Moats in the Age of AI (www.cjroth.com)

The recent unveiling of DeepSeek’s open-source AI model challenges the conventional notion that leading AI firms like OpenAI and Anthropic possess substantial competitive moats. By creating a competitive AI system for under $6 million, DeepSeek demonstrates that the barriers to building advanced models are eroding, as modern programming languages and AI tools now allow for rapid redevelopment of complex software previously thought to require extensive resources. This shift indicates that traditional software complexities and lengthy codebases may no longer safeguard against emerging competitors, compelling AI companies to explore secondary moats through proprietary platforms and unique training data. Furthermore, the ongoing scarcity of compute resources and data center infrastructure underlines a new landscape of competitive advantages. Companies like Nvidia thrive due to the soaring demand for AI chips, while those capable of securing vital resources such as energy and water find themselves in an advantageous position. The rise of on-device AI also introduces a paradigm where traditional cloud solutions may face increased competition, allowing companies like Apple to strengthen their position. As the AI market becomes more commoditized, successful players will likely rely on unique partnerships, infrastructure, and geographical advantages, redefining the notion of what constitutes a sustainable competitive edge in this evolving landscape.