

……极致的古典个人主义，就是安·兰德当年大力鼓吹的那种，《阿特拉斯耸耸肩》之类。它的副作用是，家庭纽带瓦解了——先是有了这个家庭单元。我们现在看到，西方世界的离婚率高得吓人。人结婚很晚。西方社会基本上已经不是家庭单元，而是一个个个体。大城市里，人合住在一起的已经不是家人，就是随便一些人。

And capitalism forces this hard. And there are big problems in this too. Why, for example, did, like, that same Marx and, like, the leftist, like, philosophers, uh, why did they, like, criticize capitalism? Because through this individualization, uh,

资本主义把这个推得很猛。这里面也有很大的问题。比如马克思，还有那些左翼哲学家，他们为什么批判资本主义？正是因为这种个体化，

well, Marx put it even harder: “There is no crime that capital would not commit for 300% profit.”

马克思说得还更狠：「没有一种罪行是资本为了 300% 的利润不敢犯的。」

Yes. Well, well, that's the direct criticism. But there's also the criticism of why all of this can quickly turn into fascism. Because an individualistic society, it's fairly weak, when everyone is focused on himself. A person, if he runs into some injustice with this Leviathan, with the state — if there's no strong community, and usually a community has always, well, been the extended family, then the state can simply

对。这算是直接的批判。还有一种批判是：为什么这一切会很快滑向法西斯主义？因为个体化的社会相当脆弱，每个人只盯着自己。一个人要是跟利维坦、跟国家撞上了，遇到某种不公，如果没有强大的共同体——而共同体通常一直就是大家族——那国家可以直接把他

grind him up,

碾碎，

grind him up, and nobody will protect him. And such side effects

碾碎，没人保护他。这类副作用

are present too. And globally, like, the demographic rate is falling. And we, like, see how in South Korea now, how in Japan, everything is, like, getting annihilated. So here maybe it's our culture, uh, of our, like, big clans, of our — the thing that, like, Kazakhs love toi, right, for instance, like, getting together. This is a really important institution, in that,

也同样存在。而且从全局看，人口指标在下滑。我们看到韩国现在、日本，一切都在湮灭。所以这里，也许能靠我们的文化，靠我们那些大家族——哈萨克人爱办 toi，对吧，比如说，大家聚一聚。这是个非常重要的制度，因为

uh, everyone, like, several times a year all the families get together, uh,

一年里全家族要聚好几次，

a community gets built, uh,

共同体就这么长出来了，

a way of transmitting information, a way of transmitting knowledge, norms of behavior. This is very important,

这是传递信息的方式，是传递知识、传递行为规范的方式。这非常重要，

yes, this is very important. And we know, for example, like, in Israel, for instance, it's a very developed capitalist country, but at the same time it also leads on demographics. Like, on average per household, five children per Israeli household, and that this national culture of theirs, they very strongly have big families, even that

对，这非常重要。比如以色列，那是个非常发达的资本主义国家，但同时人口上也领先。平均每个以色列家庭五个孩子，他们的民族文化里多子女家庭非常强势，甚至

yes it gets pushed, and they even have this interesting structure: they have these very deeply religious conservative families, and there are, uh, liberal Jews — usually liberal Jews, they come out of the conservative ones, that is, a family, for example, has five children

对，这是被使劲推着的，而且他们的结构还挺有意思：有非常虔诚保守的家庭，也有自由派犹太人——自由派犹太人通常出自保守家庭。比如一家有五个孩子，

they all, like, got a good education. Like, three stayed conservative and also keep, like, the demographics going. Two become liberal, like, professors somewhere, like, building companies. And this balance stays with them. In that there's very

都受了很好的教育。三个留在保守那边，继续把人口撑住；两个变成自由派，去哪儿当教授、去办公司。这个平衡就一直保持着。就是说有非常

over there lots of kids, and here one child each, right,

那边多子女，这边一家一个孩子，对吧，

there's a very conservative wing, there's a liberal wing. So, well, for instance, there's the Israeli model. It's interesting, of course, what the Kazakh model will be like, but globally this is also such a challenge, like, under capitalism. We, we hope that Kazakhstan will get richer, that the standard of living will rise, but that these side effects will pass us by thanks to, like, our culture as well, our institutions, uh, that, like, formed with us over centuries, and the Soviet Union couldn't destroy them.

