Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

US may be losing edge to China in the AI race

Jan 24, 2025 - morningbrew.com
DeepSeek, a leading AI lab in China, has developed a new artificial intelligence model that rivals the best offerings from Silicon Valley, according to industry executives. The model, which focuses on reasoning, reportedly outperforms or matches OpenAI’s similar model in tasks related to math and coding, although these results have yet to be independently verified. DeepSeek's recent releases include a model launched on Christmas Day that competes with OpenAI’s GPT-4o, and their latest model, R1, is open-source and available for free, potentially attracting users away from subscription-based models like OpenAI’s ChatGPT Pro.

The significance of DeepSeek’s R1 model lies in its ability to learn from scratch using humanlike thought patterns, without imitating humans, as noted by Nvidia's Senior Research Manager, Jim Fan. This development marks a shift towards reasoning-focused AI models, which are prioritized by tech leaders aiming for artificial general intelligence. The open-source nature of DeepSeek’s R1 could disrupt the current AI landscape by offering a cost-free alternative to proprietary models.

Key takeaways:

  • DeepSeek, a Chinese AI lab, has developed AI models that are on par with top American models, according to industry executives.
  • DeepSeek's new reasoning-focused model reportedly outperforms or matches OpenAI's similar model in math and coding tasks, though results are not independently verified.
  • The new model, R1, is open-source and free, potentially attracting users from paid models like OpenAI's ChatGPT Pro.
  • DeepSeek's AI learns from scratch using humanlike thought patterns, which is significant for achieving artificial general intelligence.
View Full Article

Comments (0)

Be the first to comment!