Dan Ives has a message for anyone worried that AI is about to pop like the dot-com bubble: relax. The Wedbush analyst thinks we're in mid-1996, not late 1999. That matters because in his framework, 1996 is when the real money gets made — spending is accelerating, doubts are still thick in the air, and the biggest gains haven't happened yet.

Earnings Season Will Prove the Skeptics Wrong

Ives is expecting a solid fourth-quarter earnings season from Big Tech, and his field checks suggest enterprise AI demand at Microsoft (MSFT), Alphabet (GOOGL), and Amazon (AMZN) remains strong. But here's the thing he thinks Wall Street is missing: an estimated $3 trillion in AI spending over the next three years, with over $550 billion in capital expenditures already in motion.

Why does that matter? Because AI is evolving. We're moving past the infrastructure phase — the chips, the data centers, the plumbing. Now it's about platforms that actually monetize usage, workloads, and execution. That's where the next wave of winners emerges.

Three Companies Built for the Consumption Phase

Snowflake: The Consumption King

Snowflake Inc. (SNOW) is basically the poster child for AI consumption economics. Customers don't pay for software seats — they pay for data processed. More AI workloads mean more consumption revenue. Ives puts it simply: "Enterprise consumption phase is ahead of us into 2026."

The market hasn't caught on yet. Snowflake trades at an enterprise value around 10 to 11 times next-twelve-months revenue, well below peers trading at 15 times or higher. It's being treated like a post-hype laggard, but as enterprises scale their AI projects, Snowflake's model should capture that growth disproportionately.

MongoDB: The Database Behind the Models

MongoDB Inc. (MDB) has become the database of choice for modern AI applications, especially those requiring real-time, flexible data structures for large language models and analytics. Ives sees the "launch of LLM models across the board" as a tailwind.

Wall Street analysts are projecting over 40% long-term revenue growth, driven by expanding annual recurring revenue and higher attachment rates for cloud services. MongoDB's Atlas platform — its cloud subscription engine — already makes up the majority of revenue and is growing faster than legacy segments, signaling deeper adoption.

Palantir: Turning AI Into Action

Palantir Technologies Inc. (PLTR) is one of the few pure plays actually converting AI into operational decisions for enterprise and government clients. Recent multi-year contracts in defense and healthcare — some exceeding $100 million annually — show Palantir's AI platform is moving from proof-of-concept to mission-critical infrastructure.

The company has already guided toward stronger cash flow and margin expansion, which is rare in the AI ecosystem. As Ives notes, "true adoption of generative AI will be a major catalyst for the software sector." Palantir is betting it can be the execution layer for that adoption.

What This Means for Your Portfolio

If Ives is right and this really is 1996, the risk isn't that you're chasing yesterday's winners. It's that you're missing the ecosystem shift as AI spending moves from buildout to consumption. Snowflake, MongoDB, and Palantir aren't playing the hype game anymore — they're positioned for the phase where earnings and actual usage do the talking, not just narratives and promises.