Another Kimi K2 test. I asked it to research Ed Zitron's AI Bubble thesis: The Core Argument: Discrepancy Between AI Revenue and Capex Zitron's central thesis posits a significant and unsustainable disconnect between the massive capital expenditures being poured into generative AI and the comparatively minimal revenues being generated. Unlike previous technological advancements like cloud computing, which demonstrated scalable utility and profitability, generative AI is viewed as a service layered on existing infrastructure with a vastly overestimated total addressable market. "This entire bubble has been inflated by hype, and by outright lies perpetuated by a tech media that's incapable of writing down what's happening in front of their faces." - Not sure whether the numbers below are correct (hallucination), but the rough idea is that companies are spending ~$100B in yearly capex vs less than ~$10B in genAI revenues. Both OAI and Anthropic are losing money and that is likely to continue for some time. The MSFT revenue number below is questionable. Only Nvidia is profitable in this business, and their future sales depend on the eventual profitability of the others. Enterprise genAI usage remains modest, so everything rides on the HOPE for huge future improvements in the usage and economic value of AI agents and such.