Refresh
Deepseek does not pass the navy
The American navy has issued a new warning to sailors, warning against Deepseek IA due to “security and ethical concerns”, ” According to CNBC. It seems that the alert was issued by the US Navy on Friday January 24, 2025, which is a few days before the application stops new registrations and experienced a breakdown on Monday January 27, 2025.
As indicated by CNBC, the warning of the American navy can be read, “We would like to bring to your attention a critical update concerning a new model of AI called Deepseek,” said the email. The memo said it was “imperative” that team members do not use Deepseek AI “for any task related to work or personal use.”
The American navy confirmed the authenticity of the opinion and referred to its generative AI policy. The warning is essentially equivalent to a ban against Deepseek IA and its different models, asking the recipients to “refrain from downloading, installing or using the Deepseek model for any title”.
Apparently, the American navy had to have its reasoning beyond the breakdown and reported malicious attacks that hit Deepseek AI three days later.
President Trump says that Deepseek IA should be “alarm clock” for American AI companies
President Trump, only two weeks after his second term, commented Deepseek, saying: “The liberation of Deepseek IA of a Chinese company should be alarm to our industries that we must be focused on the laser to compete to win because We have the greatest scientists in the world, ” According to the Washington Post.
He described the launch of Deepseek AI as a “alarm clock”, adding that competitors in the United States – potentially Openai, Nvidia and Google – must be “focused on the victory on victory”. Trump’s comments were also likely to reflect the impact of Deepseek News on the US stock market. Most of the technological actions have slipped, but the leader of the GPU of AI, Nvidia, had his worst day recorded.
It is a comment without surprise, but the follow -up declaration was a little more confusing because President Trump would have said that the breakthrough of Deepseek in the more efficient AI “could be positive because technology is now also available for companies American ” – This is not exactly the case, however, because the newcomer to AI does not yet share these details and is a Chinese owned company.
Nvidia calls Deepseek an “excellent AI progression”
Beyond Sam Altman of Openai sharing his reflections on Deepseek IA and promising many more chatgpt, Nvidia also commented publicly, calling Deepseek an “excellent progression of AI”.
The response made after the decrease in record stock prices of 600 billion dollars yesterday, the most important decline that shares have ever seen and largely the result of the performance of Deepseek and the cost of the model of Ia. Beyond being impressed by R1, it is clear that Nvidia wants to remain a key element of the story.
The complete written declaration indicates, “Deepseek is an excellent AI progression and a perfect example of time testing. Deepseek’s work illustrates how new models can be created using this technique, taking advantage of widely available models and calculations which are fully in accordance with export control. Inference requires a large number of NVIDIA GPU and high performance networking. We now have three scaling laws: pre-training and post-training, which continue, and the new test time scale. »»
Who really has a Deepseek AI?
Deepseek was founded in mid-2023 by the Chinese Director of Heart Funds Liang Wenfeng, who is the CEO of the company. Liang founded High -Flyer, a hedge fund that uses AI to create trading strategies in 2015 – then according to a Washington Post profileused this experience to develop major language models with its new Deepseek company.
How close Deepseek’s ties close to the Chinese government? Inevitably, the new success of the AI application attracted a lot of attention, but it has apparently not always considered an AI star in China.
According to Matt Sheehan (an expert in the Chinese AI industry cited in the profile of the Washington Post), Deepseek was “not the” chosen “of the start-ups of the Chinese AI” and that he ” Taken the world by surprise, and I think of a great extent, they took the Chinese government by surprise “.
But Deepseek is now far from being an unknown – and it will be interesting to see if or how it is from the Chinese government in order to appease these growing fears of privacy.
Is Deepseek sure to use?
We have made our own in -depth comparison of the compression of Deepseek in Chatgpt, but since then, some confidentiality alarm ringtones have been retained on the application.
As indicated by the BbcAustralian Minister of Sciences, Ed Husic, told ABC News earlier in the day that there were a lot of unanswered questions around “data management and privacy” with Deepseek. “I would be very careful, this type of problem should be carefully weighed,” he warned.
Deepseek Privacy Policy It is quite open that “we store the information we collect in secure servers located in the People’s Republic of China”. This information includes your email address, telephone number, date of birth and cat history.
None of this is very different from the confidentiality policies of Chatgpt or Gemini, but the harvest of this information in China – and the fact that it is combined with “the actions you have taken outside the service” of advertisers ” – is forced to keep these alarm rings of the bells sound stronger in the coming days.
Will Deepseek media threw last?
The editor of Techradar, launches Ulanoff, wrote a beautiful dismantling of the Deepseek media threw – wondering if the chatbot, which is not yet multimodal, is worthy of the column thumbs that it obtains and (quite reasonably) suggesting That it is unlikely that last in the United States, given the recent misfortunes of Tiktok.
Commenting on the decreases in the course of NVIDIA action and others, he notes “Without any information or real proof that Deepseek and his investors are transparent and truthful, investors began to draw their ia dollar from the stock market American”.
Even if we accept that Deepseek is a breakthrough, there are understandable ban on its longevity in the United States. As Ulanoff says, “it doesn’t matter how good this application will not survive the current American climate”.
Sam Altman of Openai answers
Sam Altman of Openai has now commented publicly Deepseek for the first time, declaring on X (formerly Twitter) that the AI model is “impressive” – and I can’t help but hear this in the voice of Patrick Bateman in American psycho business card scene.
But he was also generally optimistic about Openai’s response, declaring that “we will obviously provide much better models” and that it is “legitimate invigorating to have a new competitor”. Altman also does not think that the news changes the image in terms of fleas, declaring that “more calculation is more important than ever to succeed in our mission”.
The markets do not seem to agree, the Nvidia flea giant suffering from the largest market value diving of a day in the history of the United States yesterday.
A deep fast refreshment
A little confused about Deepseek? Here is a quick primer. The free AI chatbot was published on January 20, but exploded in popularity in recent days when fans of technology have achieved its meaning. As Marc Andreessen’s venture capital noted on X (formerly Twitter), “Deepseek R1 is the spoutnik moment of Ai”.
The application is currently at the top of the free graphics on the App Store and the Apple Play Store in the United States and many other countries, although it is made in China – which has been the subject of A commercial ban on the advanced chips of Nvidia.
Ironically, it is these commercial restrictions that seem to have triggered ingenuity behind Deepseek, which was created using a small quantity of the enormous computing power which is behind the main models of today’s AI.
Reference tests show that he can do tasks such as answering questions and generate code as well as current AI models. However, you may find it difficult to create a Deepseek account – he was forced to take a break following a major cyber attack.
