DeepSeek just flipped the AI script in favor of open-source—and the irony...
DeepSeek, the Chinese AI startup, shocked the world after unveiling an AI model last week that rivals leading models like OpenAI’s o1, while claiming it cost far less to develop and required far fewer Nvidia chips—and giving it away for free. The fallout sent Nvidia’s stock plummeting today and left observers wondering: What does it mean for the most deep-pocketed AI startups, OpenAI and Anthropic, which sell their models to consumers and companies, as well as highly funded competitors like Mistral and Cohere?
The current moment is deeply ironic
Toronto-based AI developer and consultant Reuven Cohen told Fortune that DeepSeek released its AI model as open-source, allowing researchers, developers, and users access to the underlying code and its “weights” to use, modify, or improve. This move challenges the closed approach taken by companies like OpenAI, which was initially founded as a nonprofit company that shared its research openly but has since become more closed off.
While DeepSeek did not release the training data for its R1 model, there are suggestions that it may have used outputs from OpenAI's o1 to enhance its model's reasoning abilities. This process, known as "reverse engineering," has been ongoing as open-source developers analyze and learn from models like o1.
Impact on existing business models
This shift towards more open-source AI development poses a significant challenge to companies like OpenAI and Anthropic, as well as their business models, cost structures, and technological assumptions. The success of DeepSeek suggests that models can self-improve by learning from others, potentially disrupting the market dynamics of the AI industry.
Patrick Moorhead, founder of Moor Insights & Strategy, notes that companies with proprietary leanings need a significant advantage to compete successfully in this evolving landscape. The rise of open-source AI could potentially impact the valuation and competitiveness of companies like OpenAI and Anthropic.
Implications for the future
William Falcon, CEO of Lightning AI, highlights the potential shift in the industry as companies may reconsider using OpenAI models at scale due to the capabilities and cost-effectiveness of open-source alternatives. This development raises questions about the long-term viability of existing AI companies and their revenue models.
Despite the challenges posed by DeepSeek's approach, experts like Daniel Newman and Vaibhav Srivastav believe that companies like OpenAI and Anthropic still have advantages in the application layer of AI development. The value lies not just in building models but in how they are integrated into real-world applications.
The role of open-source collaboration
While the emergence of DeepSeek as an open-source disruptor may challenge established players in the AI industry, it also underscores the importance of open collaboration in driving progress. Experts emphasize the need for continued innovation and learning from various sources to advance AI capabilities.
Despite the market disruptions caused by DeepSeek's approach, companies like OpenAI and Meta are likely to continue researching and adapting to maintain their competitive edge in the evolving AI landscape.
