DeepSeek has spread.
This week, DeepSeek, a Chinese artificial intelligence laboratory, broke into mainstream consciousness Its chatbot app rises to the top of the Apple App Store chart ((and Google Play,). DeepSeek’s AI model is trained using computationally effective technology Leading Wall Street analysts – and technicians – Question whether the United States can maintain its lead in the AI race and whether demand for AI chips will remain.
But where did DeepSeek come from and how did it so quickly rise to international reputation?
DeepSeek’s Trader Origins
DeepSeek is backed by Gaofei Capital Management, a quantitative hedge fund in China that uses AI to inform its trading decisions.
AI enthusiasts Liang Wenfeng Wenfeng reportedly started to get involved in trade in 2019 when a student at Homo sapiens started to get involved in trading, and he launched High Flying Capital Management in 2019 as a hedge fund with a focus on developing and deploying AI algorithms.
In 2023, High-Flyer started with DeepSeek, a lab dedicated to researching the separation of AI tools from financial businesses. The lab uses High-Flyer as one of its investors to spin into its own company, also known as DeepSeek.
From day one, DeepSeek built its own data center cluster for model training. But like other AI companies in China DeepSeek is affected by U.S. export ban. To train one of its latest models, the company was forced to use the NVIDIA H800 chip, a smaller-functioning version of the chip H100 that the U.S. company can use for U.S. companies.
It is said that DeepSeek’s technical team is biased towards Young. company It is reported that actively recruiting Ph.D. in AI researchers from top Chinese universities. DeepSeek also hires people without any background in computer science According to the New York Times, to help its technology better understand a wide range of subjects.
The powerful model of DeepSeek
DeepSeek launched its first set of models in November 2023 – DeepSeek Encoder, DeepSeek LLM and DeepSeek Chat.
DeepSeek-v2 is a universal text and image analysis system that performs well in various AI benchmarks – and runs much cheaper than comparable models at the time. It forces DeepSeek’s domestic competition, including Bytedance and Alibaba, reduces the use price of some models and gives others complete freedom.
DeepSeek-V3launched in December 2024, joins only DeepSeek’s infamous.
According to DeepSeek’s internal benchmarks, DeepSeek V3s are both superior to downloadable, publicly available models, such as those of Meta camel and “closed” models that can only be accessed through the API, such as Openai’s GPT-4O.
Also impressive is DeepSeek’s R1 “inference” model. DeepSeek claims to be released in January R1’s O1 model of Openai on key benchmark and OpenAI’s O1 model.
As an inference model, R1 effectively performs fact checks, which helps it avoid some of the traps that usually trip over the model. Compared to typical non-disputed models, the inference model takes longer (usually longer to minutes) to reach the solution. The advantage is that they tend to be more reliable in fields such as physics, science, and mathematics.
However, there is one drawback to the other models of R1, DeepSeek V3, and DeepSeek. Artificial intelligence developed in Chinese, they are Benchmarking China’s Internet regulator ensures that its response “embodies core socialist values.” For example, in DeepSeek’s chatbot app, R1 won’t answer questions about Tiananmen Square or Taiwan’s autonomy.
A destructive approach
If DeepSeek has a business model, it is not clear what the model is. The company puts its products and services at prices well below market value and offers them to others for free. It also doesn’t take investors’ moneydespite a lot of risk interest.
The way DeepSeek tells is that efficiency breakthroughs allow it to remain extremely cost-competitive. Some experts dispute However, the company provides the figures.
In any case, developers have adopted the DeepSeek model, which is not open source, as the phrase is generally understood but is available under a loose license that allows commercial use. According to Clem Delangue, CEO of Hugging Face, Hugging Face is one of the platforms that host the DeepSeek model, Developers embracing faces create 500+ R1 “derived” models There were 2.5 million downloads in total.
DeepSeek’s success for larger, more established competitors has always been Described as “Ascending AI” and “Overpromotion.” The success of the company is at least partly for NVIDIA’s share price fell 18% In January, Response to the public From Openai CEO Sam Altman.
Microsoft Announces DeepSeek availability on its Azure AI Foundry serviceMicrosoft’s platform brings AI services under a single banner. CEO Mark Zuckerberg said when asked about the impact of DeepSeek on Meta’s AI spending, CEO Mark Zuckerberg said Spending on AI infrastructure will continue to be a “strategic advantage” Used for meta. March, Openai is known as DeepSeek’s “state subsidy” and “state control”, It also suggested that the US government consider banning the DeepSeek model.
On NVIDIA’s fourth-quarter earnings call, CEO Jensen Huang highlighted DeepSeek’s “outstanding innovation”, Saying it and other “inference” models are very useful for NVIDIA because they require more computation.
at the same time, Some companies are banning DeepSeek,entire nation and government,,,,, Including South Korea. New York State Ban DeepSeek from being used in government equipment.
As for what the future of DeepSeek may have, it is not clear. The improved model is given. But the U.S. government seems to be People are cautious about foreign influences they consider harmful. In March, the Wall Street Journal reported The United States may ban DeepSeek on government equipment.
This story was originally published on January 28, 2025 and will be updated regularly.
