Why does DeepSeek make Western AI practitioners feel insecure?

By Carl Franzen

Compiled by: Xiaobai Navigation Coderworld

Until a few days ago, only the most hardcore geeks (I say this as one) had heard of DeepSeek, a ChineseAIThe company, a subsidiary of the equally uniquely named High-Flyer Capital Management, a quantitative analytics firm founded in 2015, has become perhaps the most closely watched company in Silicon Valley over the past few days.

这主要归功于DeepSeek-R1的发布，这是一个新的大语言模型（LLM），它能够进行类似于OpenAI当前最佳模型o1的”推理”——在回答困难问题和解决复杂问题时需要花费数秒或数分钟，通过步骤式或”思维链”的方式对自身的分析进行反思。

Not only that, DeepSeek-R1 scores as well as or better than OpenAI’s O1 on various third-party benchmarks (tests that measure how well an AI can answer questions on a variety of topics), but it also reportedly cost only about $5 million to train and used far fewer graphics processing units (GPUs) than are strictly prohibited in the United States (OpenAI’s home base).

But unlike o1, which was only available to paid ChatGPT Plus tier subscribers ($20 per month) and higher tiers like the $200 per month Pro tier, DeepSeek-R1 was released as a fully open source model, which explains why it quickly climbed the AI code-sharing ladder.CommunityHugging Face tops the list of most popular and active models.

And, since it is completely open source, people have fine-tuned and trained the model in many ways to adapt it to different specific tasks, such as making it small enough to run on mobile devices, or combining it with other open source models. Even if you want to use it for development purposes, the API cost of DeepSeek is more than 90% lower than the equivalent o1 model from OpenAI.

Most impressively, you don’t even need to be a software engineer to use it: DeepSeek offers a free website and mobile app for US users, and its R1-powered chatbot interface is very similar to OpenAI’s ChatGPT. However, DeepSeek once again surpasses OpenAI by connecting this powerful reasoning model with web search — something OpenAI has not yet achieved (web search is currently only available on the less powerful GPT series of models).

An obvious irony

There’s a rather interesting, if disturbing, irony here, given OpenAI’s original goal of democratizing AI for the masses. As Nvidia senior research manager Jim Fan said at X: “We’re living in a timeline where a non-US company is continuing OpenAI’s original mission of truly open, cutting-edge research that empowers everyone. It doesn’t make sense. But the most interesting outcomes are often the most likely to happen.”

As user X @SuspendedRobot put it (citing reports that DeepSeek appears to have been trained on ChatGPT-generated question-answer output and other data): "OpenAI has learned from the entire internetsteal"The fact that DeepSeek steals data from them and gives it back to the public for free reminds me of an English folk tale."

Meta in crisis, due to open source Llama lagging behind?

But it wasn’t just fans who noticed DeepSeek’s success. Based on my conversations and readings with various engineers, thinkers, and leaders, the open source availability of DeepSeek-R1, its high performance, and the fact that it seemed to “come out of nowhere” to challenge the previous leaders in generative AI, sent shockwaves throughout Silicon Valley and beyond. If “everyone” is going crazy about it, as my exaggerated title suggests, it is at least a hot topic in tech and business circles.

A message posted on Blind, an anonymous Silicon Valley gossip-sharing app, is going viral suggesting that Meta is in crisis over the success of DeepSeek, because it has so quickly surpassed Meta’s own efforts to become the king of open source AI via the Llama model.

“This is a game changer.”

User X @tphuang made a convincing point: "DeepSeek has commoditized AI beyond the top tier. The first picture made it clear to me. R1 is much cheaper than US labor costs, which means many jobs will be automated in the next 5 years." He later also pointed out why DeepSeek's R1 is more attractive to users than OpenAI's o1:

“O1 has 3 huge problems:

1) Too slow

2) Too expensive

3) Lack of control/over-dependence on OpenAI by end users.

The R1 solves all of these problems. Companies can buy their own Nvidia GPUs to run these models without worrying about extra costs or slow/unresponsive OpenAI servers.”

@tphaung also raised a thought-provoking analogy question: "Will DeepSeek become the Android of LLM?"

