Breaking the AI Monopoly: DeepSeek $6 Million Power Play

- Open-source large language model DeepSeek R1 shows high performance with just $6 million training cost - Major changes in the global AI competitive landscape predicted [Unblock Media] The open-source large language model 'DeepSeek R1' is demonstrating performance comparable to existing massive AI models with just a $6 million training cost, overturning the perception that China only produces closed AI systems. Moreover, with Nic Carter raising the possibility of the 'demise of OpenAI's moat', Trump's $500 billion 'Stargate' project, and UK Prime Minister Keir Starmer's data center development plans, significant changes in the global AI competitive landscape are predicted. As DeepSeek R1's cost efficiency and open-source innovation are highlighted simultaneously, developers who have spent billions of dollars building large AI models are facing new challenges. The AI hardware provider Lambda Labs estimated the computational workload required to train GPT-3 at approximately 3.14 exaflops (10^18 floating-point operations). Based on this, they calculated that the GPU usage costs would amount to over $12 million. A VentureBeat report noted that by achieving high performance with a relatively small amount of $6 million, DeepSeek R1 opens up the possibility for startups and SMEs with less capital to compete in the AI race. However, there are also voices expressing that the nature of open-source may raise security and ethical issues more easily, which requires further discussion. Nic Carter's remarks about the potential collapse of OpenAI's moat emphasize that DeepSeek R1 is more than just a low-cost model. Forbes, citing Carter's comments, analyzed that the market barriers built by global AI corporations with massive funds and data-computing resources might no longer be 'invincible' due to open-source models. Carter pointed out that the emergence of a high-performance AI model without major investments could signal disruption not only to OpenAI's monopoly but to the entire existing AI ecosystem. With the relaxation of AI regulations in China and the impact of U.S. chip export bans intertwining, the global AI market is becoming more dynamic. Although the U.S. banned the export of high-performance computing chips in October 2022, Chinese companies quickly responded by shifting to domestic chip manufacturing or bypassing it via cloud services. CNBC reported that the U.S. sanction, instead of slowing down China's AI industry development, inadvertently accelerated the establishment of a self-reliant ecosystem. Additionally, in August 2023, China relaxed AI development regulations and abolished financial penalties for companies operating outside industrial standards to more actively support innovation. Trump's 'Stargate' project and the UK's AI infrastructure expansion efforts are further heating up this global competition. On January 22, 2023, Trump announced the $500 billion 'Stargate' AI infrastructure project, invested in by OpenAI, Oracle, and SoftBank, with the ambition to make the U.S. the world's AI capital. This project includes plans to build high-performance computing and AI data centers in the U.S., creating over 100,000 jobs. Trump emphasized at the Davos World Economic Forum that deregulation in the energy sector would further boost this goal. Meanwhile, UK Prime Minister Keir Starmer also announced a similar strategy in January, presenting a blueprint to accelerate data center development and expand AI infrastructure to maintain global competitiveness. The reassessment of existing AI development processes and changes in the global competitive landscape seem inevitable. The example of DeepSeek R1 has proven that high-performance models can be implemented at low costs, and governments and corporations around the world are quickly responding to this trend by adjusting regulations and making large-scale investments. These movements may signal the end of an era where the capital strength of a few companies determines success and failure in the AI industry. However, as the pace of technological advancement accelerates, risk factors including ethical and security issues are also emerging. Major foreign media outlets like VentureBeat, Forbes, and CNBC commonly point out that the AI market is moving away from a monopolistic structure towards a new phase with numerous competitors.