Microsoft offers a DIPSic artificial intelligence model on its cloud platform

Microsoft has recently announced that the agency’s cloud subscribers can access the Dipsic R1 from AI Foundry, which is said to be able to analyze and deduce.

Does Microsoft come to peace with Dipsic?

A few days after the significant success of Deepsic artificial intelligence in AppStore and the unprecedented welcome of users, Microsoft has announced in its blog post that the R1 model will be available on the agency platform and on the AI ​​Foundry platform. He got. The model is said to have the ability to analyze and infer and show high capabilities in understanding and conclusions.

It should be noted, however, that Openai, a very close partner of Microsoft, had recently announced that Dip-Sick’s makers have violated the terms and conditions of the company’s services, including ChatGPT, and have used unauthorized data to advance their goals in Deepseek-V3 model development. . However, it seems that Microsoft does not want to be lagging behind under any circumstances, and even Openai’s displeasure with recent Deepsic actions does not make them a popular and seemingly powerful model on its cloud platform.

The Microsoft Cloud Cloud Development Team has announced that the Dipsic R1 has been undergoing security reviews, and the automatic evaluation of model behaviors alongside security reports shows that there is no threat to the company’s customers. Microsoft has also pointed out that holders of Copyotel Plus devices will be able to run a separate version of the model on their computers in the near future.

Microsoft security experts had recently announced that Deepsic developers had stepped down a lot of data through the API used in OpenAI products in the summer of 2008. It is more interesting to know that Microsoft is one of the biggest shareholders of Openai, and even the company’s time executives were aware of it; But all of the said things did not make Microsoft refuse to provide an interesting and popular model that has been welcomed by users these days.

It is still unclear whether Microsoft has made changes to the Dipsic R1 to reduce censorship, or we are still faced with the same limited and loyal version of the Chinese ruling party. According to studies by the Newsguard Institute of Credit, the R1 model in 5 % of cases provides non -dedicated answers or avoids answering questions related to the day’s news. This is true for 5 % of the questions about China and its government.