DeepSeek, a Chinese artificial intelligence start-up, has demonstrated its latest AI assistant. The assistant achieves performance comparable to or even surpassing some of the world’s best chatbots. This was accomplished using significantly less computational power and fewer advanced AI chips, particularly cutting-edge GPUs.
Expert @_KarenHao on what Deepseek may mean for the future of journalism and Gen AI.
"DeepSeek will be very inspirational for AI researchers who now realise they can develop approaches that require significantly less money."
✍️ A piece by @Marina__Adamihttps://t.co/yoUP9oZiP7
— Reuters Institute (@risj_oxford) February 1, 2025
This development suggests a potentially disruptive method for developing large language models (LLMs). It offers a more efficient alternative to current strategies. It raises questions about future infrastructure investments and the demand for AI chips.
“While AI technology has provided hugely important tools… the business model depends on hype.”
Excellent piece on the significance of DeepSeek by @kenanmalik https://t.co/2YFyWRBr8E
— John Tasioulas (@JTasioulas) February 2, 2025
As of this writing, Nvidia’s stock has plunged more than 22% from its recent peak. Similar declines have been seen in other leading AI tech stocks.
“While privacy fears are justified, the main beef Silicon Valley has is that China’s chatbot is democratising the technology“
[DeepSeek has ripped away AI’s veil of mystique. That’s the real reason the tech bros fear it | Kenan Malik | The Guardian] https://t.co/gxZyqf80yC
— Norio Nakatsuji (@norionakatsuji) February 2, 2025
Despite constraints, DeepSeek managed to develop AI models like DeepSeek-V3 and DeepSeek-R1 with cutting-edge capabilities on a reported training budget of around $6 million.
“The AI startup has upended the industry by developing a model that costs much less to produce – and is available free to a universe of tinkerers“
[DeepSeek: cheap, powerful Chinese AI for all. What could possibly go wrong? | John Naughton | The Guardian] https://t.co/HcRyHfyVyb
— Norio Nakatsuji (@norionakatsuji) February 1, 2025
Its white paper outlines an innovative approach that combines software ingenuity with new training techniques to maximize the potential of older Nvidia GPUs. DeepSeek’s models have shown impressive results next to market-leading alternatives from Meta Platforms in benchmarks for problem-solving, mathematical reasoning, coding, and general knowledge.
DeepSeek’s economic AI breakthrough
The DeepSeek AI Assistant rapidly ascended to become the most-downloaded free application on the App Store in the United States. DeepSeek’s open-source platform makes the code publicly accessible, allowing businesses and developers to customize AI models without incurring the high costs of proprietary systems. The company reportedly offers commercial pricing that is significantly lower than that of its competitors for those seeking to integrate DeepSeek’s most advanced models into existing tech infrastructure via an application programming interface (API).
Several companies have switched to using DeepSeek in a bid to save costs, marking a significant shift in the competitive landscape of artificial intelligence technology in Europe. Startups have been quick to adopt DeepSeek, while larger, established firms are taking a more cautious approach, opting to watch and wait before making a transition. On 20 January, the Hangzhou-based company released DeepSeek-R1, a partly open-source ‘reasoning’ model that can solve some scientific problems at a similar standard to models developed by OpenAI.
Earlier this week, DeepSeek launched another model, called Janus-Pro-7B, which can generate images from text prompts much like OpenAI’s DALL-E 3 and Stable Diffusion, created by Stability AI in London. Researchers inside China say the success is expected and fits with the government’s ambition to be a global leader in artificial intelligence. In 2017, the Chinese government announced its intention for the country to become a global leader in AI technologies and applications by 2025.
By 2022, the Chinese Ministry of Education had approved 440 universities to offer undergraduate degrees specializing in AI. That year, China supplied almost half of the world’s leading AI researchers, while the United States accounted for just 18%. DeepSeek’s rise reflects China’s broader objectives and substantial investments in the AI sector, which continue to drive innovation and competitiveness on the global stage.