有非常保守的一翼，也有自由派的一翼。这就是以色列模式。当然，哈萨克模式会是什么样，很有意思，但从全局看这也是个挑战，资本主义之下的挑战。我们希望哈萨克斯坦越来越富，生活水平越来越高，但这些副作用能绕开我们——靠的也是我们的文化、我们的制度，那些在我们这儿形成了几百年、连苏联都没能摧毁的东西。

Uh-huh.

嗯。

And here, of course, we probably need, like, new Kazakh philosophers, like, who will write about this, research this, and we'll, like, find the answers to all these questions. And,

这里当然大概需要新一批哈萨克哲学家，来写这些、研究这些，我们才能找到所有这些问题的答案。还有，

well, I think that's how it'll be. And we do have philosophers, and we have political scientists, right, all of this needs to be talked through, this needs to be talked about. I'm listening to your answers right now and it's just so pleasant for me. Ayn Rand, Atlas Shrugged. From other interviews I see that you, like, well, you read Hobbes too, and, well, that is, uh, what books do you read in general? How much do you read a day or a month? How many books? What do you focus on more? — Well, right now, of course, while I'm busy with Higgsfield, the company, I read practically nothing at all.

我觉得会是这样。哲学家我们有，政治学者也有，对，这些都需要讲出来，需要去谈。我现在听您的回答，真的觉得特别舒服。安·兰德，《阿特拉斯耸耸肩》。我看别的采访，您还读霍布斯，那您平时都读什么书？一天或者一个月读多少？多少本？更偏重哪方面？——现在当然，忙 Higgsfield 这家公司，我基本什么都读不了。

Uh-huh. But on vacation I read Bernanke, his book, where he wrote about the history of the Fed in great detail, from the earliest views. But globally I got lucky, probably, in that when I was working specifically on artificial intelligence,

嗯。休假的时候读了伯南克的书，他把美联储的历史写得特别细，从最早那些看法写起。但从大局看，我大概挺幸运的，就是当年专门做人工智能的时候，

Uh-huh.

嗯。

I had the time to go broad. That is, that classical, that's, probably, the English kind of education, where you need to study everything in breadth. And I was probably just lucky that I had the time for it. And yes, I really was interested in understanding, like, where, well, like, the social institutions we live by now came from, how they appeared. And, of course, well, we were strongly influenced by, like, Western countries, so Western, like, philosophy, the history of Western philosophy, I studied it. And yes, that's starting, obviously, like, Aristotle, Plato, like, Greek thought

我有时间横向铺开地学。就是那种经典的、大概算英式教育吧，什么都得横着学一遍。我大概只是运气好，有时间干这个。是的，我确实很想搞明白，我们现在赖以生活的这些社会制度是从哪儿来的，怎么出现的。当然，西方国家对我们影响很大，所以西方哲学、西方哲学史我都学了。是的，那就从亚里士多德、柏拉图，从希腊思想开始，

all the way from the classics,

从古典一路开始，

yes, and onward

对，再往后

that's a whole university curriculum then, if you got all the way to to the modern era, to the English philosophers. I mean it's amazing, that's literally a university education,

那这就是一整套大学课程了，如果你一路读到近代，读到英国哲学家。真让人吃惊，这简直就是大学教育，

yes. Well, John Locke, probably, and David Hume — those are the two philosophers who, probably, most strongly, like, shaped my views.

对。约翰·洛克，还有大卫·休谟，大概是塑造我观念最深的两位哲学家。

John Locke — that's political science. That's practically, that's the structure of society, right. That's

约翰·洛克那是政治学。讲的基本就是社会怎么构造，对吧。这

a really serious thing.

是很严肃的东西。

What, what drew you in — the social, the social contract, an agreement between people? What, what drew you in? Dam—,

是什么吸引了你——社会契约，人和人之间的约定？到底是什么吸引你？哎呀——

well, John Locke essentially created liberalism. I mean, he is essentially the father of liberalism. Liberalism is a system that in the end, in the end, still won out, made, you know, America the richest.

