• 목록
• 답변
• 글쓰기
• 게시판 리스트 옵션
  • 수정
  • 삭제

Palmer Luckey, the founding father of digital reality company Oculus VR, on Wednesday labelled DeepSeek’s claimed finances as "bogus" and accused too many "useful idiots" of falling for "Chinese propaganda". Lucas Hansen, co-founder of the nonprofit CivAI, mentioned while it was difficult to know whether or not DeepSeek circumvented US export controls, the startup’s claimed coaching price range referred to V3, which is roughly equal to OpenAI’s GPT-4, not R1 itself. DeepSeek mentioned that its new R1 reasoning model didn’t require highly effective Nvidia hardware to attain comparable performance to OpenAI’s o1 model, letting the Chinese firm practice it at a significantly decrease cost. Given OpenAI’s widespread use in business and education, the potential impression is regarding. Nvidia won't, nonetheless, must be redesigned to make use of HBM2 to proceed selling to Chinese customers. "GPT-4 completed training late 2022. There have been a variety of algorithmic and hardware enhancements since 2022, driving down the associated fee of training a GPT-4 class mannequin. "It’s plausible to me that they'll prepare a model with $6m," Domingos added. "It’s straightforward to criticize," Wang mentioned on X in response to questions from Al Jazeera about the suggestion that DeepSeek’s claims should not be taken at face value. While the authenticity of the claims stays unverified, cybersecurity experts warn that if true, the breach could expose sensitive user information, including industrial projects, tutorial research, and private communications.

However, even one of the effectively-resourced AI corporations is not immune to cybersecurity threats. Just two weeks after its official launch, China-based mostly AI startup DeepSeek has zoomed previous ChatGPT and change into the primary Free DeepSeek v3 app on the US App Store. DeepSeek just lately overtook OpenAI's ChatGPT as the top free app on the Apple App Store in the US and a lot of different nations. The free AI app is also expected to make a dent available in the market share of the top US AI firms and will result in significant price reductions. It seems to have similar functionality to market leader ChatGPT and it rocketed to the highest of app stores world wide. While there's broad consensus that DeepSeek’s launch of R1 a minimum of represents a significant achievement, some outstanding observers have cautioned in opposition to taking its claims at face value. OpenAI has yet to affirm or deny the breach but is taking the claims "seriously".

In a paper final month, DeepSeek researchers said that the V3 mannequin used Nvidia H800 chips for training and cost lower than $6 million - a paltry sum in comparison with the billions that AI giants reminiscent of Microsoft, Meta and OpenAI have pledged to spend this 12 months alone. Two databases were exposed with over a million traces of log streams containing chat history, API Keys, backend particulars, and other extremely sensitive information was uncovered. Tech billionaire Elon Musk, one of US President Donald Trump’s closest confidants, backed DeepSeek’s sceptics, writing "Obviously" on X underneath a post about Wang’s claim. Alibaba owns the South China Morning Post. DeepSeek returned an inventory of the five most important occasions as of October 2025, which included subjects like the Israel-Hamas conflict escalation and economic challenges in China. But then it added, "China is just not neutral in observe. Its actions (economic support for Russia, anti-Western rhetoric, and refusal to condemn the invasion) tilt its place nearer to Moscow." The identical question in Chinese hewed far more closely to the official line. "If they’d spend more time working on the code and reproduce the DeepSeek thought theirselves it is going to be higher than talking on the paper," Wang added, using an English translation of a Chinese idiom about individuals who engage in idle discuss.

OpenAI CEO Sam Altman has acknowledged that it price greater than $100m to prepare its chatbot GPT-4, whereas analysts have estimated that the model used as many as 25,000 extra superior H100 GPUs. He did not respond on to a question about whether or not he believed DeepSeek had spent less than $6m and used much less superior chips to prepare R1’s foundational mannequin. In a analysis paper launched final week, the DeepSeek improvement workforce stated they had used 2,000 Nvidia H800 GPUs - a less superior chip initially designed to comply with US export controls - and spent $5.6m to train R1’s foundational mannequin, V3. OpenAI, the corporate behind ChatGPT and different advanced AI models, has been a frontrunner in artificial intelligence research and improvement. ✅ Fair AI growth will be a key differentiator within the industry. DeepSeek’s success alerts a shift towards cost-effective AI improvement. A analysis note from investment and financial firm Jefferies mentioned DeepSeek’s success could go two ways: Either drive the industry to "pursue extra computing energy to drive even quicker model improvements" or "refocus on efficiency … The firm had started out with a stockpile of 10,000 A100’s, nevertheless it wanted extra to compete with companies like OpenAI and Meta.