I’m not trying to sound the alarm bells for the sake of clicks, but the patterns I’m seeing between the dot-com bubble and potential AI bubble are too eerily similar to ignore.
If you read tech and financial news and watch podcasts like me, you no doubt are hearing similar sentiments right now.
After the family was asleep the other night I went down a rabbit hole on YouTube trying to learn more about the dot-com bubble. I remember the highlights (Pet.com, Netscape, fortunes lost, etc.) but wanted to dig deeper.
During my search I found a short, super helpful video (worth watching below) that made me sit up in my seat and say, “oh sh*t that sounds familiar!”
Massive amounts of risk capital. FOMO. Low interest rates. Rookie entrepreneurs everywhere. It all sounds like the current day AI race.
Of course if you’re an eternal optimist “this time is different.”
But what if it isn’t?
What if this time it’s exactly the same thing 25 years later?
Dot-com vs AI Era
Markets have a way to humble careless investors and those who ignore history, so I decided to dig further and create a comparison table to identify similarities.
Dimension | Dot-com era (1995-2000) | AI era (2023-2026) |
VC investment | ~$240-250B nominal (~$450-500B in 2025 dollars); 80%+ of VC into internet cos at peak | ~$270-380B already; AI took 50-61% of all global VC in 2025; on pace to exceed dot-com totals in half the time |
Gateway product | Mosaic (1993) / Netscape Navigator (1994) | ChatGPT (late 2022), widely called “the new Netscape” |
Picks-and-shovels leader | Cisco (routers, switches); peaked at 130-200x forward earnings | Nvidia (GPUs); ~35-55x, with far stronger profitability and cash flow |
Iconic hype-defining company | Netscape (also AOL, Yahoo) | OpenAI / ChatGPT |
Strong later entrant | Rival browsers / platform players | Anthropic |
Poster children for excess | Pets.com, Webvan (high burn, weak unit economics) | Thinly capitalized AI app-layer and “AI-washed” startups; mostly still private |
Nasdaq / index P/E | 100-200x at peak (NDX trailing >100x) | ~25-40x (Nasdaq-100 trailing ~32x) |
Tech sector forward P/E | ~50-58x | ~25-35x |
Shiller CAPE (S&P) | ~44 | ~38-42 |
Dominant stock group | “Four Horsemen”: Microsoft, Intel, Cisco, Dell | “Magnificent 7” |
Interest rates | Rose to 6.5% peak (May 2000); slashed to 1% by 2003 after the crash | Rose to 5.25-5.50% (2023); easing to mid-3% range by mid-2026 |
Survival rate | ~50% of VC-backed cos alive after 5 years; under half of ~900 IPOs survived independently by 2004 | TBD; shakeout expected at the app layer even if the tech wins |
Infrastructure overbuild | Fiber optic networks | Data centers, power, grid capacity |
Circular financing | Telecom vendors financing customers | Nvidia investing in customers who buy its chips |
Top 10 stocks as % of S&P 500 | ~27% | ~36-40% |
IPO activity | Hundreds of speculative IPOs in 1999-2000 | Few, highly selective offerings so far |
The comparison was shocking to me. Early capital, infrastructure, controversy, massive losses, etc. In some cases we look hotter and scarier now than 25 years ago.
In Summary
Honestly, I want all startups to succeed. The world is better when entrepreneurs and investors win. But the size and concentration of AI right now is starting to look familiar.