约翰·洛克实质上创造了自由主义。他基本上就是自由主义之父。自由主义这套系统到最后还是赢了，把美国变成了最富的。

Made our civilization, you could say, as of toda—

可以说，塑造了我们今天的文——

Yes, still, and in the world too, yes, made our civilization in the sense that, you know,

对，在世界范围也一样，塑造了我们的文明，就体现在

capital economies, markets — in the end they, yes, turned out to be efficient, civilization got built around that. John Locke's ideas seemed very logical to me, given that, uh, well, precisely, I'm from a family of doctors, that kind of Soviet intelligentsia. And of course, in childhood the Soviet Union wasn't something negative for me.

资本经济、市场，最后确实被证明是高效的，文明就是围绕这个建起来的。约翰·洛克的思想在我看来非常有逻辑，也正因为我出身医生家庭，属于那种苏联知识分子。当然，小时候苏联对我来说并不是什么负面的东西。

Mm-hm.

嗯。

And then, when I studied history, uh, and understood how it all actually was, and when I read John Locke — you're kind of reading a man who was writing these things in the 15th century.

后来学历史，明白这一切实际是怎么回事，再读约翰·洛克——等于是在读一个在十五世纪就写下这些东西的人。

Mm-hm.

嗯。

And he writes things that were happening in the 20th century in Kazakhstan. He kind of spelled it out in plain terms that, for example, you can't do it this way, you have to do it that way. And, well, and you also read, obviously, his opponents, his adversaries. And at that time all of it was for them so intellectual, so philosophical. And then you see that in the 19th–20th century it was all implemented, and there's already data, so it's roughly clear who was right and who wasn't. And David Hume is, probably, the kind of philosopher who strongly influenced me when it comes to studying artificial intelligence. Well, psychology,

而他写下的东西，正是二十世纪在哈萨克斯坦发生的事。他等于是掰开揉碎地讲清楚了：比如这样做不行，得那样做。当然你也会读他的对手、他的反对者。在当时，这一切对他们来说都还停留在智识和哲学层面。后来你看到，十九、二十世纪这些全都被实践了，已经有数据了，大致能看清谁对谁错。至于大卫·休谟，大概是从研究人工智能这个角度对我影响很大的哲学家。心理学，

yes, gnoseology, epistemology. And he laid the foundations of all of this in that it's actually very hard to talk about some kind of objective truth, for instance, that there are different views, in that you can't always argue things in natural language. And at Higgsfield too this also helps. We make decisions very fast, and we don't like arguments inside the company. Mm-hm. Because, because, like, David Hume proved a very long time ago that basically for any thought, uh, in natural language, uh, you can produce arguments in natural language, because natural language isn't mathematically precise and there are always a lot of holes in it. And for any point of view, if you have a big enough space for argumentation, then in natural language you can argue anything one way or the other, or the other.

对，认识论、知识论。他给这一切打下了基础：其实很难谈什么客观真理，比如说存在不同的观点，用自然语言并不总能把道理论证清楚。这一点在 Higgsfield 也很有用。我们决策非常快，公司内部不喜欢争论。嗯。因为休谟很早就证明了，任何一个想法，用自然语言都能找出论据来支持，因为自然语言不是数学上精确的，里面永远有一堆漏洞。只要论证空间足够大，任何一种观点，用自然语言都能往这边或那边论出来。

Mm-hm.

嗯。

And that's also our policy, that we always move data driv—, data-driven. We always want to get data from the market, rather than sit around, rather than reflect, rather than plan, but move fast. And it's one thing when we just want to do it that way, and another thing when you have the knowledge that this has actually been studied for several centuries already. Uh, and this has, like, such a very — such views have such a strong foundation.

这也是我们的一条方针：永远按 data-driven 走。我们总想从市场拿到数据，而不是坐着反思、做计划，而是快速行动。我们只是想这么干，和你知道这件事其实已经被研究了好几个世纪，这是两回事。这种看法背后是有很扎实的根基的。

Right after the conference that Oskar Blisbegov put on, I liked your talk. I also gathered from that talk that you're interested in the stock market. Where would you invest? Which companies, in your opinion, are going to grow?

