Community discussions
Every time AI stocks surge, the same tired chorus starts again: “This feels like the dot-com bubble.” It’s the laziest take on Wall Street — and it ignores every number that matters.
The late-1990s internet boom was built on speculation. Today’s AI economy is built on profitability, scale, and indispensable infrastructure. The difference isn’t subtle — it’s financial gravity.
The Dot-Com Era: Revenue Optional
At the turn of the millennium, companies with no earnings and no viable business models were being valued in the billions. Pets.com, Webvan, and dozens like them burned cash in pursuit of “future users.” Most never shipped a product.
Less than ten percent of households had broadband. Internet advertising barely existed. The Nasdaq wasn’t priced for innovation — it was priced for imagination. When the illusion broke, there was nothing underneath.
The AI Era: Real Businesses, Real Cash Flow
Now look at 2025.
NVIDIA is reporting $28 billion in quarterly revenue with gross margins above 50%.
Microsoft’s AI-enhanced products are adding billions per quarter in new recurring revenue.
Amazon, Google, and Meta are driving efficiency and advertising profits directly from AI integration.
Even infrastructure players like Super Micro, Palantir, and SoundHound AI are seeing triple-digit revenue growth.
This isn’t a promise of profit — it’s profit realized. AI has already become a line-item on corporate income statements, not a bullet point in investor decks.
The Overlooked Giant: Micron and the Memory That Powers AI
There’s one piece of the puzzle that most analysts still overlook: memory.
AI models don’t just need compute — they need massive, high-bandwidth data flow. Without memory, every GPU in the world is just a dead chip.
That’s why Micron Technology (MU) sits at the center of this revolution. Its HBM3E memory powers NVIDIA’s most advanced chips and the next generation of AI accelerators. Micron isn’t a side player; it’s the oxygen supply for the entire AI ecosystem.
As AI models grow larger, memory density and speed become the true bottlenecks — and Micron owns that lane. In fact, its technology is arguably as valuable as NVIDIA’s GPUs themselves.
Profits Don’t Lie
In 2000, the top ten tech firms generated about $40 billion in profit combined. In 2025, the AI-exposed leaders are on pace for more than $400 billion.
That’s not a bubble — that’s a transformation in productivity and profit generation.
Productivity, Not Promises
The internet was a concept in 1999. AI is infrastructure in 2025 — deployed, measurable, and indispensable.
It’s not selling domains or traffic; it’s powering code generation, logistics, advertising, healthcare, and national defense. Every major enterprise on the planet is embedding AI to save costs or create new revenue.
The Bottom Line
Calling this the “next dot-com bubble” is like calling electricity a fad after seeing the first lightbulb.
The dot-com era sold ideas. The AI era sells results.
And without Micron’s memory, there is no AI.
This isn’t speculation — it’s the monetization of intelligence itself.