February 25, 2025
By Anjali Kochhar
Since its January 2025 launch, DeepSeek AI has continued to shake up the global tech landscape, challenging the traditional dominance of Silicon Valley. Its open-source approach has spurred innovation and forced the industry to rethink the closed-door policies of Western tech giants, writes author Chandran Nair in an article in the South China Morning Post.
DeepSeek-R1 distinguished itself not only by being open-source but also through its remarkable efficiency and cost-effectiveness. Utilising reinforcement learning and a “mixture of experts” technique, the model activates only the necessary networks in response to specific prompts. This selective activation significantly reduces computing power requirements, making the model more accessible and affordable.
“DeepSeek is a gesture of goodwill, and in that sense a deep provocation. It is also a symbol of what happens when bullies hard-wired with a sense of exceptionalism and privilege are confronted by a competitor that refuses to be intimidated or sidelined,” Nair shares his take in the article.
Notably, DeepSeek reported training its V3 model at approximately $6 million, a fraction of the $100 million expenditure by OpenAI for GPT-4 in 2023.
The release of DeepSeek-R1 elicited a spectrum of reactions worldwide. While some praised its open-source approach, others raised concerns about potential security risks and intellectual property issues. Critics accused DeepSeek of bypassing export controls and appropriating intellectual property from Western companies.
However, Nair notes these criticisms often overlooked the fact that companies like OpenAI had previously trained models on vast internet data without explicit consent, a practice that had come under legal scrutiny.
In the financial sector, DeepSeek’s emergence had immediate repercussions. Nvidia, a leading supplier of AI hardware, experienced a significant stock sell-off, with shares dropping 17% due to concerns over reduced demand for its products, shares the author in their article at SCMP.
However, Nvidia’s CEO, Jensen Huang, countered this narrative by emphasising that DeepSeek’s innovations would, in fact, increase the demand for AI hardware. He highlighted that DeepSeek’s reasoning feature is compute-intensive, necessitating greater computing capacity and, consequently, more advanced hardware.
Beyond the corporate sphere, DeepSeek’s influence extended into academia. Leading Chinese universities, including Shenzhen University and Zhejiang University, integrated DeepSeek’s models into their AI curricula. This initiative aligns with China’s strategic objective to become a global leader in AI education and research, fostering innovation and practical application among students.
Looking ahead, DeepSeek’s commitment to open-source development and cost-effective AI solutions positions it as a formidable player in the AI industry. Its approach has prompted a reevaluation of existing AI business models, with discussions emerging about the sustainability of proprietary systems.
Experts suggest that the industry may witness competitive dynamics similar to those in the budget airline sector, where cost-cutting measures could impact user experience.
DeepSeek’s launch of its open-source AI model has not only challenged Silicon Valley’s hegemony but also ignited a global discourse on the future of AI development. By prioritising transparency, efficiency, and collaboration, DeepSeek exemplifies how innovation can be achieved beyond traditional Western paradigms, potentially leading to a more inclusive and dynamic technological ecosystem.
About the author
Anjali Kochhar covers cryptocurrency and blockchain stories in India as well as globally. Having been in the field of media and journalism for over four years now, she has developed a sharp news sense and works hard to present information that goes beyond the obvious. She is an avid reader and loves writing on a wide range of subjects.