Web entrepreneur Arnaud Bertrand was blunt at X about the staggering implications of DeepSeek’s success: “It’s hard to overstate how game-changing this is. This is not just about AI, it’s also a huge blow to the United States’ misguided attempts to block China’s technological development.”Xiaobai NavigationThe great irony is that without this limitation, DeepSeek might not have existed (as the old saying goes, necessity is the mother of invention).”

Review issues

However, some have voiced caution about DeepSeek’s rapid rise, arguing that as a startup operating in China, it must necessarily comply with the country’s laws and content censorship requirements. In fact, when I used the iOS version of DeepSeek in the United States, I found that it would not answer certain questions.

As a member of the news media, I of course attach great importance to freedom of speech and expression, which is one of the most fundamental concepts I firmly support.

However, I also feel compelled to point out that OpenAI’s models and products (including ChatGPT) also refuse to answer a range of questions — particularly those involving human sexuality and adult/NSFW content, even if these questions are mundane.

Of course, this isn’t an apples-to-apples comparison. For some, resistance to relying on foreign technology will make them skeptical of DeepSeek’s ultimate value and usefulness. But its performance and low cost are undeniable.

In an era when 16.5% of goods in the United States are imported from China, it is difficult for me to justify the censorship concerns orSafetyThe risks are to warn against using DeepSeek-R1—especially when the model code can be downloaded for free, used offline,SafetyThe environment runs on the device and can be fine-tuned at will.

I do detect some existential crisis thinking about the “decline of the West” and the “rise of China” in the heated discussion around DeepSeek. Some have linked this to the fact that American users joined the Xiaohongshu app when TikTok was briefly banned, and were amazed by the quality of life in China shown in the videos shared there. The emergence of DeepSeek-R1 takes place in the context of this narrative - one in which China looks (and by many indicators is) rising, while the United States looks (and by many indicators is) declining.

The first, but definitely not the last, Chinese AI model to rock the world

Nor will this be the last Chinese AI model to threaten the dominance of Silicon Valley giants—even if those giants, like OpenAI, are raising more money than ever before for their efforts to develop artificial general intelligence (AGI, or programs that outperform humans in most economically valuable tasks).

Just yesterday, another Chinese model from TikTok’s parent company ByteDance, Doubao-1.5-pro, was released, which performed comparable to OpenAI’s non-inference GPT-4o model in third-party benchmarks, but at only 1/50 the cost.

Chinese models are getting so fast and so good that even people outside the tech industry are taking notice: The Economist just published an article about DeepSeek’s success and other Chinese AI efforts, and political commentator Matt Bruenig posted on X: “I’ve been using Gemini, ChatGPT, and Claude for NLRB document summarization for almost a year now. Deepseek is better than all of them at this. Its chatbot version is free. The price of using its API is 99.5% less than OpenAI’s API. [shrug emoji]”

How does OpenAI respond?

No wonder OpenAI co-founder and CEO Sam Altman said today that the company will bring its yet-to-be-released second-generation inference model series o3 to ChatGPT, which will be available even to free users. OpenAI seems to still be carving its own path with more proprietary and advanced models - setting industry standards.

But the question is: With DeepSeek, ByteDance, and other Chinese AI companies nipping at its heels, how long can OpenAI stay ahead in producing and publishing new cutting-edge AI models? And if it does fall behind, how fast and how severe will its decline be?

OpenAI does have another historical precedent to draw on, though. If DeepSeek and Chinese AI models do what Google’s open-source Android did to mobile — taking over a large portion of the market for a while — you only have to look at how the Apple iPhone took over the high end of the market with its closed, proprietary, all-in-house approach and steadily expanded downward from there, particularly in the U.S., to the point where it now owns nearly 60% of the domestic smartphone market.

Still, for all those who are shelling out big bucks to use AI models from leading labs, DeepSeek shows that the same functionality might be available at a lower price, with greater control. In an enterprise setting, that might be enough to win.

The article comes from the Internet:Why does DeepSeek make Western AI practitioners feel insecure?

Related recommendations: Dialogue with Dragonfly Managing Partner: How to not rely on luckcryptocurrencySuccess in the field?

True success comes from technical understanding, continuous learning, and value creation, not just the pursuit of short-term gains. Original: When Shift Happens Compiled by: Yuliya, PANews In the field of cryptocurrency investment, finding the next 100-fold return opportunity is the dream of every investor. As the world's top cryptocurrency investment fund, Dragonfly…