China's DeepSeek V3 has emerged as a competitive player in the AI landscape, outperforming leading models from OpenAI and Meta on key benchmarks while being developed at a fraction of the cost. The model is reported to be 95% cheaper than ChatGPT and more powerful than GPT-4. Its rapid development and open-source nature have raised concerns about the implications for AI regulation in the US and EU, as it is expected to challenge the dominance of proprietary models. Additionally, a Chinese team claims to have reverse-engineered OpenAI's architecture, further democratizing AI technology. DeepSeek V3 has gained significant traction, with impressive download numbers for its 671 billion parameter model within just ten days. Western companies and governments are already assessing and adopting this state-controlled technology.
DeepSeek is very popular. And rightfully so. https://t.co/6zglQ5zQ4n
Deepseek-v3 is insanely popular. A 671B model's download numbers in just 10 days are unbelievable https://t.co/nfrgkkT27G
How China’s DeepSeek-V3 AI model challenges OpenAI’s dominance. https://t.co/bkrZCvt6JX