DeepMind CEO: The World Is Not Ready for What's Coming

章节 1：商业竞赛加速人工智能发展

📝 本节摘要：

哈萨比斯认为，若能沿着更审慎的科研路径推进，人工智能或可先用于科学与医学；但聊天机器人和基础模型的商业成功催生了激烈竞赛。竞赛带来资金、产品和公众参与，也让严谨研究更难兼顾。旁白指出，巨额资本投入正推动产品快速面世。

[原文] [Demis Hassabis]: I don't sleep very much, partly because of too much work, but also I have trouble sleeping. Shane and I and others who've been doing this for a long time, we understand it better than anybody, the enormity of what's coming. I'm actually pretty surprised there isn't more of that being discussed, given that, you know, even our timelines, which were—there were some very short timelines out there, but even ours are 5 to 10 years, which is not long for institutions or things like that to be built to handle this. So, I'm quite worried about what those sorts of systems will be able to do maybe in 2, 3 years' time. So, we're working on cyber defense in preparation for a world like that where maybe there's millions of agents, you know, roaming around on the internet. But then there'll be rogue actors, maybe rogue

[译文] [Demis Hassabis]: 我睡得不多，一部分是因为工作太多，另一部分是我本来就不太睡得着。我、谢恩（Shane）以及其他长期从事这项工作的人，比任何人都更了解即将到来的一切有多重大。说实话，我很惊讶，大家对此讨论得还不够多。毕竟，即使按我们的时间表——外面有些预测的时间非常短——我们估计也只剩五到十年。对于建立机构或其他机制来应对这件事来说，这并不算长。所以我很担心，这类系统两三年后可能会做到什么程度。我们正在准备网络防御，以应对那样的世界：也许会有数百万个智能体在互联网上四处活动。但也会有不法行为者，也许是失控的

[原文] [Demis Hassabis]: nations, maybe rogue organizations, maybe people building on top of open source. I don't know. Like, obviously it's very difficult to stop that. Then something may go wrong, and hopefully it's just sort of medium-sized, and then that will be a kind of warning shot to humanity across the bow.

[译文] [Demis Hassabis]: 国家、失控的组织，或者基于开源模型进行开发的人。我不知道。显然，这很难阻止。之后可能会出问题，希望规模只是中等程度，那就像是向全人类船头开的一记警告炮。

[原文] [Speaker A]: That is Demis Hassabis. He co-founded in 2010, shared the 2024 Nobel Prize in Chemistry for AlphaFold, and now runs all of Google's AI research. He is the most measured man at the top of this industry, and on his own company's podcast, he admits the world may not act until something goes wrong. Here is his full warning.

[译文] [Speaker A]: 这就是德米斯·哈萨比斯。他于2010年共同创立了相关机构，因 AlphaFold 共同获得2024年诺贝尔化学奖，如今负责谷歌的全部人工智能研究。他是这个行业高层中最沉稳的人；但在自家公司的播客上，他承认，世界或许要等到出事才会采取行动。下面是他完整的警告。

[原文] [Speaker A]: In terms of all of those missing pieces, I mean, I know that there's this big race at the moment to release commercial products, but I also know that Google DeepMind's roots really lie in that idea of scientific research. And I found a quote from you where you recently said, "If I'd had my way, we would have left AI in the lab for longer and done more things like AlphaFold, maybe cured cancer or something like that."

[译文] [Speaker A]: 说到这些尚未补齐的环节，我知道现在大家都在竞相推出商业产品，但我也知道，谷歌 DeepMind 的根基其实在于科学研究。我看到你最近说过这样一句话：“如果能按我的想法来，我们会让人工智能在实验室里多待一阵，多做些像 AlphaFold 这样的事，也许还能治愈癌症之类的。”

[原文] [Demis Hassabis]: Mhm.

[译文] [Demis Hassabis]: 嗯。

[原文] [Speaker A]: Do you think that we lost something by not taking that slower route?

[译文] [Speaker A]: 你觉得我们没有选择那条更慢的道路，是不是因此失去了些什么？

[原文] [Demis Hassabis]: I think we lost and gained something. So, I feel like that would have been the more pure scientific approach. At least that was my original plan, say 15, 20 years ago, when almost no one was working on AI. We were just about to start DeepMind. People thought it was a crazy thing to work on. But we believed in it, and I think the idea was that if we made progress, we would continue to incrementally build towards AGI, be very careful about what each step was, and the safety aspects of it, and so on; analyze what the system was doing, and so on. But in the meantime, you wouldn't have to wait till AGI right before it was useful. You could branch off that technology and use it in really beneficial ways to society,

[译文] [Demis Hassabis]: 我觉得我们既有所得，也有所失。我认为，那会是更纯粹的科研路径。至少十五到二十年前，我最初的计划就是如此，那时几乎没人研究人工智能。我们刚准备创办 DeepMind，大家觉得这事很疯狂，但我们相信它。我们的想法是，如果取得进展，就一步步朝通用人工智能（AGI）推进，谨慎对待每一步及其安全问题，并分析系统在做什么。不过，在此期间，我们不必等到通用人工智能出现，它才有用。我们可以从这项技术中分出一些成果，用来切实造福社会，

[原文] [Demis Hassabis]: namely advancing science and medicine. So, exactly what we did with AlphaFold, actually, which—it's not a foundation model itself, a general model, but it uses the same techniques, you know, transformers and other things, and then blends it with more specific things to that domain. So, I imagined a whole bunch of those things getting done, which would be hugely beneficial. You'd release them to the world, just like we do with AlphaFold, and indeed do things like cure cancer, and so on, while we were working on the more AGI track in the lab. Now, it's turned out that chatbots were possible at scale, and people find them useful, and then they've now morphed into these foundation models that can do more than chat and text, obviously, including Gemini.

