The world of artificial intelligence (AI) is in a state of flux, with the rise of open-weight models from China challenging the dominance of closed models from the US. This shift has significant implications for the future of AI hardware investments and the revenue streams of AI model makers. While there has been a lot of hype surrounding the capabilities of these new models, the reality is that the growth in annualized revenue run rates for AI companies like OpenAI and Anthropic is not keeping pace with the massive investments being made in datacenters and hardware. This raises a deeper question: are we witnessing another AI bubble, or is this a genuine transformation in the industry? In my opinion, the answer lies somewhere in between. The growth in AI model revenues is slowing, and the market for AI platforms is growing much slower than for the models themselves. This suggests that the industry is maturing, and AI is becoming an extension of traditional machine learning and even high-performance computing (HPC). However, the fact that Nvidia is the only company that can afford to give away its foundation models raises concerns about the long-term growth of GenAI model licensing and subscriptions. The company's monopoly power in AI hardware could be used to underwrite its free models, potentially leading to legal issues. The future of AI model makers and the industry as a whole depends on finding a balance between open-source models and closed foundation models. The challenge is to create a sustainable business model that allows for innovation and competition while also generating sufficient revenue to cover the enormous costs of AI hardware. Personally, I think that the key to success lies in finding a middle ground between open-source and closed models, where companies can license their own models and control their own GenAI fates. This approach would allow for experimentation and deployment of AI tools while also generating revenue through software and support licenses. In conclusion, the future of AI is uncertain, but one thing is clear: the industry is at a critical juncture, and the decisions made today will shape the future of AI for years to come. The challenge is to find a balance between innovation and sustainability, and to create a business model that can support the growth of AI while also generating sufficient revenue to cover the enormous costs of AI hardware.
