Is AI’s real impact on stocks about margin expansion, not revenue growth? Looking for flaws in this thesis.

I’ve been thinking through what’s actually holding the market back right now, and I want to pressure test this with people who think critically about macro, earnings, and second-order effects.

My current view is that a big piece of market hesitation has been geopolitical risk plus uncertainty around how AI actually translates into broad public-market earnings. Lately, it feels like the geopolitical side may be easing a bit, which leaves the AI question as the bigger unresolved variable.

Here’s the core of what I’ve been wrestling with:

We have private and quasi-private AI/LLM businesses growing at a crazy pace. Companies like OpenAI and Anthropic appear to be scaling revenue extremely fast, and enterprises everywhere seem to be adopting LLM tools in some form, whether directly for employees or by building internal workflows and products on top of them.

But when I look across the broader market, I don’t yet see a clean “AI adoption -> obvious revenue explosion” showing up everywhere. That made me question the whole thesis at first. If these AI companies are growing so fast, where is the matching payoff across the rest of the economy?

My current answer is that I may have been looking in the wrong place.

Maybe the real impact is not that AI suddenly makes every company’s revenue explode. GDP can only grow so fast. At some point, even the fastest-growing AI businesses have to approach an asymptote simply because the economy is finite.

Instead, maybe what AI does across the S&P 500 is this:

• companies use AI to make employees much more productive
• some roles get partially automated, and some eventually get eliminated
• revenue may grow only modestly, or even stay relatively similar
• but labor cost per unit of output drops materially
• operating margins improve, potentially by a lot
• earnings rise even without some giant top-line acceleration

If that is the right framework, then maybe the market is still underestimating what broad AI adoption does to profits, even if it is overfocused on whether AI directly creates new revenue everywhere.

That leads me to the next part of the thesis:

If this happens broadly, then stocks could rip on earnings and margins, while at the same time unemployment rises because companies need fewer people for the same output. If enough labor gets displaced, it also seems possible that we get strong disinflation, or even deflationary pressure, because businesses can produce more efficiently and at lower cost.

That is where I’m less certain.

Questions I’m trying to think through:

• If AI is primarily margin expansion rather than revenue expansion, is that enough to justify a major leg up in equities?
• If companies become structurally more efficient, do those gains mostly accrue to shareholders, or do they get competed away through lower prices?
• Does broad AI adoption end up being deflationary, disinflationary, or weirdly still inflationary because of infrastructure/capex/energy constraints?
• If unemployment rises from labor displacement, how does the Fed even react if earnings are strong but labor weakens?
• Would rates stay higher because inflation is sticky elsewhere, or lower because AI becomes a deflationary force?
• Does government policy eventually have to offset the labor shock somehow?

My tentative view is that the market may be too focused on “where is the AI revenue?” and not focused enough on “what if AI massively lifts margins without needing revenue to explode?”

That said, I know enough to know there are probably holes in this argument, and I’d genuinely like to hear them.

I’ve done very well investing over the years by trying to identify moments where the market is misreading a big transition, and this feels like one of those moments to me. I’m posting this because I want the smartest criticisms of the thesis before I size into this read harder.

Where is this argument wrong, incomplete, or too early?