The main technology analyst of the I / O fund, Beth Kindig, discusses competition between Deepseek and Chatgpt on the money winner.
In China Artificial Intelligence (AI) The exit by startup Deepseek of new AI models which competes with companies at the top of American technology companies did a tour of the markets on Monday and aroused concerns about American companies to lose their advantage in the AI ​​race to Chinese rivals.
Deepseek published his model R1 AI last week which, according to him, is 20 to 50 times cheaper to use than the O1 model of Chatgpt-Maker Openai, according to the task, according to an article on the Official WeChat account of the Company .
Deepseek’s ai The assistant, propelled by his new Deepseek-V3 model, vaulted after Chatgpt at the top of the Apple App Store. The company said that the model had been formed with less than $ 6 million in calculation power of what it said was 2,000 NVIDIA H800 chips to reach a level of performance equally with the most advanced models of OPENAI and META. The startup also deployed its updated image generation model called Janus-Pro Monday.
The progress made by Deepseek with what he declared to be lower, paces at lower capacity and a lower cost than expenses for the training of AI by American competitors caused a market sale and a debate on the question of whether the Chinese company has reversed the advantage of American companies in the edge of the company Race AI.
Startup of startup Deepseek in the face of hacking, blocks questions about the CCP
The DEEPSEEK AI chatbot blocked the questions criticizing the Chinese Communist Party, found a Fox Affairs review. (Illustration Photo by Justin Sullivan / Getty Images / Getty Images)
Marc Andreessen, co-founder and general partner of the venture capital company Andreessen Horowitz, wrote in an article on X that “Deepseek R1 is the spoutnik moment of AI”, in reference to the first advance of the Union Soviet on the United States in the spatial race of the Cold War era of the Cold War era ,.
The co-founder of Anthropic, Jack Clark, wrote in his newsletter “Import Ai”, R1 is significant because it largely corresponds to the O1 model of Openai on a range of reasoning tasks and questions the idea that Western IA societies have an important lead in Chinese tasks. “
While some figures in the technological sector and investors in the space of IA see the progress of Deepseek as indicating the arrival of a new phase of AI competitionOthers are less convinced that it poses a wide challenge to the American technological industry.
What is the AI ​​Chinese Deepseek application?
The Chinese AI startup Deepseek has gone around the markets with the release of its new AI models. (Getty Images / Getty Images)
Dan Ives, Managing Director and World Manager of Technological Research at Wedbush Securities, wrote in a note on Monday to investors that, although the LLM of Deepseek has clearly impressed the technological sector, it should not be considered a rival for more capable companies like Nvidia on this subject scene.
“No US Global 2000 will use a Deepseek Chinese start-up to launch their infrastructure and use of AI,” wrote Ives. “In the end, there is only one flea company in the world by launching autonomous, robotic and wider use of AI and it is Nvidia. The launch of a competitive llm model For the use of consumers is one thing … The launch of a wider IA infrastructure is a wider IA infrastructure is an integer, another ball game and nothing with Deepseek makes us believe something different. “”
Ives sees the sale of the technological sector stimulated by the emergence of Deepseek as a possibility of investing in technological companies that are active in AI space.
“These are only the opportunities to have the Nvidia, Microsoft, Alphabet, Palant, Salesforce, Amazon and wider ecosystem that is subject to strong pressure,” he wrote in the middle of the sale on Monday . “Deepseek impressed the technological community by this LLM model … but that does not launch 100x the capacity / algorithms which is necessary to even consider this as a competitive threat in our opinion.”
Chinese Application Deepseek Hammers Us Stocks with a cheaper open-source IA model
The AI ​​assistant of Deepseek dominated the OpenAi chatpt in the Apple App Store. (Christoph Dernbach / Picture Alliance via Getty Images / Getty Images)
Mark Malek, CIO to Siebert, observed in a note that while Openai chatgpt And other models have “large capacities in natural language treatment, while Deepseek is created to be specific to the task”.
Malek asked in his note: “Who or what was disputed by the release of Deepseek this weekend? Was it hyperscalers, data security companies, network equipment manufacturers, flea manufacturers, flea manufacturers, flea manufacturers, flea manufacturers, flea manufacturers, flea manufacturers, flea manufacturers, IC design software suppliers, AI users, etc.? ”
“No, in the opinion should be LLM (large language model), AI models like Openai, Google Deepmind, Anthropic (Claude) and perhaps the Meta Lama,” he explained. “Now let’s remember that these are software With vast resources that can easily change their algorithms to reflect the current state of research. “”
Get Fox Affairs on the move by clicking here
“Did Deepseek seek and find a more effective treatment model for AI?
