Recently, DeepSeek, a generated artificial intelligence (GENAI) chatbot, has confused the AI industry. A company based in China was established in 2023, but no one had heard it until January 24. Almost overnight, Deepseek became the hottest topic in the AI world.
The big topic of Deepseek was a claim that the AI model was built for only $ 5.58 million using the 2,000 NVIDIA (NASDAQ) H800 GPU, like Openai’s GPT-O1. It was asserted that it was a model that costs billions of dollars. make. If the Deepseek team tells the truth, it makes a big question to American investors and high -tech companies. Why are US companies spending billions of billions, even though Deep Shek seems to have pulled this in just a few?
Developers have also played an important role in becoming DeepSeek. This is because it is cheaper to build based on Deepseek, especially Openai.
Both of these elements have various concerns about the use of DeepSeek. Some people are worried about data privacy because DeepSeek has been reporting that it has captured user data from Americans and preserved it on a Chinese server. In addition, no one knows how the GENAI system captures the data is used or sold.
In addition, skepticism has questioned whether the company actually built this model for $ 5.58 million or hides the actual cost. If the company spends more to build a model, we are wondering how they avoided our chip restrictions. Another point of the controversy is whether Microsoft (Nasdaq) led by MSFT, whether DeepSeek has stolen Openai research and data to develop a model.
Each item has a major impact on Deepseek. If the company avoids the US chip export restrictions in some way, there will probably be legal results or other types of penalty for doing so. If DeepSeek simply copies Openai’s work, it looks much more innovative than initially suggested. As a result, Deepseek is not a newcomer in the AI industry, but is a framework as imitation, so it may lose some reliability.
Microsoft, Meta, and Openai have doubled AI spending
The trend does not seem to stop quickly, despite the concerns of the high -tech industry and investors overestimated US AI High -Tech giants, despite the concerns that have expressed over expenditure. The high -tech giant actually suggests that it will double in this news but to double.
During the latest revenue calls, Microsoft and Meta (Nasdaq) spent a total of $ 37.4 billion in the fourth quarter of 2024 in the fourth quarter of 2024. Money was mainly directed to what you need to continue and expand the operation of chips, data centers, and AI.
In 2025, Microsoft promises to spend $ 80 billion in AI infrastructure, and META plans to spend up to $ 65 billion. Mark Zuckerberg says that this expenditure is all long -term play and a long -term strategic advantage.
This large-scale expenditure occurs when Deepseek claims that GPT-O1-level models have been built for only $ 5.58 million. As a result, many people are wondering why companies such as Openai and Microsoft need billions of billions. The reason is that AI operations require a considerable amount of calculation, data, and energy to train and deploy a large amount of calculation, data, and energy. In fact, it may be necessary for not only American companies, but all AI companies that are seriously competing in that space need to spend hundreds of billions. 。 Suspicates believe that DeepSeek’s cost claims are far from the actual number, hiding the actual infrastructure costs or benefiting from private support.
Not only Microsoft and Meta are increasing AI spending, but Openai is said to be trying to raise $ 40 billion in a new financing round that values the company for $ 340 billion. This indicates that they are not worried about the low -cost competitors’ abilities of truly changing the industry.
Openai quietly develops Chatgpt Gov
It was OWN on all the topics of DeepSeek, but this week Openai released Chatgpt Gov. This new version of ChatGpt is specially designed for government agencies, and now can use Azure Commermercial or Azure Government Cloud to run AI models with its own private cloud infrastructure. 。
Governments and large companies hesitate to adopt GENAI for privacy concerns. Nobody wants to use confidential data for chatbot training. There is also the risk that personal information entered in the AI tool will be a future response. As a result, many companies and government agencies are completely banned from using ChatGpt for work -related tasks.
Openai is trying to solve the problem using Chatgpt Gov. Although it is dull compared to DeepSeek’s fuss and radioactive drops caused by it, there are much more easy -to -use AI products corresponding to the government and companies, and there are few in the meantime.
By providing government agencies a method of executing AI models in their own secure cloud, the government has a unique security to avoid potential risks and pitfalls that hinder these systems before. You can use Genai with confidence that you can manage privacy and compliance requirements. 。
In order for artificial intelligence (AI) to work correctly in the law and prosper in the face of an increase in issues, it is necessary to integrate an enterprise blockchain system that guarantees the quality and ownership of data input. Data. Check out the coingeek reports on this new technology and learn more why Enterprise blockchain becomes AI backbone.
Watch: A transformed AI application is coming
Title = “YouTube Video Player” FrameBorder = “0” APLAS = “Accelerometer; Clipboard-Write; Picture; Web-Share “ReferrerPolicy =” Strict-Origin-WHEN-WHEN-CROSS- ORIGIN “Allowfulscreen>