The 2000 tech bubble was defined by massive overinvestment in unused telecom infrastructure ('dark fiber'). In contrast, today's spending on GPUs sees immediate, high utilization and positive ROI for the largest buyers, indicating a fundamentally healthier market driven by real demand.
Concerns about NVIDIA investing in startups that then buy its chips are overblown. These deals are a strategic necessity to support an ecosystem of NVIDIA users (like OpenAI) against Google, which leverages its own TPU chips to create captive customers (like Anthropic).
SaaS companies are trying to preserve high gross margins, but this is impossible if they want to succeed in AI, which is compute-intensive. Lower margins should be reframed as a positive signal of user adoption and AI integration, much like the successful transition from on-prem to cloud.
While AI feels disruptive, incumbent tech giants possess all the key ingredients to win: massive datasets, huge capital for compute, and vast distribution networks. This suggests AI could reinforce their market power, making it a 'sustaining' innovation rather than one that unseats them.
The AI hardware market isn't just about NVIDIA. It's a battle between NVIDIA's full-stack system, Google's powerful TPU, and a combined effort where Broadcom builds the networking fabric and custom ASICs, with AMD serving as a plug-in alternative chip.
Google's ad model thrives because advertisers systematically overpay, overvaluing their ability to retain customers. AI agents will eliminate this inefficiency by finding the best outcome for the user directly, shifting the business model to one based on affiliate fees for completed transactions.