So there's this interesting tension playing out in the market right now that's worth paying attention to. Tech stocks are getting hammered, but for completely contradictory reasons.

On one side, investors are spooked by how much money big companies are throwing at AI infrastructure. Amazon just announced $200 billion in AI capex for 2026, and the stock tanked about 9% year to date. Microsoft reported solid earnings with 17% revenue growth and 21% operating income growth, but the market punished it the next day because management said they're spending over $100 billion on capex this year. Down 17% year to date.

But here's where it gets weird. While mega-cap AI companies are getting sold off for spending too much, software stocks are getting crushed for a different reason entirely. SaaS companies like Salesforce and Adobe are in freefall because traders are convinced that agentic AI is going to make traditional enterprise software obsolete. They're calling it the "SaaSpocalypse."

The logic goes: if AI agents become powerful enough, why would companies need to pay for specialized software anymore? It's a compelling narrative if you buy it. But some heavy hitters in the industry aren't convinced. Nvidia's Jensen Huang basically called it illogical. His point makes sense too—SaaS companies built their software around specific industry problems that require real expertise. General-purpose AI agents might not replicate that specialized value. More likely, AI and software companies end up partnering, not competing.

If you think the software sell-off is overdone, there's an ETF play worth considering. The iShares Extended Tech-Software ETF gives you exposure to 114 North American software companies in one fund. Top holdings include Microsoft (9.7%), Palantir (8.2%), Salesforce (7.7%), Oracle (7.2%), and Intuit (5.2%). The fund's been averaging 10.4% annual returns since 2001 with a 0.39% expense ratio. It's trading at a 35.2 P/E ratio, which is a bit elevated compared to the Nasdaq-100's 32.4, but that's actually reasonable given the sector's growth profile.

What makes this interesting from an e learning perspective is how it forces you to think through the actual mechanics of AI disruption. You're not just betting on a thesis—you're making an educated decision about whether the market's fears are proportional to the actual risk. That's the kind of critical thinking that separates casual traders from investors who actually understand what they're buying.

The real question is whether both these AI narratives can be true at once. Is AI so transformative that it destroys one of the world's most profitable industries while simultaneously justifying massive capex spending? Probably not. Something's got to give, and if you're betting the software sell-off is overdone, an ETF like this lets you take a concentrated position without picking individual winners.