China’s DeepSeek - A Balanced Overview

It was one of the worst if the Chinese company had billion dollar in this, but it says it has been in and of course the country's largest country. Hi, welcome to another episode of Cold Fusion. Look at these stock market charts from the 28th of January 2025. What you're looking at is a bloodbath, a bloodbath in the US stock market of over $1 trillion. And the course? The release of the DeepSeek R1 AI model from China. The Chinese model is as capable as the best US models, but it's free to use, open source, more efficient and most shocking of all, it reportedly costs less than 3% of chat GPT-01 to develop. Just two years ago on this channel, we were talking about an AI arms race between companies. Today that's evolved into an AI race between countries.

这是中国公司最糟糕的情况之一，如果他们在其中投入了数十亿美元，但事实是它一直存在，当然也是该国最大的公司。你好，欢迎收看冷聚变的另一集。看看2025年1月28日的这些股市图表。你看到的是美国股市的惨状，损失超过1万亿美元。原因呢？是因为中国发布的DeepSeek R1 AI模型。这个中国的模型与美国最好的模型一样强大，但它是免费使用的、开源的、更高效，最令人震惊的是，据说其开发成本不到Chat GPT-01的3%。就在两年前，我们在这个频道上讨论的是公司之间的AI军备竞赛。如今，这已经演变成国家之间的AI竞速。

In the one corner, we have the United States. They have a long history of technological dominance. But then on the other side, we have China, a country with a very different ideology and motives. In this race to dominance, it's not about weapons, but it's about developing systems that are designed to think artificial intelligence. This race is reminiscent of the Cold War. Some have even dubbed these events as quote, the Sputnik moment of AI. The White House says that they're looking into quote, national security implications of China's DeepSeek AI platform.

在一个角落，我们有美国。他们在技术领域一直占据主导地位拥有悠久的历史。但在另一边，我们有中国，一个拥有非常不同意识形态和动机的国家。在这场争夺主导地位的竞赛中，关键不在于武器，而在于开发旨在思考的人工智能系统。这场竞赛让人联想到冷战时期。一些人甚至把这些事件称为“人工智能的斯普特尼克时刻”。白宫表示，他们正在研究中国的DeepSeek AI平台对国家安全的影响。

And to top it all off, open AI has accused DeepSeek of stealing its IP to train their own model. It's all heating up. With the United States pouring in half a trillion dollars into the Stargate AI project, the global race is on. And this ongoing battle could be one of the biggest stories in tech this year. As artificial intelligence becomes a matter of national security, the technology would be forced to move faster than it is today. What a crazy time to be alive. But before we get ahead of ourselves, what is really going on here? How did a company from nowhere do all of this? Is this all just part of the AI hype cycle? Or is this the real deal? It seems like the whole world is playing catch-up since the release, so let's try and make sense of it all.

并且更令人震惊的是，OpenAI指控DeepSeek窃取其知识产权来训练自己的模型。事情越发紧张。随着美国向“星门AI项目”投入五千亿美元，全球竞争已经展开。这场持续的对抗可能成为今年科技界最重大的事件之一。随着人工智能成为国家安全的议题，技术的发展将比现在更加迅速。这真是个疯狂的时代。然而，在我们深入探讨之前，这背后到底发生了什么？一家无名公司是如何做到这一切的？这只是AI热潮的一部分，还是货真价实的突破？看起来，自从发布以来，全世界都在努力赶上这一趋势，所以让我们试着理清这一切。

Historically, when technology meets national security threat from an ideological opponent, we get inventions like the computer and jet aircraft from the competition of World War II, but this time around, the United States was completely unchallenged in the field of AI for the most part. But that all changed on January 20, 2025, with the release of R1. DeepSeek R1, which is free, has performance reportedly on par with OpenAI's $200 a month model. And this is performance in the context of tasks such as language reasoning, mathematics, and coding. The free model also beats out anthropics Claude Sonnet and Google's Gemini. But what many people may not know is that DeepSeek does things a little bit differently to the current state of the art models. It's in part why it's so efficient, but we'll cover these details later in the episode.

