David Sacks, the Crypto and Artificial Intelligence Advisor of the White House, argued that Deepseek, a Chinese -based artificial intelligence initiative, has benefited from the outputs of OpenAI’s models. Sacks told Fox News, “There is important evidence that Deepseek has done here is to distill information from the OpenAI models, and I do not think OpenAI is very pleased with it, Kul he said. However, he did not give information about the details of these evidence.
This technique, known as distillation, allows the model of an artificial intelligence to learn from the outputs of another model to develop similar abilities. This method is a common practice in the field of artificial intelligence and is used to make models more efficient and effective. Deepseek’s ability to create a more efficient artificial intelligence model without being largely dependent on advanced GPUs using this technique draws attention. This may lead American artificial intelligence companies to take precautions against distillation technique to prevent the models they call “imitator ..
Deepseek’s artificial intelligence model R1 performs comparable with OpenAI’s O1 in mathematical and logical reasoning. The company claims that it has achieved this success by using lower cost and less calculation power. This has brought up the issues of productivity and cost efficiency in artificial intelligence research.
On the other hand, Microsoft and OpenAI explores whether Deepseek’s performance is caused by distillation of its own models or an independent research breakthrough. According to sources close to the issue, Microsoft’s security researchers observed that people whom they thought may be linked to Deepseek leak a large amount of data leaking using the OpenAI application programming interface (API). OpenAI’s technology partner and largest investor, Microsoft, said OpenAI informed about this activity.
Investigation from Italy to Deepseek
Italy’s data protection authority Garante demanded information about how Deepseek uses personal data. Garante asked for detailed information on issues such as types, resources, purposes, legal basis and whether the data were stored in China. Deepseek and its affiliated companies have 20 days to respond to these demands. This attracts attention as one of the first regulatory moves against Deepseek.
Deepseek’s rapid rise led to an increase in competition in the field of artificial intelligence and intensifying ethical debates. How the company’s methods will be evaluated by other players in the sector and how this will affect future artificial intelligence research is a matter of curiosity. In addition, these developments once again demonstrate the importance of ethical and legal standards in the development of artificial intelligence models.