DeepSeek’s Sustainability AI Approach Shakes Silicon Valley Raw Power Move 

Since its release, DeepSeek did not shy away from utterly disrupting the global tech market, causing some of the biggest companies’ stocks such as Nvidia to drop 17% on January 27 and igniting a wave of fear over Silicon Valley on whether Chinese AI models’ cost-efficient approach can surpass raw power. But what does DeepSeek’s quick rise to fame mean for the industry’s priorities, such as the development of sustainability AI. 

AI for Sustainable Development 

DeepSeek’s quick rise to fame was largely driven by its capacity to reduce computational costs related to training AI models, contrary to other US AI models that require huge amounts of energy and water resources. 

ChatGPT is a great example of its effect on sustainability in AI as it generates approximately 260 tons of CO2 monthly, demonstrating the toll AI has on the environment. 

On the other side, DeepSeek asserts it has developed a high-performance model that prioritizes AI in environmental sustainability with high performance. The Chinese company’s true innovation lies in its “test-time scaling” technique that dynamically optimizes computing resources during model operation. 

DeepSeek’s technology was highlighted by Nvidia CEO Jensen Huang at CES 2025, where he emphasized that test-time scaling, combined with pre-and post-training scaling is shaping the future of AI industry. 

With focus on AI environmental sustainability, DeepSeek has reduced the cost of training its models, while OpenAI’s GPT-4 reportedly cost over $100 million to train, when DeepSeek’s R1 model was trained for just $6 million. 

OpenAI CEO Sam Altman congratulated DeepSeek on X, acknowledging the model’s development for a fraction of the price. 

DeepSeek’s success in attaining sustainability AI also overcomes US export restrictions. Liang Wenfeng, the founder of DeepSeek stocked up 50,000 Nvidia A100 chips before the US ban in 2022. Experts acknowledged the high-end chips that DeepSeek is using with cheaper processors. But this does not mean that experts do not urge caution.  

Indul Hassan, Head of Engineering at BT Group, made it clear that DeepSeek’s success must be measured over time, moreover, the “mixture of experts” approach adapted by the company is not fully new, and other companies have already adopted it. 

Furthermore, there have been privacy concerns over DeepSeek, because it gathers and shares large amounts of user data.  

Sustainable Solutions for AI 

DeepSeek only highlights startups’ opportunity in the industry, proving that eventually, the driving forces will be all about sustainability AI and cost efficiency.  

Franny Hsiao, AI Architect at Salesforce, addressed how global competition within AI should inspire more companies to emerge, solving the very big problems in ways that serve AI and environmental sustainability and cost-effective.  

US and Chinese tech companies are in a continuous battle that is escalating over time, however, DeepSeek has shed light on AI and computational sustainability which may shape AI for the upcoming years.  

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Tech sections to stay informed and up-to-date with our daily articles.