Every company is suddenly an "AI company." Every earnings call mentions machine learning. Every press release gets the magic buzzword treatment. And investors are finally starting to ask: where's the actual revenue?
A Reddit investor put it bluntly: "I'm tired of every company adding 'AI' to their press release just to pump their stock. Half of them are trading at 50x earnings but their actual revenue growth is mid at best."
They're not wrong.
We've hit the point in the AI cycle where the market is moving from the hype phase to the show-me-the-money phase. And a lot of companies are about to get exposed.
Let's name names.
Who has real AI revenue? Nvidia is the obvious one—they're selling the shovels in the gold rush, and business is booming. Microsoft has actual AI products generating real subscription revenue through Copilot. Google is integrating AI across search and cloud. These companies can point to line items on their income statements.
Who's selling vapor? That software firm trading at 60x earnings whose "AI strategy" is a chatbot they bolted onto their legacy product? The SaaS company that rebranded their analytics dashboard as "AI-powered insights"? The enterprise platform claiming AI will "unlock new revenue streams" without providing a single number? Yeah, those.
Here's how to separate signal from noise:
1. Look for actual AI revenue. Not "AI-adjacent" or "AI-enabled." Dollars that customers are paying specifically for AI features. If the company won't break it out in their earnings, assume it's negligible.
2. Check the valuation against reality. A company growing revenue at 15% shouldn't trade at 50x earnings just because they mentioned "machine learning" three times. Compare them to peers without the AI premium—if the gap is huge, you're paying for hope, not performance.
3. Watch for capital intensity. Real AI infrastructure costs money. If a company claims to be going all-in on AI but capital expenditures are flat, they're not serious. Meta, Amazon, and Microsoft are spending tens of billions on AI infrastructure. Posers aren't.
4. Listen to management tone. Companies with real AI traction talk specifics: customer adoption rates, use cases, retention metrics. Companies faking it speak in vague platitudes about "transformative potential" and "being at the forefront of innovation."
The Reddit post asked if anyone is rotating into energy or industrials until the dust settles. Honestly? That's not crazy. When a sector gets this frothy, taking profits and moving to something boring with actual earnings isn't cowardice—it's risk management.
I'm not saying AI isn't real. It is. The technology will reshape entire industries. But we're in the phase where the market has priced in a decade of perfect execution for companies that haven't proven they can execute for a quarter.
History is littered with transformative technologies that made early investors rich and late investors poor. The internet changed everything, but that didn't stop the dot-com bubble from obliterating portfolios. AI will be the same: massive long-term impact, brutal short-term reckoning for companies that can't deliver.
2026 is the year investors start demanding proof. Revenue growth. Margin expansion. Customer adoption. If a company can't show those, the AI premium is going to evaporate fast.
If they tell you "we're investing in AI for the future," ask them when the future starts paying dividends.
