AI's Bubble Potential & The Energy Crunch

Summary

AI is essentially converting electricity into intelligence, and with an seemingly infinite demand for intelligence, the demand for power is also exploding. Investors should critically assess management's foresight in this rapidly evolving tech landscape. While the immediate future may see increased unemployment for knowledge workers due to AI advancements, the long-term outlook suggests adaptation and integration. The current AI boom is compared to 1995, with the potential for a bubble larger than the dot-com era. A key catalyst was Anthropic's release of Opus 4.6, making coding tools accessible, which has directly fueled the semiconductor rally. Models are rapidly evolving, with current ones akin to early-career professionals, and projected to reach seasoned expert levels by year-end. Enterprise adoption, particularly for coding and productivity tools, is accelerating, leading to companies burning through AI inference budgets much faster than anticipated. Individuals are advised to embrace AI to enhance their roles, as those who don't risk becoming irrelevant. Globally, there's a race among AI models, with GPT-5.5 currently leading, followed by Opus 4.7, Gemini 3.1, and Grok 4.3. The demand for AI infrastructure, especially energy, is a critical bottleneck, with nuclear power and advanced energy storage expected to play significant roles. Hyperscalers are drastically increasing capital expenditures, signaling continued expansion of AI's brainpower. Beyond GPUs, the energy play in AI infrastructure remains largely underestimated by Wall Street. The potential for massive IPOs from companies like SpaceX, Anthropic, and OpenAI could reshape market indexes. Investors are advised to believe in management's competence to navigate the future of AI and its associated demands.