For three years, investors fixated on the inflationary side of artificial intelligence: more compute, more data centres, more semiconductors, more electricity, more capex. More than $650 billion will pour into AI infrastructure this year alone — spending once reserved for governments and wars.
But the most important consequence of AI may prove to be the exact opposite. Deflation. Oil is falling. The Iran risk premium is fading. European growth is more resilient than expected. And quietly, AI is moving from the laboratory into the operating model of the firm.
The first phase of the supercycle was about building intelligence. The next phase is about using intelligence to eliminate cost. The market is rotating — from AI creation to AI monetization, from capex to profit expansion, from intelligence scarcity to intelligence abundance.
Five market signals
Tap each signal to expand the read.
-
Oil keeps falling, risk assets are recovering, and markets are stripping out the geopolitical premium that dominated recent weeks. The immediate inflation threat is fading and the macro backdrop is turning more supportive for risk-taking.
Signal Markets are rotating from geopolitical fear back toward growth and earnings.
-
Germany is still the weak link: services deteriorating, consumer demand fragile, auto-sales growth slowing. Yet the broader Eurozone keeps dodging the recession many expected. Europe isn’t strong — it’s simply more resilient than consensus feared.
Signal Europe’s slowdown is a growth problem — not yet a recession problem.
-
European tech is losing momentum — not because AI is ending, but because investors are turning selective. The first phase rewarded almost anything associated with AI. The next will reward companies that produce measurable economic returns from it.
Signal The market is separating AI infrastructure, AI productivity and AI hype.
-
While investors watch AI spending, Oracle is cutting 21,000 jobs — perhaps one of the most important AI stories of the year. First-order effect: productivity. Second: labour displacement. Third: margin expansion. Fourth: earnings growth. This is what deployment looks like.
Signal The AI productivity cycle has started.
-
European auto demand is weak, yet Tesla is recovering. This is no longer a story about vehicles — it’s about platforms: software, data, autonomy, charging networks, customer relationships. The value pool of mobility is migrating from manufacturing toward orchestration.
Signal Platforms are becoming more valuable than products.
Four hidden patterns
The structure underneath the signals.
AI is becoming deflationary
The market still fixates on AI capex. The next phase is AI cost reduction — cutting labour requirements, compressing decision cycles, automating workflows, removing friction. The next winners may not build intelligence; they’ll remove the most cost from the system.
The great rotation has started
The first AI trade rewarded suppliers selling accelerated computing. The next may reward users deploying intelligence to improve margins and returns on capital — from infrastructure construction to productivity extraction, from investment to monetization.
Europe may benefit more than expected
Consensus says Europe is losing the AI race. Too simplistic. It lags in models and hyperscale — but leads in industrial base, manufacturing, transport, utilities, healthcare and finance: a dense concentration of AI beneficiaries. The race is shifting from creation to application.
The next repricing comes from earnings
Three years priced AI as a capex story — data centres, GPUs, power, semis. Markets ultimately reward earnings, not spending. As AI moves into operations, the impact surfaces in margins, free cash flow, returns on capital and EPS. That’s where the next repricing occurs.
The market is looking at the wrong side of AI
Spending is yesterday’s story; profitability is tomorrow’s. The first phase demanded enormous investment in compute, data centres, energy and AI factories. The next may generate something more powerful — a global productivity shock, a margin-expansion cycle, a deflationary wave capable of reshaping industries, labour markets and financial markets.
Most investors still ask who is building intelligence. The more important question is who will profit from the falling cost of it. The first AI trade rewarded builders and spending. The next may reward operators and profitability.
What to carry into the next session
- Oil is falling and geopolitical risk is fading.
- Europe is weaker than America but stronger than expected.
- The AI trade is becoming increasingly selective.
- Oracle’s layoffs may be an early signal of large-scale AI deployment.
- AI is evolving from a capex story into a productivity story.
- Lower costs may matter more than larger models.
- The market is rotating from AI builders to AI beneficiaries.
- Europe may benefit more from deployment than valuations imply.
- Earnings growth is overtaking infrastructure spending.
- The next winners remove the most cost from the economy.
Markets spent the last three years pricing the cost of building intelligence. The next decade may be defined by the profits generated by its deflationary power.