Elon Musk just co-signed one of Cathie Wood's boldest economic predictions ever, but the current GDP data tells a very different story, and the gap between her projections and reality raises serious questions about who wins and who gets crushed…
Over the weekend, Elon Musk endorsed one of Cathie Wood’s most aggressive economic claims in years. On Saturday, Musk quote-posted ARK Invest chief futurist Brett Winton on X with a four-word verdict: “The AI riptide is already underway.” Wood then quote-posted Musk with a longer thesis about AI-driven growth, and Musk circled back and replied, “You are right.”
The exchange unfolded on August 29, 2026. Winton kicked off the thread. Musk amplified him. Wood then wrote, verbatim: “When I try and explain that tokens inferenced increased 25-fold in a year and are multiplying exponentially, cascading through the economy, and are likely to push real GDP growth per year toward double digit territory, many clients and companies simply do not comprehend.” The post drew 1.1 million views, 7,789 likes, 756 reposts and 409 replies, and can be read here.
What “25-Fold” Actually Measures #
The 25-fold number is ARK’s own proprietary tokens-inferenced metric, not a government or third-party statistic. Wood translated the scale in a follow-up post: “Because they have not experienced 25-fold, most investors hear the number 25 attached to growth and mentally note 25% which, in the context of history, is really good growth. 25-fold is in another investment orbit: 100 units increasing, not to 125, but soaring to 2500!”
Per ARK’s Big Ideas 2026 research, AI compute demand surged by a similar order of magnitude while unit inference costs fell more than 90% over the same period. ARK’s 2026 Outlook cites inference costs dropping up to 99% per year on some benchmarks and training costs falling roughly 75% annually.
GDP Growth Numbers, Disentangled #
Wood has made different claims about AI fueled GDP Growth in different places.
Real GDP Growth, near double digits: This Wood said directly to Musk in her recent thread. Real GDP is adjusted for inflation.Nominal GDP Growth, 6% to 8%: From ARK’s 2026 Outlook commentary, attributed to 5-7% productivity growth, roughly 1% labor-force growth, and an inflation range of negative 2% to positive 1%. Nominal GDP includes inflation, which is why it runs higher than real GDP.Real GDP Growth, roughly 5%: A “Goldilocks” 2026 scenario Wood has floated elsewhere.
Reality Check on the Numbers #
Actual U.S. output is not tracking any of those scenarios yet. Real GDP growth came in at 1.5% in the second quarter of 2026, following the preceding quarter’s reading. That is the current-growth benchmark against Wood’s longer-run scenarios.
The inflation side of ARK’s model faces visible pressure. Nominal GDP is real growth plus inflation, so ARK’s negative 2% to positive 1% inflation assumption carries enormous weight in the 6% to 8% nominal range. Recent readings run well above that band. Headline PCE inflation was 3.7% year over year in July 2026, with core PCE at 3.34%, and the Fed’s 2% annual target remains the reference. Reuters, on August 26, 2026, ran the headline “US inflation remains elevated as GDP growth outlook brightens.”
Winners and Losers, Stock by Stock #
Wood’s follow-ups reframe the thesis as a stock-picking argument rather than a rising-tide one. She wrote that revenue at frontier AI labs is multiplying 5-fold and 10-fold over six months to one year, and that revenue growth at more mature companies on the right side of AI change is reaccelerating from 25% to 30-40%+, adding that she is seeing more “deer in the headlights.”
Her closer: “Half the solution is understanding the AI problem and the AI opportunity, company by company: revenue growth, productivity and costs, interest coverage. ‘Facts are stubborn things.’ The equity market psychology is shifting from ‘AI hype’ to ‘oh NO’ and ‘oh YES’, stock by stock.”
Winton, whose post seeded the thread, has argued that AI infrastructure’s fast payback periods and very high internal rates of return could raise the broader cost of capital, potentially hurting companies without direct AI exposure. That is a caution about who pays for the buildout, and it is also why the power, cooling, and networking suppliers behind the data centers keep surfacing in our free report on seven AI infrastructure stocks that aren’t chipmakers.
What remains is a genuinely bold long-term case, now with Musk’s endorsement, set against a 1.5% real GDP print and PCE inflation running above 3%. The gap between the projection and the current reading is the story worth watching.
Contact [email protected] for any questions or corrections.