Gigawatts, not Clicks: The upcoming decade
If you look at the last 15 years, the winning trade was almost always pure software. Scalability was cheap, distribution was instant, and "Software is eating the world" mantra was all you really needed. You could ignore the physical world and still win.
AI has changed that game, but the change had to happen regardless. You can ignore physical reality for only so long. We are moving from an era of "Clicks" to an era of "Gigawatts." Every new model training run is essentially a massive energy auction. The limit on growth is no longer how many users you can acquire, but how much power you can pull from the grid.
America is still ahead in the game, no matter the Deepseek moment, but China is not far behind. The geopolitics is shifting at an unprecedented rate. We are moving towards true multi-polarity, which means less of "dollar inflation is global taxation" and more of "onshoring manufacturing".
The upcoming decade's multibaggers will coast on these trends. Away from America-focus to global-focus. Companies that build the physical infrastructure - chips, energy, materials, and more. Trump wants Greenland for a reason. Uranium and copper are more interesting than the next B2B SaaS. Gold, silver, and Bitcoin are more interesting than a new social media app.
How are you guys positioning? What are your physical world plays?
2 Comments