Technical stocks Fell. Giant companies such as Meta and Nvidia faced a barrage of questions about their future. Social media executives have taken social media to announce their concerns.
Everything was due to the start of the unknown Chinese artificial intelligence Dibsic.
Deepseek caused waves all over the world on Monday as one of its accomplishments – that it has created a very strong model of artificial intelligence with much less money than many artificial intelligence experts – raised a set of questions, including whether American companies are competitive in Amnesty International Next.
Deepseek is “The moment of Sputnik of artificial intelligence”, Mark Anderson, a technical project, a technical project, to publish On social media on Sunday.
How can the company heard a few people have such an effect?
What is a dipcique?
Deepseek is an emerging company established and owned by high Chinese stock trading company. Its goal is to build artificial intelligence techniques similar to the Chatgpt Chatbot from Openai or Gemini from Google. By 2021, Deepseek acquired thousands of computer chips from the American chip maker NVIDIA, which is an essential part of any effort to create strong AI systems
In China, the startup company is famous for seizing young and talented artificial intelligence researchers, high salaries and an opportunity to work in advanced research projects. Liang Wenfeng, a Chinese businessman.
Over the past few years, Deepseek has released many large language models, a type of technology that supports Chatbots such as ChatGPT and Gemini. On January 10, she released the first free Chatbot app, which was based on a new model called Deepseek-V3.
Why was the stock market reaction now?
When Deepsek presented Deepseek-V3 the next day for Christmas, it corresponds to the best chat capabilities from American companies such as Openai and Google. It was impressive alone.
But the team behind the new system also revealed a greater step forward. In a research paper explaining how technology built, Deepseek said it used only a small part of the computer chips on which artificial intelligence companies depend to train their systems.
The best companies in the world usually train Chatbots with giant computers that use up to 16,000 chips or more. Depsic engineers said they only need about 2000 Nafidia chips.
Why is this important?
Since late 2022, when Openai was launched in the boom of artificial intelligence, the prevailing idea was that the strongest artificial intelligence systems cannot be built without investing billions of dollars in specialized artificial intelligence chips. This means that only the largest technology companies – such as Microsoft, Google and Meta, are all based in the United States – that can build leading technologies.
(New York Times prosecution Openai and its partner, Microsoft, claiming that copyright violation of news content related to artificial intelligence systems. The two technological companies denied the claims of the case.)
But Deepseek engineers said they only need about $ 6 million of crude computing power to train their new system. It was approximately 10 times, which eliminated Meta in building the latest artificial intelligence technology.
How did Deepseek made its technique with fewer artificial intelligence chips?
The best artificial intelligence engineers in the United States says that Deepseek’s research sheet has developed smart and impressive ways to build artificial intelligence technology with fewer chips.
In short, startups showed a more efficient way to analyze data using chips. Artificial intelligence systems learn their skills by identifying patterns of huge quantities of data, including text, images and sounds. Deepseek described a way to publish this data analysis through many specialized artificial intelligence models – what researchers call a “mix of experts” – while reducing the lost time by transferring data from one place to another.
Others used similar methods before, but transferring information between the forms tends to reduce efficiency. This Deepseek did this in a way that allowed him to use the lower computing power.
Tim Deatmers, a researcher at the Allen Institute of Artificial Intelligence at Seattle and Professor of Computer Science at Carnegie Mellon University who specializes in building effective artificial intelligence systems. “Deepseek uses ways that anyone can repeat.”
Deepseek’s research paper raised questions about whether large American companies can maintain great progress in artificial intelligence. Many experts believe that artificial intelligence technology will become a commodity, as many companies sell the same product.
Is Deepseek technology as good like Openai and Google systems?
Deepseek-V3 can answer questions, solve logical problems and write their own computer programs effectively like anything in the market, according to standard standards.
Before Deepseek released its technology, Openai revealed a new system, It is called Openai O3Which looked stronger than Deepseek-V3. But Openai did not issue this system to the wider audience.
Openai O3 is designed for reason “through problems involved in mathematics, science and computer programming. Many experts have indicated that Deepseek did not build a logical model in this manner, which is seen as the future of artificial intelligence
Then on January 20, Deepseek released his thinking form called Deepsek R1, and experts impressed. In the end, investors and others sent us to panic in late last week and during the weekend, where they realized the importance of the new Deepseek technology.
The technology giants in the United States build data centers with specialized artificial intelligence chips. Is this still important, given what Deepseek did?
Yes, it is still important.
Large numbers of artificial intelligence chips can still help companies in several ways. With more chips, they can do more experiments because they explore new ways to build artificial intelligence, in other words, more chips can still give companies a technical and competitive advantage.
Experts said that more chips are also needed to operate the new strain of artificial intelligence models. This requires more computing power when people and companies use it.
Wasn’t the United States limited to the number of NVIDIA chips sold to China?
Yes. To maintain the leadership of the United States in the global artificial intelligence race, the Biden administration has established rules that limit the number of strong chips that can be sold to China and other competitors.
But the impressive performance of the Deepseek model raised questions about the unintended consequences of the restrictions of the US government. China researchers have forced creativity with a wide range of available online tools.
Some experts continue to argue with us for the restrictions of American trade, saying they were not only developed and that they will have a greater impact on China’s capabilities on creating artificial intelligence over the years.
Does Deepseek technology mean that China is now ahead of the United States in artificial intelligence?
no. The world of Openai’s O3 O3, and his performance in standard standard tests was more impressive than anything else on the market. But experts are concerned that China is jumping forward in open source artificial intelligence systems.
What is exactly an open source AI?
Like many other companiesDeepseek “opens” the latest artificial intelligence system, which means that it has shared the basic computer code with companies and other researchers. This allows others to build and distribute their own products using the same technologies.
This is part of the reason that Deepseek and others in China have managed to build an Amnesty International Competitive Systems quickly and inexpensive.
In the world of artificial intelligence, the open source first collected steam in 2023 when Meta freely shared the artificial intelligence system called llama. At that time, many assumed that the open source ecosystem will not flourish unless companies such as meta-giant companies that have huge data centers are full of specialized chips-in opening their technologies.
But Dibsic and others have shown that this ecosystem can flourish in ways that extend beyond American technology giants.
Why is this important?
Many experts have argued that major American companies should not open their technology sources because of It can be used to spread misleading information or cause another serious harm. Some American legislators discovered the possibility of preventing or suffocating this practice.
But other experts have argued that if the organizers penetrated the progress made in open source technology in the United States, China will get a great advantage. If the best open source technologies come from China, then these experts are arguing, then American researchers and companies Sibnion are their systems at the top of these technologies.
In the long run, this can put China at the heart of research and development of artificial intelligence, which may speed up its efforts to build a wide range of artificial intelligence techniques, including independent weapons and other military systems.