就在 Oskar Blisbegov 办的那场会议之后，我很喜欢你那场演讲。我从演讲里还看出来，你对股市有兴趣。你会投到哪里去？你觉得哪些公司会涨？

Yes, globally, of course, there's a view that the market is very overheated, because, well, the valuations — I mean the largest companies, Nvidia, Google is already worth 4 trillion, Meta is worth trillions, well, some valuations that used to seem completely un—

对，从全球看，当然有一种观点认为市场过热了，因为，估值——最大的那几家公司，Nvidia、Google 已经值 4 万亿了，Meta 也是万亿量级，有些估值放在以前完全觉得不

unreachable,

可能达到，

yes. And, uh, it all depends, of course, on the horizon. I really, well, believe in artificial intelligence, and I understand that this is fairly, well, this really is a redivision of the world economy. That's why the valuations we're seeing are fair, because we operate in this market. Mm-hm.

对。当然，这都取决于时间跨度。我很相信人工智能，我也明白这确实是全球经济的一次重新划分。所以我们看到的这些估值是合理的，因为我们本身就在这个市场里做事。嗯。

And that's why we're seeing such gigantic cash flows, even, you know, with those same numbers, yeah, yeah, we, in less than a year, we have several hundred million. And the run rate revenue

所以我们才会看到这么巨大的现金流，就拿同样这些数字来说，对，对，我们不到一年，就有几亿。而 run rate revenue

is 300 million now, right, or higher.

现在是 3 亿，对吧，甚至更高。

Well, already more,

嗯，已经更多了，

more than 300 million. I don't know, do you grow every single week or what.

超过 3 亿。我都不知道，你们是每周都在涨吗？

Well, we grow 30% a month.

我们每个月涨 30%。

That's insane — in a week, right? Two weeks ago we were talking,

太夸张了，一周就这样？两周前咱们聊的时候，

yeah. And so it turns out — where do such, well, gigantic flows come from? Because GPUs are very expensive. That is, as I said earlier, one country controls the entire market, and there's a big shortage of GPUs on the market, there aren't enough GPUs. That is, the demand that people have right now for their own productivity — everybody wants to, like, answer an email not by writing it himself anymore, but with the help of, like, ChatGPT or Claude. Everybody wants to, like, not write the code, but to ask the neural net. Everybody wants to, like, not to — not to design a little brochure for social media, but wants to generate it. And it's like this all over the world. And to make presentations automatically, and to plan, to do research. That is, essentially the entire white-collar economy, and it has gigantic demand for neural networks. And what does that mean? It's gigantic demand, ultimately, for these graphics cards. And in the end they're produced by one company in the world, on a small island.

对。那这么巨大的资金流是从哪儿来的？因为 GPU 非常贵。就像我前面说的，整个市场被一个国家控制，市场上 GPU 严重短缺，根本不够用。现在人们对自己生产力的那种需求——每个人都想回邮件不用自己写，用 ChatGPT 或者 Claude。每个人都不想自己写代码，想直接问神经网络。每个人都不想自己给社交媒体设计一张小图，而是想直接生成出来。全世界都这样。还有自动做 PPT，做调研规划。本质上整个白领经济，对神经网络的需求是巨大的。这意味着什么？意味着最终对这些显卡的需求是巨大的。而生产它们的，全世界归根到底只有一家公司，在一个小岛上。

Yes, Nvidia. And inside Nvidia — it's a client of TSMC, which directly, uh, manufactures all of this. That's on the small island of Taiwan,

对，Nvidia。而 Nvidia 自己是 TSMC 的客户，真正把这些东西造出来的是 TSMC。就在台湾那个小岛上，

yeah,

对，

the shortage is gigantic, and that's exactly why prices are so high. If the market looked different, if TSMC had a lot of competitors, a whole lot more GPUs would be produced, then prices would be lower, valuations would be lower. But since the shortage is gigantic,

短缺极其严重，所以价格才这么高。如果市场是另一个样子，如果 TSMC 有很多竞争对手，GPU 的产量会多得多，价格就会更低，估值也会更低。但正因为短缺极其严重，

a subscription to ChatGPT would cost, like, not 20 dollars but 2 dollars, and, accordingly, valuations would be, like, 10 times smaller. But because of this very shortage, on the one hand valuations are gigantic, on the other hand there are, yeah, these geopolitical risks, in that if production stops, then, well, it'll be very hard, world

ChatGPT 的订阅就不是 20 美元，而是 2 美元，估值也会小 10 倍。但正因为这种短缺，一方面估值巨大，另一方面也有地缘政治风险——如果生产停摆，那会很麻烦，全世界的

development will slow down. That'll hit harder than the Strait of Hormuz.

