DeepSeek’s founder has at least two reasons to celebrate this week. Alongside the Chinese Lunar Year, the company had the App Store’s most downloaded app with its AI agent, while the US market melted down due to the new model being so accurate and cheap to train compared to comparable models from the likes of OpenAI.
Now, Business Insider reports that DeepSeek’s founder, Liang Wenfeng, went viral on Chinese social media, mostly due to the plummeting of the US stock market. On Weibo, viral threads call Wenfeng a “Guangdong AI hero.” One of the posts, with over 18 million views, said, “DeepSeek is a megahit: internet users buzz about three AI heroes of Guangdong.”
Alongside DeepSeek’s CEO, Moonshot AI’s founder, Yang Zhilin, and the AI scientist He Kaiming, who is an author of one of the most-cited papers on machine learning, were praised over the Chinese social media.
Here’s why DeepSeek hype is real
Almost out of nowhere, DeepSeek went from another GPT company to the hottest startup around. The main reason was that the Chinese company open-sourced its models while revealing in its R1 research paper that the highly sophisticated AI was trained at a fraction of the cost of OpenAI’s o1, which caused the stocks of several big tech companies to plummet.
After that, even OpenAI’s CEO Sam Altman commented on the DeepSeek hype, saying he’s excited to “have a new competitor.” Altman promises he has even more to announce to counter DeepSeek. Still, they have a major difference. The Chinese startup only uses 3% to 5% of the resources OpenAI needs for similar progress with ChatGPT.
One of the problems with the current AI software concerns the cost of developing and using the product. Advanced models like o1 can cost tens of millions to develop. The process requires high-end graphics cards (GPU) that provide the necessary computing power and energy expenditures.
That’s why finished products like ChatGPT o1 can’t be available for free without limitations. On the other hand, DeekpSeek researchers took another approach for R1, finding ways to train an advanced reasoning model without access to the same hardware.
It’s not just that, but DeepSeek made access to R1 much cheaper than OpenAI’s ChatGPT, which is a significant development. Add in the open-source nature of DeepSeek models, and you can see why developers would flock to test the Chinese firm’s AI and why DeepSeek would surge in the App Store.