历史上，当技术遇到来自意识形态对手的国家安全威胁时，例如在二战中竞争所催生的计算机和喷气式飞机的发明。而这次，美国在人工智能领域基本上没有遇到什么挑战。然而，这一切在2025年1月20日发生了改变，那天发布了R1。DeepSeek R1是免费的，其性能据说可以与OpenAI每月收费200美元的模型媲美。在语言推理、数学和编程等任务中，免费模型甚至超越了Anthropic的Claude Sonnet和谷歌的Gemini。然而，许多人可能不知道的是，DeepSeek与当前最先进的模型有一些不同之处。这也是它如此高效的部分原因，但我们将在本集稍后详细介绍这些细节。

Because there's no competition for that level of AI performance for free, users have been flocking to it, with DeepSeek becoming number one in Apple's App Store. But here are the stats of why people's jaws are dropping. The AI was built in two months, and reportedly cost less than 5.6 million to build. The AI company Anthropic says that 100 million to 1 billion is the general amount needed to develop an AI system from scratch. And to that end, meta plans to spend 65 billion on AI. So creating something that performs this well with just 5.6 million dollars is groundbreaking. But all may not be as it seems. More on that later. Much more.

由于没有其他免费的 AI 性能提供竞争，用户们纷纷涌向这款应用，使得 DeepSeek 成为苹果应用商店的第一名。以下是让人瞠目结舌的数据。这款 AI 在两个月内完成建造，成本 reportedly 低于 560 万美元。而据 AI 公司 Anthropic 所说，从零开始开发一个 AI 系统通常需要投入 1 亿到 10 亿美元。为了达到这一目标，Meta 计划在 AI 上耗费 650 亿美元。因此，仅用 560 万美元就创造出如此优秀的产品，实属突破。但事情可能并不像看起来那么简单。后续还有更多内容要揭晓。

I think there are two very important things that people need to know about what's happening with DeepSeek AI and the way it's being interpreted on Wall Street. The first is, it doesn't matter if it's a Chinese government scyop or not. The technological innovation of having an LLM train itself through reinforcement learning is impressive. The cost efficiency of doing inference with only 7 billion parameters rather than 700 billion parameters is impressive. The possibility of being able to do more model training and inferencing with less usage of power and less chips is impressive. It doesn't mean though that chip demand is at risk. What I think it means is you're more likely to see an acceleration of AI everywhere all over the economy. DeepSeek R1 being open source means that its code is freely available for whoever wants to use it and for whatever they want to use it for. Users can modify it as they please all for free. This is totally the opposite approach of open AI which is pretty ironic.

我认为有两件非常重要的事情，人们需要了解关于DeepSeek AI的动态及其在华尔街的解读。第一，不管这是否是中国政府的策略，其技术创新令人印象深刻，即通过强化学习让大型语言模型自我训练。只用70亿个参数进行推理，而不是7000亿，成本效益显著。能够用更少的电力和芯片进行更多的模型训练和推理，这种可能性同样令人惊叹。但这并不意味着芯片需求会受到威胁。我认为这意味着我们更有可能看到人工智能在整个经济领域的加速应用。DeepSeek R1作为开源项目，意味着其代码对任何想使用它的人都是免费的，并且可以随意修改。这与OpenAI的做法完全相反，这真的很讽刺。

This is all horrific news for US AI companies because it means that suddenly their costs are all out of balance. DeepSeek with its 671 billion parameters can run locally on a stack of M4Mac pros. In contrast, investors and companies have poured billions of dollars into American AI servers. After the shock of this release, now it looks like US companies have been spending too much money using too much energy and charging too much for the services that they've been providing. Maybe in the future it's not going to be so much the models that would make the most money but the applications that run on top of them. Has this all been a massive mistake from US investors? No one knows for sure and that's why the markets are selling off.