[译文] [Demis Hassabis]: 也就是推动科学和医学进步。事实上，AlphaFold 做的正是这件事。它本身不是基础模型，也不是通用模型，但运用了相同的技术，比如 Transformer 等，再结合针对该领域的专门方法。我原本设想能做出很多这类成果，带来巨大益处。我们可以像发布 AlphaFold 那样把成果开放给全世界，甚至实现治愈癌症之类的目标，同时继续在实验室推进通用人工智能方向的研究。后来我们发现，聊天机器人可以大规模推出，而且人们觉得有用；如今它们又发展成了基础模型，显然不只会聊天和处理文本，例如 Gemini。

[原文] [Demis Hassabis]: They can do images and video, and all sorts of things. And that's also been very successful commercially and in terms of a product. And I love that, too. I've always dreamed of having the ultimate assistant that would help you in everyday life, make it more productive, maybe even protect your brain space a bit as well from an attention perspective, so that you can focus and be in flow, and so on. Because, you know, today with social media, it's just noise, noise, noise. And I think AI that works for you can help us with that.

[译文] [Demis Hassabis]: 它们能处理图像、视频和各种内容。在商业上和产品层面，这也非常成功，我同样很喜欢。我一直梦想拥有一个终极助手，帮你处理日常生活，让生活更有成效；也许还能从注意力的角度，保护一点你的脑力空间，让你能够专注、进入心流。如今社交媒体上全是噪音、噪音、噪音。我认为，真正为你服务的人工智能能在这方面帮上忙。

[原文] [Demis Hassabis]: So, I think that's good, but it has created this pretty crazy race condition where there are many commercial organizations, and even nation-states, all rushing to improve and overtake each other. And that makes it hard to do rigorous science at the same time. We try to do both, and I think we're getting that balance right. On the other hand, there are lots of pros to the way it's happened, which is, of course, that a lot more resources are coming into the area, so that's definitely accelerated progress.

[译文] [Demis Hassabis]: 所以我觉得这很好，但它也造成了一场相当疯狂的竞赛：许多商业机构，甚至国家，都在争相改进，想要超越彼此。这让人很难同时开展严谨的科学研究。我们努力两边兼顾，我觉得目前把握住了平衡。另一方面，这种发展方式也有不少好处，当然就是有更多资源投入这个领域，肯定加快了进展。

[原文] [Demis Hassabis]: And also, I think the general public are actually, interestingly, only a couple of months behind the absolute frontier in terms of what they can use. So, everyone gets the chance to feel for themselves what AI is going to be like. And I think that's a good thing, and then governments sort of understand this better.

[译文] [Demis Hassabis]: 而且有意思的是，普通公众能用到的技术，实际上只比最前沿落后几个月。这样每个人都有机会亲自感受人工智能会是什么样子。我觉得这很好，也能让各国政府更了解这项技术。

[原文] [Speaker A]: Look at what that race costs. In 2025, Alphabet spent about $91 billion on capital expenditure, most of it on AI computing infrastructure, servers, data centers, and networking. In February 2026, it told investors it would spend 175 to 185 billion this year, roughly double. By August, it had raised the range again to between 195 and 205 billion dollars. Nobody spends 200 billion dollars a year on something they would rather keep in the lab. Money on that scale has to earn a return, and it earns it through products shipped fast to billions of people. So, when the man running Google's AI says he would have gone slower, ask who got their way instead. Keep listening. Quick thing before we carry on: most of the people who watch this channel are not actually subscribed. If these videos are worth your time, take 1 second and hit subscribe. It's free. It genuinely helps.

[译文] [Speaker A]: 看看这场竞赛的代价。2025年，Alphabet 的资本支出约为910亿美元，其中大部分用于人工智能算力基础设施、服务器、数据中心和网络。2026年2月，公司告诉投资者，今年计划支出1750亿至1850亿美元，约为前一年的两倍。到了8月，它又将区间上调至1950亿至2050亿美元。没人会每年花2000亿美元在一件宁愿留在实验室里的事上。如此规模的资金必须带来回报，而回报来自迅速推出产品，送到数十亿人手中。所以，谷歌人工智能的负责人说自己本希望慢一点时，不妨想想最终是谁说了算。请继续听下去。继续之前插一句：看这个频道的大多数人其实还没有订阅。如果这些视频值得你花时间，就花一秒钟点个订阅。完全免费，确实能帮到我们。