发展会放缓。这比霍尔木兹海峡那一下还狠。

Well, it'll be hard, because prices will rise sharply. It'll cause demand destruction, in that there's this risky situation where everything is controlled by one country. That's,

会很难，因为价格会暴涨。会造成 demand destruction，因为现在这个局面本身就有风险：所有东西都被一个国家控制。这，

well, it seems to me they'll relocate production sooner, sooner or later, probably they'll protect themselves through diversification. The planet has to, humanity has to protect itself somehow, right. Well, and then the logic of the market is capitalism too. Everyone will have to strive for efficiency, which means they'll drive costs down.

我觉得早晚会把生产迁走，大概会靠分散化来自保。地球总得、人类总得想办法保护自己吧。再说市场逻辑也是资本主义，所有人都得追求效率，那就会把成本压下来。

Yes. It is happening — Donald Trump keeps gathering the CEOs of all the tech companies, saying that everything is going to be relocated. But there are several fundamental things there, in that this intellectual property, how to produce all of this, it's also with TSMC. And other countries have, uh, a time lag before they can reproduce it. And TSMC itself wants to move, uh, to America, but what also blocks it there is that TSMC is Taiwan. If they move everything to America — they build all the fabs along the coast of Taiwan. If China starts to invade, it'll have to destroy all the fabs. And they're also strategically built around the island. So that's a factor too. And there's a fundamental scientific and technical factor.

对。这件事正在发生：Donald Trump 一直把所有科技公司的 CEO 召集起来，说要把生产都迁走。但这里有几个根本性的东西：这套知识产权、这些东西具体怎么造，也都在 TSMC 手上。别的国家还有一个时间差，得等他们能复制出来。TSMC 自己也想搬到美国，但还有一点卡着：TSMC 就是台湾。如果他们全搬去美国——工厂现在全建在台湾沿岸。中国真要打，就得把这些厂全毁掉。而且这些厂在岛周围是按战略布局的。这也是一个因素。另外还有一个根本性的科技因素。

The fabs are there to defend Taiwan, right? Because, because China also needs Nvidia's product,

工厂本身就是在保护台湾，对吧？因为中国也需要 Nvidia 的产品，

right? Yes. And

对吧？对。而且

interesting, so for now your recommendation is not to sell Nvidia? Yes. And globally there are also these risks, but long-term, long-term, obviously, the market will sort it out, yeah, the classic phrase,

有意思，所以目前你的建议是 Nvidia 别卖？对。全球层面确实有这些风险，但长期看、长期看，市场自然会解决，对，就那句老话，

obviously, production will be relocated, obviously, data centers will be built, a new, like, supply chain will be created, and long-term AI, it really does boost productivity. Mm-hm.

生产肯定会迁走，数据中心肯定会建起来，新的供应链会搭出来，长期看 AI 确实在拉高生产力。嗯。

And we're already seeing this now, the fact that

而且我们现在已经看到了，

the productivity, the economy, of the American economy has grown a lot. And that's also a factor in why it's growing so fast. And there are the companies, there's Anthropic, OpenAI, right now they're strongly — there's also Elon Musk's. Right now these three companies, they're fighting for liquidity, and all three plan to go for an IPO this year. And xAI is already going out first, in June. And then Anthropic, OpenAI. Uh, they

美国经济的生产力大幅提高了。这也是它涨得这么快的一个因素。有这么几家公司：Anthropic、OpenAI，现在还有 Elon Musk 那家。这三家公司现在在抢流动性，三家今年都计划 IPO。xAI 第一个上，六月。接着是 Anthropic、OpenAI。他们

everyone's waiting for Anthropic. For some reason everyone's waiting for Anthropic,

大家都在等 Anthropic。不知道为什么，都在等 Anthropic，

yeah, everyone's waiting for Anthropic, because they've grown up to a 40 billion run rate in 4 months. That is, in all of history there's never been

对，都在等 Anthropic，因为他们 4 个月就涨到了 400 亿 run rate。历史上从来没有过

such growth. And you have to understand that in the end one of them will win — Anthropic, or OpenAI, or xAI.

