Chinese startup DeepSeek announced last Monday, January 27, that it would temporarily limit registrations due to a cyberattack after the brand’s artificial intelligence assistant suddenly became very popular.
It is worth noting that on Monday, the mentioned AI developer encountered disruptions in the operation of its website. The artificial intelligence assistant, designed by DeepSeek, has become a top-rated free app available on Apple’s App Store in the United States.
The startup resolved issues related to the mentioned app’s programming interface and the inability of users to log in to the website. The outages were the longest for DeepSeek in about 90 days and coincided with a surge in popularity.
Last week, the startup launched a free virtual assistant. The brand noted that this digital product differs from similar functional systems of other players in the artificial intelligence industry in that it uses less data at a lower cost. It is possible that this peculiarity of the virtual assistant developed by DeepSeek will become a kind of turning point in the approach to investing in the artificial intelligence industry on a global level. Obviously, for investors, a strategy involving smaller amounts of financial injections is preferable. If this approach is eventually scaled up, players in the artificial intelligence industry will face significant changes in the economic aspect of their activities. Huge investments have already become something like a dogma, which the virtual product designed by DeepSeek challenges with its example.
The Chinese developer’s popular app is powered by the DeepSeek V3 model, which, according to the startup, tops the list of leaders among open-source models and competes with the most advanced closed-source models in the world.
According to media reports, the virtual product designed by DeepSeek was perceived by Silicon Valley as an event. In this case, the dominance of the United States in the global artificial intelligence industry was actually challenged. It is also important that the success of DeepSeek has led to all sorts of doubts about the effectiveness of Washington’s export control regime regarding Beijing. The United States has limited shipments of advanced chips and equipment for manufacturing microcircuits of the appropriate category to China. Despite this, DeepSeek has managed to develop more than just an incredibly competitive but also the most popular technology product in its segment. It is worth noting that Washington’s export restrictions nevertheless have an impact on Beijing’s capabilities in the technology sector, but these measures are probably not critical in terms of China’s fundamental potential to move forward, including in the area of artificial intelligence.
On Monday, technology stocks turned out on a downward trajectory. The relevant tendency was because DeepSeek provoked questions about the sustainability potential of the current dominance of the largest players in the global artificial intelligence industry and the need for large-scale financial injections into AI development. On Monday, the sharp drop was demonstrated by the value of stocks, for example, companies such as Nvidia and Oracle.
Returning to the topic of Washington’s export restrictions against Beijing, it is worth noting that chips are necessary for training and ensuring the subsequent operation of artificial intelligence systems. China is still heavily dependent on external shipments of microcircuits. In this context, it is worth noting that DeepSeek uses Nvidia’s chips. Without microcircuits from the United States, the success of the mentioned startup would not have been possible. It is worth noting that chips are a kind of basic material element that provides an opportunity for the development of the artificial intelligence industry.
Currently, information is actively spreading that DeepSeek has developed an advanced product, spending less than $6 million, which is almost the minimum amount by the standards of the area of machine intelligence, and using Nvidia’s less powerful chips, rather than the latest items of this company. As noted by the media, not everyone believes in the mentioned information. For example, some technology executives suggest that there may have been an export control violation in this case. It is implied that DeepSeek could have used Nvidia’s chips, the power of which exceeds the figures provided for by Washington’s restrictive measures against Beijing. At the same time, this is just an assumption, not a proven fact.
The media draws attention to the lack of information about the company behind DeepSeek. This startup from Hangzhou was founded in 2023. It is worth noting that in the mentioned year, the search engine giant Baidu released the first Chinese large language model. Since then, dozens of big and small technology companies from the Asian country have developed their own large language models. At the same time, the virtual product designed by DeepSeek is the first to be recognized by the United States artificial intelligence industry as matching or even surpassing the performance of cutting-edge US models.
