Investors are moving beyond the glamour of GPUs toward AI software and services with predictable revenue and margins — here’s where the smart money is looking next.
Illustration by IMF Alpha editorial · Reviewed by Pedro Marini
Wall Street has a new obsession: sustainable AI profits, not just chip-led growth.
Last year felt like a hardware festival — data-center racks, H100 backorders, valuation multiples that treated semiconductors like an endless growth story. That craze hasn't vanished. But quietly, investors are starting to prize recurring revenue, gross margins and real pricing power in AI software and platforms.
This isn't a rejection of chips. Think of it as a market maturing. Chips spark activity; software is where the durable money lives. The pattern resembles the cloud transition in the 2010s: first the infrastructure frenzy, then a drift toward SaaS businesses that turned usage into predictable revenue. The similarity is striking.
Why it matters
Three signals worth watching
Examples and caveats
A quick investor checklist
A human note: this shift is partly behavioral. After a year of breathless chip coverage, portfolio managers want stories they can explain to pension committees. They want evidence — contracts, unit economics, and the often-murky but decisive art of sales execution.
If chips supplied the fuel, the next profitable leg of AI will be written by companies that actually turn capability into cash flow. Look past the fireworks and into the income statement; that's where the lasting winners will be decided.
As lawsuits and privacy rules squeeze scraped training sets, synthetic data firms are drawing capital and corporate deals. Practical wins, hidden risks.
From web-scraping lawsuits to paid, privacy-preserving feeds and synthetic substitutes — firms are buying better data to train safer, more valuable models.
Smaller models, smarter chips and privacy-first apps are turning phones and PCs into autonomous AI hubs — and the ripple effects will hit chips, apps and search.