这种增速。而且要明白，最后赢的一定是它们中的某一个——Anthropic、OpenAI 或者 xAI。

But it will most likely be the first company that will be worth 10 trillion dollars.

而这大概率会是第一家值 10 万亿美元的公司。

So you believe in that more? Yes. Not even Nvidia, not Apple, not Meta, but specifically a company out of artificial intelligence,

所以你更信那边？对。连 Nvidia 都不是，不是 Apple，不是 Meta，就是做人工智能的公司，

yeah, because they have the strongest talent. They have this velocity, they grow very fast. If they, if nothing breaks inside them, then the candidates for the ones who'll reach 10 trillion — that's Google.

对，因为他们的人才最强。他们有那种 velocity，涨得非常快。只要内部不出问题，能走到 10 万亿的候选里，就是 Google。

Mm-hm.

嗯。

That's Anthropic or OpenAI. Well, there's a strong battle going on between them. And it's possible Meta joins the race, while Mark Zuckerberg, he's on vacation, everyone's joking there, he's performing at the UFC, well, he's not, like, running a serious company, but when he does get involved, Meta — one of the contenders is Amazon, also giant resources. Micr—, yeah, Bus could still show something,

就是 Anthropic 和 OpenAI。他们俩打得很凶。Meta 也可能加入这场竞赛——现在 Mark Zuckerberg 在休假，大家都在开玩笑说他跑去打 UFC，好像没在正经管公司；但等他真下场，还有 Meta。另一个有力选手是 Amazon，资源同样巨大。Micr——对，Bus 也可能还会出手，

yeah, and these companies will get to 10 trillion. And you have to understand where those 10 trillion will come from. There is a large number of other technology companies, those classic SaaS companies, B2B, the SaaS market, which — the whole, the entire American enterprise is built on it. There is, for example, a company called PagerDuty. It's this classic infrastructure company. It has around a billion in recurring revenue and a valuation on the exchange of around, like, 2 billion. That is, a multiple of just two. Nobody believes in this company, because everyone reckons that if Anthropic is going to be worth 10 trillion dollars, then all of these layers, they'll burn up.

对，这些公司会做到 10 万亿。得想清楚这 10 万亿从哪来。还有一大批别的科技公司，那种经典的 SaaS 公司、B2B，整个 SaaS 市场——美国的 enterprise 全建在这上面。比如有家公司叫 PagerDuty，典型的基础设施公司，大概 10 亿的 recurring revenue，股市估值也就 20 亿左右，倍数只有 2。没人看好这家公司，因为大家都在算：如果 Anthropic 值 10 万亿美元，那这些中间层就全烧没了。

Mm-hmm.

嗯。

All of these B2B SaaSes. And that's a point of view, many people in the Valley share it, and there is such a probability. So globally, if you look long-term, then of course you're bullish on these AI companies.

就是所有这些 B2B SaaS。这是一种看法，硅谷很多人都认同，也确实有这种可能。所以从全局看、看长期，当然是看多这些 AI 公司。

And, well, it seems to me the optimal portfolio is, well, "optimal" — this is naturally not investment advice. The optimal, sort of, risky portfolio is an AI portfolio, AI companies and gold, because gold, yeah,

我觉得最优的组合是——说"最优"，这当然不算投资建议——最优的风险组合就是 AI 组合，AI 公司加黄金，因为黄金，对，

everyone, yeah, well,

大家都，对，

everyone sees how overheated all of this is and, well, and gold has very, like, quickly, well, been growing,

大家都看得到这些东西被炒得多热，而且黄金涨得也很快，

yeah, and it keeps growing, and at the same time, well, usually, that is, fear is growing in the market too, and, well, bullish. Kazakhstan is lucky again. And we have a lot of gold, in the gold price, which is growing, and a lot of good people. And I hope that Kazakhstan's economy in the near future will only grow, only up and up. And in conclusion, one more question. No, of course, there are many questions, but we'll probably discuss them later. But this question, which I ask practically everyone who works with artificial intelligence: how dangerous is artificial intelligence? How true is it that artificial intelligence could develop a will of its own? And could it develop a life of its own, its own, its own motivation? How do you look at this?