这对美国的人工智能公司来说是个可怕的消息，因为这意味着他们的成本突然失去了平衡。名为DeepSeek的模型拥有6710亿个参数，可以在多台M4 Mac Pro上本地运行。相比之下，投资者和公司已经在美国的人工智能服务器上投入了数十亿美元。在这次发布的震惊之后，现在看来美国公司在使用过多能源和提供服务收费过高方面花费了太多钱。也许将来最赚钱的未必是模型本身，而是运行在它们之上的应用程序。这一切难道是美国投资者犯下的巨大错误吗？没有人能确定，这也是为什么市场正在抛售的原因。

One bright spot for US companies though is that users of AI systems may not feel comfortable in giving their data directly to China especially in corporate settings. In order to compete, Sam Altman, CEO of the chatGPT maker OpenAI, has announced that their GPT-30 mini model will now be given away for free. As for Mark Zuckerberg and Meta, they're internally panicking but it's not just the Americans. Over in China, the effect is the same. Other Chinese tech giants such as Bitedance, the maker of TikTok, Alibaba and Tencent have freaked out and had to cut the prices of their AI models to compete and despite the low price charged by DeepSeek, it remains profitable while its rivals lose money.

对美国公司来说，有一个积极的因素是，用户可能不愿意在企业环境中直接将数据提供给中国。为了竞争，ChatGPT的制造商OpenAI的首席执行官 Sam Altman 宣布，他们的GPT-30迷你模型将免费提供。而对于马克·扎克伯格和Meta，他们在内部感到恐慌，但这种情况不仅限于美国。在中国，情况也是一样的。其他中国科技巨头，如抖音的制造商字节跳动、阿里巴巴和腾讯，也感到压力，不得不降低其AI模型的价格以保持竞争力。尽管DeepSeek的收费较低，但它依然盈利，而它的竞争对手则在亏损。

Interestingly, OpenAI told the financial times that they have evidence that DeepSeek was using the output from chatGPT to train its own model. In fact, last year, they blocked OpenAI API accounts that they believe belonged to DeepSeek, suspecting theft. The US government's official stance is that it is possible that IP theft has occurred. It should also be noted that it seems like Chinese AI developers are still managing to get their hands on top of the line in videographics cards despite US sanctions. But that begs the question, who are DeepSeek and how did DeepSeek seemingly overnight build this thing?

有趣的是，OpenAI 告诉《金融时报》，他们有证据表明 DeepSeek 正在利用 ChatGPT 的输出内容来训练自己的模型。实际上，去年 OpenAI 封锁了一些他们认为属于 DeepSeek 的 API 账户，怀疑这些账户涉及盗窃行为。美国政府的官方立场是，知识产权盗窃有可能发生。还要注意的是，尽管有美国的制裁措施，中国的 AI 开发者似乎仍能获得顶级的视频显卡。但这就引发了一个问题，DeepSeek 究竟是谁，他们是如何在短时间内建造出这个东西的呢？

For a company responsible for one of the biggest red days in the US stock market, not a lot is known about the founder and the team behind DeepSeek, but the story is interesting so far. DeepSeek founder Liang Wenfang isn't from the typical tech world. He actually has a background in finance and co-founded a hedge fund called High Flyer. His company used AI to predict market trends and help make investment decisions. And he was very successful at that and his fund now manages 8 billion. But after his initial success, he wanted more. His next goal was to build, quote, human level AI. In 2021, he started buying thousands of NVIDIA GPUs as part of his quote, AI side project. This was right before the Biden administration began limiting US exports of AI hardware to China. Liang eventually spun off his AI side project into another company and that company was DeepSeek and the R1 is their latest model.

对于一家在美国股市创造了最大跌幅之一的公司，名为DeepSeek，其创始人及团队的信息不多，但其故事非常引人入胜。DeepSeek的创始人梁文方并非来自典型的科技行业。他其实有金融背景，并共同创立了一家名为High Flyer的对冲基金。他的公司利用人工智能预测市场趋势并帮助做出投资决策。这方面他很成功，目前他的基金管理着80亿的资产。但是，在取得初步成功后，他希望更进一步。他的下一个目标是建立所谓的“人类水平的人工智能”。2021年，他开始购买数千块NVIDIA的GPU，作为他的“AI副项目”的一部分。这发生在拜登政府开始限制向中国出口AI硬件之前。梁文方最终将他的AI副项目分拆成另一家公司，而那家公司就是DeepSeek，他们最新的产品是R1模型。

