Altman assesses DeepSeek's AI

Altman assesses DeepSeek's AI

21 July 2025

OpenAI CEO Sam Altman has commented on DeepSeek AI, acknowledging the team's talent but expressing doubt that they have achieved significantly greater efficiency than OpenAI. Altman's remarks come amidst investor scrutiny regarding the substantial resources required for AI development.

DeepSeek AI gained attention for its cost-effective AI model, which rivals leading US models in various categories, including maths, science, and coding. DeepSeek's R1 model has been recognised as impressive, particularly for its performance relative to its price. However, Altman maintains confidence in OpenAI's ability to deliver superior models and views the competition as a positive motivator. DeepSeek focuses on efficiency, explainability and customisable models. Its R1 model uses a hybrid training approach, combining reinforcement learning with supervised fine-tuning.

DeepSeek's models are designed to be adaptable and customisable, potentially offering superior performance and value in specific niches. DeepSeek AI distinguishes itself through its focus on efficiency, performance, and accessibility, offering high-performance models at a lower cost than OpenAI's GPT-4. Its open-source approach further enhances its appeal to developers and businesses seeking cost-effective AI solutions.

AI generated content may differ from the original.

Published on 21 July 2025
aiopenaideepseekartificialintelligencemachinelearning
  • GPT-5: OpenAI's AI Ambition

    GPT-5: OpenAI's AI Ambition

    Read more about GPT-5: OpenAI's AI Ambition
  • AI Models' Reasoning Transparency

    AI Models' Reasoning Transparency

    Read more about AI Models' Reasoning Transparency
  • AI Revolutionises ARDS Treatment

    AI Revolutionises ARDS Treatment

    Read more about AI Revolutionises ARDS Treatment
  • AI Copyright Law Scrutiny

    AI Copyright Law Scrutiny

    Read more about AI Copyright Law Scrutiny
Altman assesses DeepSeek's AI