I've been watching a documentary on the dot-com bubble while monitoring my positions in SMH and Intel, and I can't shake the feeling that everyone is asking the wrong question.The question isn't "is AI a bubble?" The question is: how do you make investment decisions when you genuinely can't tell if you're in the third inning of a generational boom or the final chapter of a mania?Here's what's making this so difficult to read.The bull case is real: Hyperscalers like Amazon, Microsoft, and Google just posted massive earnings beats, with most of their growth coming from cloud and AI services. They're not pulling back on capex - they're increasing it. NVIDIA posted another blowout quarter. TSMC and AMD's CEOs are talking about CPU demand doubling year-over-year. Intel is on track to compete with TSMC by 2029 on advanced chip technology. The companies making money from AI are actually making money, not just talking about it.The bear case is also real: Consumer sentiment on AI is terrible. People hate it, see it as a threat to jobs and creativity. Token spend at some companies costs more than just hiring humans. Data centers are being abandoned. Valuation multiples are stretched across the sector. Every new technology that displaced humans faced backlash, but that doesn't mean the technology always won.And here's what's paralyzed me and a lot of other investors: both narratives have evidence backing them up.I see Anthropic 10x-ing revenue. I see headlines about companies being forced to adopt AI finding it more expensive than alternatives. I see analysts calling this "the 3rd inning with a man on 2nd." I see other analysts calling it a bubble built on circular accounting and hype.Meanwhile, semiconductor companies keep posting record results. But 73% of AI capex plays have underperformed the S&P this year. Intel is positioned beautifully for the agentic AI future. But will that future actually materialize on the timeline bulls expect?The dot-com comparison is both helpful and misleading. Yes, there are similarities - massive capital deployment, sky-high valuations, genuine technological transformation happening alongside pure speculation. But there are also differences. The internet in 1999 was mostly promise. AI in 2026 is already generating real revenue and productivity gains.So what's an investor supposed to do?Here's the framework I'm using, and it's not about predicting the future - it's about managing risk while staying exposed to potential upside.First, accept that you don't know. Anyone who tells you with certainty that AI is definitely a bubble or definitely the future is either lying or delusional. The smart investors I know are all saying the same thing: If you can't tell if it's the 3rd inning or the 9th, don't bet your entire portfolio on either outcome. I have significant exposure to SMH and Intel because I believe the semiconductor story has legs, but doesn't mean When companies start missing earnings, when capex gets pulled back, when actual revenue growth slows - those are signals. Right now, the hyperscalers are still spending and still growing. When that changes, the narrative will change fast. Don't just own NVIDIA. If you believe in AI, own the picks and shovels (semiconductors), the infrastructure (cloud providers), and maybe some applications. If one part of the story breaks, you're not completely exposed. Not a plan, but a plan. Mine is: if hyperscaler capex guidance drops meaningfully or if semiconductor revenue growth turns negative for two consecutive quarters, I trim by 30-50%. Your plan might be different, but have one.The truth is, we're all making educated guesses here. The AI trade could continue for years. It could end tomorrow. Both outcomes have happened in market history.What separates good investors from bad ones isn't knowing the future - it's managing risk while staying positioned for opportunity. You can believe AI is transformational and still hedge your bets. You can be skeptical of the hype and still participate in the upside.I'm holding SMH and Intel. I'm watching earnings closely. I'm prepared to be wrong. And I'm definitely not watching any more documentaries about bubbles while my portfolio is open.