But honestly, the more I've been reading up on the Liang story, the more interesting it gets. So let me know in the comments section if you want to see a dedicated episode on the DeepSeek founder. So DeepSeek R1 was trained with reinforcement learning. That means there weren't any humans who helped it learn. And the method that DeepSeek uses for their model architecture is different to most of the other players. It's a technique called mixture of experts. Sky News explains it well. Quote, where open AI is latest model, GPT-40, attempts to be Einstein, Shakespeare and Picasso rolled into one. DeepSeek's is more like a university broken up into expert departments. This allows the AI to decide what kind of query it's being asked and then send it to a particular part of the digital brain to be dealt with. This lets the other parts to remain switched off, saving time, energy and most importantly, the need for computing power. The YouTube channel Computer File explains further.

老实说，我越深入研究梁的故事，就觉得越有趣。所以，如果你们想看到一集专门关于DeepSeek创始人的节目，请在评论区告诉我。DeepSeek R1是通过强化学习训练的，这意味着它在学习过程中没有人类帮助。而且，DeepSeek在模型架构上采用的方法与大多数其他公司不同，这是一种被称为“专家混合”的技术。Sky News对此解释得很好：“Open AI的最新模型GPT-40试图集爱因斯坦、莎士比亚和毕加索于一身，而DeepSeek的更像是一所拥有多个专家部门的大学。”这使得AI可以判断它被问到了什么类型的问题，然后发送到数字大脑的特定部分来处理。这样，其余部分可以保持关闭状态，从而节省时间、能源，最重要的是减少计算能力的需求。YouTube频道Computer File对此有进一步的解释。

So maybe you ask every specific maths question. What mixture of experts will do is have trained a specific part of this network, a much smaller part to solve that problem for you. And so you basically have the early stages will root the question to different parts of the network and then only activate a small part of it. Let's say 30 billion parameters, which is a huge, huge saving. So this sort of shaded area here will activate and then that will produce your answer. You can develop systems using agents like this where you have one that's trained to do this and one that's trained to do this and you just ask the right one. Suppose I want to train a network to write my emails for me. Maybe it's very good at that. I train a different network to solve a different problem and I just ask the right one as opposed to hoping that one model can do it. So that's much more efficiently.

所以，也许你会问每一个具体的数学问题。专家混合模型的做法是训练网络中的某个特定部分，一个更小的部分来解决你的问题。基本上，网络的早期阶段会将问题分配到不同的部分，然后只激活其中的一小部分。比如说，有30亿个参数被激活，这能节省很多资源。这里有一个标记的区域会被激活以产生答案。你可以使用这样的代理系统来开发程序，一个负责这项任务，另一个负责那项任务，你只需询问正确的系统。例如，我想训练一个网络来为我写邮件，也许它非常擅长这个。我训练一个不同的网络来解决不同的问题，只需问对了系统，而不是指望一个模型能完成所有任务。这种做法效率更高。

To add to the efficiency is a process called distillation. Basically using larger models to train smaller models in targeted domains. The result is equivalent performance with significantly less computing power. And this was the big shock for AI developers and financial markets. Making chain of thought reasoning completely open and visible was an interesting choice. Open AI basically does the opposite. Daz is essentially write down a step by step process for solving the problem and slowly solve it and then write down the answer. You tend to get much better at solving problems that require multiple steps. If you want to just what is why is the sky blue it will just go to regurgitate that pretty easily from Texas learned on the internet.

为了提高效率，有一种被称为“蒸馏”的过程。简单来说，就是利用大型模型在特定领域中训练小型模型。结果是，这些小型模型可以在性能方面媲美大型模型，但所需的计算能力却大大减少。这对于AI开发者和金融市场来说是一个巨大的震惊。让思维链的推理完全开放并可见是一个有趣的选择。OpenAI基本上采取相反的做法。Daz的做法本质上是写下一个解决问题的步骤流程，逐步解决问题，然后写下答案。这样，你会在解决需要多步骤的问题上变得更擅长。如果你只是想要了解“为什么天空是蓝色的”这种问题，它可以很容易地从网络上学到的知识中重复这些信息。

But if you're asking like problem solving skills it's hard to do in one shot so you kind of take a little bit of time to just to just take you know to just work through it. Now open AI pioneered this chain of thought but they don't tell you how they do it because it's all closed. And so it's not open AI at all right in some sense. So essentially you see a kind of pracy summary version of the chain of thought but it's not their internal actual internal model which is essentially a trade secret. What R1 is doing is it's doing a chain of thought which is similar to O1 but it's fully public they've released all the models they've released all the code you can talk to it you can see the entire model log and they've also trained it with massively more limited data.

