“PMI improves not when AI is built, but when AI spreads, and we’re moving from brain builders to economy-wide beneficiaries.”
In this week’s video, I examine how 2024 closed with surprising strength with back-to-back quarters of 3.8%+ real GDP growth, a phenomenon that’s only occurred three periods this century. Not only did this occur with tariff fears but those two quarters also saw zero job creation. With two Fed cuts still projected for 2025 despite robust nominal GDP running at 8% annualized in the last quarter, we’re witnessing a productivity boom driven by AI despite the commentary you hear daily from economists, strategists and perma-bears. For 2026, it will become apparent that AI is hitting the physical infrastructure limits rather than compute constraints.
What’s changing now is not simply how much AI is being built, but where it is moving. The next phase of AI is shifting from pure cloud and software into the physical economy, edge devices, advanced packaging, on-prem infrastructure, power generation, and real-world execution. This transition broadens the opportunity set beyond mega-cap GPUs and data centers into hardware, materials, capital goods, and systems that connect intelligence to the real economy. Nvidia’s partnership with Groq highlights this shift toward architectural specialization, while enterprise AI agents, edge compute, and BYOG (bring-your-own-generation) trends signal that AI deployment is becoming more distributed, physical, and capital-intensive.
Taken together, these signals point to a global reflation regime, not a recession or stagflation outcome. Copper is breaking higher, global PMIs are improving, the dollar is weakening, financials are emerging from long-term bases, and liquidity remains supportive. This is the rare environment where technology, commodities, and real assets can rise together, a setup that historically has produced multi-year cycles. In that context, Bitcoin’s recent underperformance and deeply negative sentiment look more like consolidation than failure, especially as rising PMIs, a steeper yield curve, and a weaker dollar align with its strongest historical regimes. For investors, the takeaway is not to abandon AI but to expand the lens toward hardware, energy, financials, Bitcoin, and globally exposed cyclicals as AI’s impact moves from screens into the real economy.
Timestamps
- (00:00–07:03) GDP surprise & productivity boom: Q3 real GDP came in at 4.3%, following 3.8% in Q2 with zero job creation during both quarters. This is structurally similar to 2003-2004 post-dotcom recovery, not a bubble setup.
- (07:03–13:17) Reflation indicators: Copper rising, MSCI World ex-US closing at highs, dollar in downtrend, Korea/Taiwan exports surging, capital goods orders implying PMIs near 60 all pointing to global reflation, not recession.
- (13:17–22:29) Regime identification: Financials breaking out, multi-decade commodity bases awakening, gold rising alongside equities this is early-cycle reflation where tech and commodities rise together (only happened twice in 60 years).
- (22:29–27:06) Bitcoin setup: Despite sentiment at lows and 67 days below the 50-day MA, liquidity is positive and Bitcoin historically thrives in weaker dollar + rising PMI environments. Ethereum outperforming signals network effects kicking in.
- (27:06–32:04) Multi-decade breakouts: John Roque charts showing Citigroup (16-year base), Freeport-McMoRan, Exxon, DuPont (chemicals), mining stocks, Delta Airlines, Eli Lilly, Tesla—all breaking out or setting up.
- (32:04–36:31) AI bubble concerns vs. reality: Tech capex is Manhattan Project-scale, but balance sheets are strong, free cash flow solid. The real risk isn’t valuation, it’s whether revenues materialize in 2025-2026.
- (36:31–40:20) Market immune system & turbulence model: Built a covariance-based early warning system tracking 99 assets across stocks, bonds, commodities, crypto, currently not flashing red, suggesting the rally has room.
- (40:20–45:14) Andre Karpathy wake-up call: “I’ve never felt this much behind as a programmer” signals that AI agents and enterprise adoption are about to surge, driving hyperscaler revenues and forcing job redesign across all sectors.
- (45:14–End) AI collaboration mandate: The productivity leap is real, but requires daily habit-building with AI tools. This isn’t about answers, it’s about making better probabilistic decisions through systems thinking and dialogue.