对，一直在涨。而且通常，就是说，市场上的恐慌也在涨，同时又看多。哈萨克斯坦又走运了：我们黄金多，金价还在涨，好人也多。希望哈萨克斯坦经济接下来只会往上走，一路向上。最后再问一个问题——当然问题还有很多，那些以后再聊。这个问题我几乎会问每一个做人工智能的人：人工智能到底有多危险？说人工智能可能产生自己的意志，这有几分真？它会不会产生自己的生命、自己的动机？您怎么看？

Indeed, yes. Uh, such a risk does exist. And the notion of will, or there's also this notion of consciousness — it still, well, this notion has no scientific term for it. There's a giant number of philosophical theories about what consciousness is. And there are certain schools of thought that already consider that modern LLMs have already attained consciousness. For example, recently Richard Dawkins, he there, uh, he had a collaboration with Anthropic, and he there came to the opinion that in Claude, in this Anthropic model, consciousness has already appeared. There, Richard Dawkins, the famous scientist, stated this. And then the question is how big the risks of this are for humanity. And there really are these local risks that artificial intelligence is already being used for military purposes. We know there's a technology company like Palantir, which is already using artificial intelligence for military purposes. It's a company that's very opinionated overall about how the world system of the whole world should look. It's a very opinionated company and it has this very powerful resource in artificial intelligence. So there really are, uh, various local risks already. There is this long-term risk with super artificial intelligence, which will greatly surpass, like, us, all of humanity. And here the question is, uh, what, how are we going to coexist with it?

确实，是有这种风险。意志这个概念，还有意识这个概念，到今天都还没有一个科学的说法。关于什么是意识，哲学理论多得吓人。也有一些思想流派认为，现在的 LLM 已经具备意识了。比如最近 Richard Dawkins，他跟 Anthropic 有过一次合作，结论是 Anthropic 的 Claude 这个模型已经出现意识了。Richard Dawkins，著名科学家，是他说的。接下来的问题是，这对人类的风险有多大。眼下确实有一些局部风险——人工智能已经被用于军事。我们知道有 Palantir 这样一家科技公司，已经把人工智能用在军事上。这家公司对整个世界秩序该长什么样非常有主见（opinionated），手里又握着人工智能这么强的资源。所以各种局部风险确实已经存在。还有一个长期风险，是超级人工智能，它会远远超过我们、超过整个人类。问题就是，我们要怎么跟它共存？

Coexist, yes. Yes, there's probably an upside in that artificial intelligence, it doesn't go down the classic evolutionary path that, like, other species, apart from humans, well, all the species on earth, they evolved along. And this food chain got built up, where one species dominated another species, one species there destroyed, another species literally there fed on it, yeah. Yeah. Well, we're biological, yeah,

共存，对。可能有一个好处：人工智能不走那条经典的进化路径——地球上除人以外的其他物种，都是沿着那条路进化过来的。食物链就是这么搭起来的：一个物种压过另一个物种，一个物种把另一个消灭掉，另一个真的就被吃掉，对。对。我们是生物体，对，

we're a different organism, yeah, in general, and there too entirely.

我们是另一种有机体，对，完全不一样。

So it, kind of, doesn't enter this race of classic evolution. And possibly these classic risks, uh, they're absent, because after all it's a silicon, uh, form of life, not, like, not carbon-based, like, like we are. And possibly we won't have a conflict over resources.

所以它不参与经典进化这场竞赛。那些经典风险可能就不存在了，毕竟它是硅基生命，不像我们是碳基。我们之间可能不会为资源起冲突。

It only needs electricity.