但如果你问的是解决问题的技能，这很难在短时间达到，所以需要花一些时间来逐步解决。现在，OpenAI开创了这种思维链的方法，但他们不会告诉你如何做到这一点，因为这都是机密的。从某种意义上说，这并不是真正的OpenAI。因此，你看到的只是一个思维链的简要版本，但不是他们内部的实际模型，因为这基本上是一个商业机密。而R1所做的是它执行一种与O1类似的思维链，但它完全公开发布了所有模型和代码。你可以与之交互，查看整个模型日志，并且它们使用的数据量非常有限。

As mentioned earlier things may not be as they seem that cost figure of five point six million dollars to create the model may not be complete. In fact in a paper released by Deepseek themselves they mentioned that that five point six million dollar figure includes only the official training of Deepseek v3 and does not include costs of prior research experiments on architectures algorithms or data that does put a question mark on all the headlines we've been seeing that this thing was built for under six million dollars but whatever the real figure is it's likely to be much less than what US companies have been spending. In the latest news Deepseek has also dropped an open image model and at this rate a video model will probably soon follow and it might even rival open AI Sora or Google's anticipated VO2. In terms of search interest right now Deepseek now outpaces chat GPT and it became one of the most downloaded apps on the app store and then towards the end of January things absolutely blew up and went wild.

正如前面提到的，事情可能并不像它们看起来的那样，创建这个模型的成本为五百六十万美元的数字可能并不完整。事实上，在Deepseek自己发布的一篇论文中，他们提到这个五百六十万美元的数字只包括了Deepseek v3的正式训练费用，并不包括先前在架构、算法或数据上的研究实验费用。这让我们对各种声称该项目成本低于六百万美元的报道产生了疑问。不过，无论实际数字是多少，它可能仍然远低于美国公司所花费的金额。最新的消息是，Deepseek还推出了一个开放图像模型，照这样的速度，视频模型可能很快也会出现，并且可能与OpenAI的Sora或谷歌的预期VO2媲美。在目前的搜索兴趣方面，Deepseek现在超过了Chat GPT，成为应用商店中最受欢迎的下载应用之一。到了1月末，情况进一步爆发并变得异常火热。

China during Chinese New Year went crazy. First Alibaba comes out with Quinn 2.5 max it's a very capable AI that could one-shot this code animation. Just asking a computer to code an animation and then it goes out and does it is so intuitive that I think kids of the future will believe that this is how coding always worked. Alibaba's Quinn 2.5 max outperforms Deepseek and even GPT40 in some tasks and then there's Kimi K1.5 released around the same day. It's also a great performer, is multi-modal and can browse the web in real time. Okay before you all rush out to sign up to Deepseek please be aware of something it collects data such as chat history any text or audio inputs uploaded files key stroke patterns basically anything you input into the model. Now open AI does similar things but the difference is that with Deepseek your data goes straight to servers in the people's Republic of China so I guess the question is do you want to be spied on by the US or do you want to be spied on by China?

在中国农历新年期间，中国的科技界非常热闹。首先，阿里巴巴推出了Quinn 2.5 max，这是一款非常强大的人工智能，可以轻松完成代码动画。只需向计算机请求动画代码，它就能自动完成，这种直观的操作方式可能会让未来的孩子们认为编程一直如此简单。阿里巴巴的Quinn 2.5 max在某些任务上的表现超过了Deepseek甚至GPT-40。与此同时，Kimi K1.5也在大约同一时间发布，表现也很出色，它可以进行多模态操作并即时上网浏览。 在大家纷纷想要注册使用Deepseek之前，请注意一点：它会收集用户的数据，比如聊天记录、上传的文字或音频输入、文件、击键模式，基本上你输入到模型里的任何东西都会被收集。OpenAI也做类似的事情，但区别在于，使用Deepseek时，你的数据会直接传输到中国的服务器。所以问题来了：你是愿意被美国监视，还是愿意被中国监视？

