Every generation thinks they've broken the old rules—right up until the moment the rules break them.
The Summary
- Fast Company revisits the "new era" delusion that preceded the 1907 panic, the dot-com crash, and now the AI boom—each time investors convinced themselves the fundamentals no longer apply.
- W. Brian Arthur's 1996 theory of "increasing returns" justified dot-com valuations by pointing to winner-take-all dynamics, high upfront costs, and network effects—the same arguments today's AI bulls use to explain why companies with no profits deserve billion-dollar valuations.
- The pattern repeats: legitimate innovation creates space for marginal players and outright frauds, amplifying the boom until systemic risk hits critical mass.
- The warning isn't that AI won't matter, it's that belief in exceptionalism makes everyone blind to the mechanics of collapse.
The Signal
The Netscape IPO in 1995 opened at double its pricing and closed the day with a $2.9 billion valuation despite zero profits. That wasn't a bug in market logic. It was a feature of what economist W. Brian Arthur called "increasing returns economics." High upfront investment, near-zero marginal costs, and network effects create winner-take-all markets where the fastest firm captures disproportionate rewards. Venture capitalists internalized this framework and began pouring capital into anything with a domain name.
Fast forward to 2025 and swap "dot-com" for "AI company." The same mechanics are in play. Training foundation models requires massive upfront capital. Inference costs trend toward zero as compute commoditizes. Network effects accumulate around the models developers actually use. The logic holds, which is why smart people keep funding companies burning hundreds of millions per quarter.
"New era thinking doesn't just benefit innovators—it creates space for marginal players and outright hucksters."
But here's where the pattern gets dangerous. Every boom attracts not just legitimate builders but a second wave of opportunists who understand the narrative better than the technology. In the dot-com era, that meant companies going public with nothing but a website and a press release. In crypto's 2021 peak, it meant celebrities launching dog-themed tokens. In AI, it's companies rebranding customer service chatbots as "agentic AI platforms" and raising at half-billion-dollar valuations.
The article traces this pattern back to 1907, when financial journalist Alexander Dana Noyes watched markets operate "on the assumption that we were living in a New Era; that old rules and principles and precedents of finance were obsolete." That assumption preceded the Panic of 1907. It preceded the dot-com crash. It preceded 2008.
Key mechanics of every "new era" collapse:
- Genuine innovation creates a plausible narrative for exceptional returns
- Capital floods in, lowering the bar for what counts as "innovative"
- Marginal players and frauds amplify the boom while concentrating systemic risk
- When the marginal players collapse, contagion spreads to legitimate companies
The Glass-Steagall repeal in 1999 removed the firewall between commercial and investment banking, allowing the dot-com frenzy to bleed into the broader financial system. We're watching a similar regulatory unwinding now. Calls to relax AI safety requirements, fast-track model deployment, and classify AI infrastructure as critical (thus eligible for federal support) all follow the same script.
The uncomfortable truth is that increasing returns economics is real. Network effects are real. Winner-take-all dynamics are real. But those dynamics don't suspend the laws of financial gravity. They just make the crash steeper when reality reasserts itself. The companies that win will be the ones building actual moats, not just burning capital to claim territory in markets that don't exist yet.
The Implication
If you're building in AI, the question isn't whether the boom is real. It's whether your business model survives the moment capital stops flowing to "potential" and starts demanding cash flow. The companies skating by on narrative will evaporate. The ones with actual revenue, customer retention, and unit economics will own the aftermath.
For investors and observers, the lesson is simpler: every boom thinks it's different. None of them are. Watch for the moment when fundraising velocity matters more than profitability, when every pitch deck includes the phrase "paradigm shift," and when skeptics get dismissed as out of touch. That's not the beginning of the new era. That's the top.