只需要电。

Well, all of that, in fact, is very solvable when there's a super-strong artificial intelligence. That's not a very hard problem, uh, to build a Dyson sphere around the sun and get, like, an infinite amount. Fine, yes, an infinite amount of energy, yeah,

其实等有了超强的人工智能，这些都很好解决。绕着太阳造一个戴森球、拿到无限的量，这不算多难的任务。好，对，无限的能量，对，

of energy, yeah. Globally it's interesting how this is going to happen, but it seems like

能量，对。从大处看，这事会怎么发生很有意思，但感觉

there's no such superconflict. Whereas if some kind of, well, like, the way at one point Homo sapiens destroyed the Neanderthals, because a struggle of species was taking place. And that, of course, is what we're hoping for. But for me, probably the highest risks are that artificial intelligence, it's now actively used in biology, in medicine, and there are already very strong open-source models that, basically, like, any student can download off the internet and synthesize some new viruses, create something in this field. Those really are high risks.

不会有那种超级冲突。而不是像当年 Homo sapiens 灭掉尼安德特人那样，那是物种之间的斗争。我们当然希望是前一种。但对我来说，最高的风险可能是：人工智能现在被大量用在生物学、医学上，已经有很强的开源模型，基本上任何一个学生都能从网上下载下来，去合成某种新病毒、在这个领域做点东西。这才是真正的高风险。

That's true, that's dangerous. And that, in my view,

这倒是，这很危险。而且在我看来，

isn't regulated as strongly as it ought to be regulated. Mm-hmm.

监管力度远远不到它该有的程度。嗯。

Everyone there discusses some risks of artificial intelligence, but I would put this as the, sort of, I'd put it as the highest risk, because yes, this really does need to be regulated, this needs to

大家都在讨论人工智能的各种风险，但我会把这一条排在最高的风险上，因为这确实需要监管，需要

be watched, because it's very, well,

盯着它，因为它太，嗯，

it has become accessible, because very strong neural networks, they're now available in open access, the ones that, like, do protein folding. There's also, uh, a great Kazakh startup, Orlan Biotech, and the founder of this startup. So they work on neural networks for biology, like, uh, for protein folding, for things like that. And he was the one who told me about all of this. He himself is, like, a doctor of science in this field. And really, yes, things like that are frightening.

太容易拿到了，因为很强的神经网络现在都公开可得，就是做 protein folding 的那些。还有一家很棒的哈萨克斯坦创业公司，Orlan Biotech，还有这家公司的创始人。他们做的就是给生物学用的神经网络，做 protein folding 这类东西。这些都是他跟我讲的，他自己就是这个领域的博士。确实，这种事挺吓人的。

Yerzat, thank you very much for the interesting answers. I wanted to invite you onto the program once more, to invite you for a recording once more after some time and talk in more detail about Higgsfield, about what you're planning to do, about the films, about the series that you're planning to shoot or are already shooting, and about feature films, about how you interact with representatives of the creative industries of Kazakhstan or of the world. So I think it would be very interesting for our audience to learn your personal opinion about this. Yes,

Yerzat，非常感谢您这些有意思的回答。我想再请您上一次节目，过段时间再录一期，更细地聊聊 Higgsfield，聊聊您们打算做什么，聊聊电影、您们打算拍或已经在拍的剧集，还有长片，以及您们怎么跟哈萨克斯坦和全世界创意产业的人打交道。我觉得我们的观众会很想听听您个人对这些的看法。对，

thank you very much for the invitation. I'll be very glad to take part once more.

非常感谢邀请。很乐意再来一次。

Yes. Well, and in conclusion I, here, gave a book to Yerzat. "And Life Taught Me." I wrote a book about Nurlan Smagulov. I hope to write a book about you. What life will teach you. So I recommend it for reading.

对。最后，我刚送了 Yerzat 一本书，《生活教会了我》。我写的是 Nurlan Smagulov。希望以后也能写一本关于您的——写生活会教给您什么。所以推荐大家读一读。

Yes. Thank you. It's a great honor to receive a book from you. Thank you. Well, it only remains for me to say goodbye to our viewers. Until we meet again. All the best.

对。谢谢。能收到您的书是莫大的荣幸。谢谢。那我只剩下跟观众们道别了。下次再见，一切都好。