I can't tell you what to do but that's just a heads up but in terms of privacy there is a bright side does mean that Deepseek can run locally on a machine without an internet connection for complete privacy. Here's the YouTube channel some ordinary gamers running it locally. Code things for you so for instance I can ask Deepseek like write me code for a simple login webpage so at this moment in time it'll think it'll be like all right the user is asking for code to create a simple login page so first it's going to structure HTML then it's going to style then it's going to validate and then here it is it's actually writing me the HTML code so we're sitting in a world where like I feel so scared for like junior coders these days because god damn AI is really coming for some of the jobs that people least expected to lose first so again it writes this actual login page and of course once it's done it'll actually also provide you a preview in this chat box software so you can see it for yourself before you actually throw it into you know production or testing or whatever so right here I'm just going to hit that preview button and boom there it is it actually and as I was making this video

我不能告诉你该怎么做，不过只是提醒一下，关于隐私方面有一个好消息，那就是Deepseek可以在没有互联网连接的情况下在本地运行，从而实现完全的隐私保护。这里有一个YouTube频道，一些普通玩家正在本地运行它。它能帮你写代码，比如我可以让Deepseek为我编写一个简单的登录网页的代码。此时，它会认为用户在请求创建一个简单登录页面的代码，所以首先它会布局HTML，然后进行样式设计，再然后验证，最终它会为我写出HTML代码。我们生活在一个时代，在这个世界里，我为现在的初级程序员感到担忧，因为人工智能真的开始抢占一些人们最初未曾想到会失去的工作岗位。它写好了这个登录页面后，当然，它还会在这个聊天软件中提供一个预览，所以你可以在实际投入生产或测试之前看到成品。在这里，我只需要点击那个预览按钮，画面就出现了，正当我在制作这个视频的时候。

Deepseek at the start of the week had to quote temporarily limit user registrations due to large scale malicious attacks this was also a warning to many as it seems like the program may not be as ready as it seems so what does Sam Altman think is only directly referenced the company once saying deepseeks r1 is an impressive model particularly around what they're able to deliver for the price we will obviously deliver much better models and it's also legit invigorating to have a new competitor we will pull up some releases we'll see what's around the corner for open AI but the joke is AI took chat chp t's job but in all seriousness I don't think that this is over I believe that this is just the beginning of major competition what we're seeing here is the technological version of Thucydides trap basically it states when arising power challenges and existing power conflict arises in an interview with waves republished in the China Academy back in mid 2024 Deepseeks founder Liang made his ambitions clear he said quote for years Chinese companies have been accustomed to leveraging technological innovations developed somewhere else and monetizing them through applications but this isn't sustainable this time our goal isn't quick profits but advancing the technological frontier to drive ecosystem growth why is silicon valley so innovative because they dare to try when chat gpt debuted China lacked confidence in frontier research from investors to major tech firms many felt the gap was too wide and focused instead on applications but innovation requires confidence and young people tend to have more of it end quote with such a mindset deepseek may force AI innovation forward and China could be at the forefront of the global AI race competitors around the world will be forced to reduce their costs and rethink how they're creating AI models efficiency will be the aim of the game we don't know how it will play out but we do know that we'll be having some rapid advancements in the coming years if we do remain positive we could see breakthroughs in medical science material science mathematics and even theoretical physics in the long term we could make products for cheaper make them longer lasting and produce them more efficiently but on the flip side

