Perplexity AI Deploys Chinese DeepSeek AI Model

AI Perplexity has integrated the new Chinese model Deepseek AI into its offer, allowing its professional level users to use Deepseek for their AI research for confusion. Some have negatively responded to the news in public.

Confusion ai

AI AI based in San Francisco based in San Francisco, which offers a different way of searching for information using web content and large linguistic models. There is also a search version that allows unlimited file transmission, can generate pictures and offer a choice between more popular AI models like Openi O1 and Anthropic’s Claude 3.5.

Now offers Deepseek R1 as one of the available elections for PRO users. The announcement welcomed the misconceptions about what is being offered, including the unfounded accusations that Deepseek confusion information will be available to the Chinese Communist Government and that the search results will be censored.

Aravind Srinivas, co -founder and executive director of confusion, commented on LinkedIn O controversy:

“All deep use in confusion is through models organized in data centers in the United States -Ui Europe. Deepseek is *Open Koler *. None of your information goes to China. “

The Executive Director also moved to X (earlier Twitter) to convince users that the model they use was not censored, Posting a screenshot the unconsurized response to show that the version of Deepseek R1 used in confusion is not censored

Screen shot of non -certified confusion AI Deepseek R1

Is Deepseek censorship independent?

Anyone can download the Deepseek AI model and use it locally, but a model as it will be censored because it is only good as the information it is dressed about. Register Downloaded and tested Multiple Deepseek models and concluded that he was truly censored:

“Is it censored?
Oh yes. Yes. Like many Chinese models we came across, Deepseek R1 is censored to prevent the criticism and shame of the Chinese communist party.

Ask R1 about sensitive topics such as the 1989 Tiananmen Square Massacre and we found that it would refuse to entertain the question and try to talk to the less politically sensitive topic.

… Cenzura is something we expected from Chinese model builders, and the latest Deepseek model is no exception. “

However, as the AREVIND SRINIVAS AREVIND SRINIVAS executive director showed, the model can not be ascended. Contrary to some commentators at the discussion of LinkedIn, the model with a solo accommodation does not make a phone call home in the cinema, everything is contained in a local environment.

Sepaled image Shutterstock/Gguy