在本周初，Deepseek 由于大规模的恶意攻击，不得不暂时限制用户注册。这也给很多人敲响了警钟，似乎这个项目还没有准备得那么充分。那么，Sam Altman对此有何看法？他仅在谈论公司时提到了一次，称Deepseek的R1是一个令人印象深刻的模型，尤其是在他们所能提供的性价比方面。他表示：“我们显然会推出更好的模型，同时也为有新竞争对手出现而感到振奋。我们将迎接一些发布，看看OpenAI会有什么新动向。”虽然有人开玩笑说AI抢了ChatGPT的工作，但认真来说，我认为这并未结束，这只是激烈竞争的开始。我们看到的是技术版的修昔底德陷阱，基本上说的是，当一个新兴势力挑战现有势力时，冲突就会发生。 在2024年中期，中国科学院发表的一次访谈中，Deepseek的创始人梁先生明确表达了他的雄心。他表示：“多年来，中国公司习惯于利用其他地方开发的技术创新，通过应用来实现盈利，但这种方式不可持续。我们这次的目标不是快速盈利，而是推动技术前沿，以促进生态系统的增长。硅谷为何如此创新？因为他们敢于尝试。当ChatGPT问世时，中国的投资者和大型科技公司对前沿研究缺乏信心，许多人认为差距过大，而转而关注应用。但创新需要信心，年轻人往往更有信心。” 有了这样的心态，Deepseek可能会推动AI创新，并使中国在全球AI竞争中处于前沿。全球的竞争者将被迫降低成本，并重新思考如何创造AI模型，效率将成为关键。我们虽然不知道事情会如何发展，但我们知道在未来几年会有快速的进步。如果我们保持积极乐观，可能会在医学、材料科学、数学、甚至理论物理领域取得突破。从长远来看，我们可以制造更便宜、更耐用和更高效的产品。但另一方面，也有潜在挑战。

what about nefarious uses and bad actors geopolitically also what happens to all of the humans through this transition as AI rapidly improves that's for the future to decide and I have done a video on that topic years ago before AI blew up so you can check it out after this one but as usual in all of this let's just keep a close eye and see where this goes anyway that's about it from me and that is where we are with deepseek r1 how it works so efficiently and the absolute shock that it's caused around the world although a lot of people may find consumer AI annoying these days there's no getting around it it's here to stay and improving with each week

关于那些涉及地缘政治的不法用途和不良行为者，以及在人类快速转向人工智能的过程中，人类的命运将如何，这是未来需要决定的问题。我几年前就这个话题做过一个视频，在人工智能大爆发之前，你可以在看完本视频后去查看。不过，一如既往地，我们要密切关注事态的发展。总的来说，这就是我对于DeepSeek R1的看法，它的高效运转及其在全球引起的震撼。尽管如今许多人可能觉得消费者级人工智能有些烦人，但不可否认的是，它已经成为常态，并在每周不断进步。

it's going to be an important part of everyday life soon but how does AI work anyway well now there's a fun and easy way to learn about that and many other stem subjects with today's sponsor brilliant brilliance course on artificial neural networks is perfect for that i've used it to brush up on some background context when i was making AI episodes each lesson on brilliant allows you to play with concepts a method proven to be six times more effective than watching lecture videos plus all content on brilliant is crafted by teachers researchers and professionals from mit cow tech duke microsoft google and more learn at your own pace to brush up on a project for work or just for your own self development and curiosity

它即将成为日常生活中一个重要的部分，但人工智能到底是如何运作的呢？现在有一种有趣且简单的方法可以让你学习这个以及其他许多STEM学科，那就是我们今天的赞助商Brilliant。Brilliant的人工神经网络课程非常适合这个目的。我之前在制作AI相关内容时，就用它来补充一些背景知识。Brilliant的每节课都让你可以动手操作概念，这种方式被证明比观看视频讲座有效六倍。而且，Brilliant上的所有内容都是由来自麻省理工学院、加州理工学院、杜克大学、微软、谷歌等机构的教师、研究人员和专业人士精心打造的。你可以根据自己的节奏学习，无论是为了工作项目补充知识，还是为了自我成长和个人求知欲。

to try everything brilliant has to offer for a full 30 days visit the url brilliant.org slash cold fusion to get started you'll also get 20% off an annual premium subscription thanks for watching my name is de gogo and you'll have been watching cold fusion and i'll catch you again soon for the next episode cheers guys have a good one thanks

要体验 Brilliant 30 天的全部精彩内容，请访问网址 brilliant.org/coldfusion 开始。您还可以享受 20% 的年度高级订阅折扣。感谢观看，我是 De Gogo，您刚刚收看了 Cold Fusion。我们下期再见，祝大家一切顺利，谢谢！