The Last Time Investors Really Got Excited For Tech Infrastructure

Episode Summary

Executive Summary: The transcript combines promo segments for Bloomberg’s Stock Movers and Levittown with a long Odd Lots discussion on the late-1990s telecom/internet bubble and its parallels to today’s AI infrastructure boom. Blair Levin explains how deregulation, digitization, and investor enthusiasm drove massive fiber and wireless buildouts, why the boom eventually cracked, and why infrastructure can outlast the bubble even when individual firms fail.

Main Topics: Bloomberg podcast promotions (Priority: 2/5): The transcript opens and later interrupts with promotions for Stock Movers and Levittown, positioning Bloomberg’s short-form market updates and investigative deepfake series for podcast listeners. Late-1990s telecom bubble as an AI analog (Priority: 5/5): The hosts frame the telecom buildout as the closest historical parallel to current AI infrastructure spending, emphasizing costly capex, network effects, and bubble dynamics. Telecommunications Act of 1996 and deregulation (Priority: 5/5): Blair Levin argues the Act helped convert protected analog systems into a digital, competitive environment by empowering the FCC and lowering barriers to competition. Infrastructure investment and debt financing (Priority: 5/5): The conversation explains why investors rushed into telecom networks, how expectations of explosive traffic growth justified heavy borrowing, and why those assumptions proved too aggressive. Signs of bubble fatigue and collapse (Priority: 4/5): The discussion identifies the AOL-Time Warner merger as a symbolic turning point when the market began recognizing that broadband had made some internet-era assumptions obsolete. Who captures value in technology booms (Priority: 4/5): The speakers debate which parts of a boom generate durable returns—platforms, infrastructure, or tools—drawing a comparison to today’s AI stack and cloud providers.

Key Arguments: The telecom era was a 'dual bubble': one in internet applications and another in telecommunications infrastructure. Deregulation mattered because it pushed the industry toward digital competition, which lowered costs and improved communications over time. Wireless number portability and access-charge reform were crucial but underappreciated policy changes that enabled competition. Investors misread traffic growth, assuming data demand would double every quarter rather than every year, which led to overbuilding. Infrastructure networks were not worthless after the crash; many survived and later enabled major winners like Google and Facebook to buy capacity cheaply. The AOL-Time Warner merger marked a key shift in sentiment because it revealed that broadband would undermine AOL’s dial-up model. AI may follow a similar pattern: even if some companies fail, data-center and cloud infrastructure could retain long-term value. The biggest returns may go to a few platform winners and the established cloud companies regardless of the broader bubble outcome.

Data Points: Stock Movers length: five minutes or less - Bloomberg’s new short-form market audio report format Levittown subject: dozens of young women - The podcast series centers on victims of deepfake abuse in a New York suburb Levittown setting history: five years ago - The area was described as a checkerboard of potato farms on Long Island Levittown origin: America’s first suburb - Used to describe Levittown, New York Wireless users pre-Act: less than 10 million - Blair Levin describing the wireless market in the early 1990s Access charge example: five cents a minute - What Bell companies sought to charge AOL for dial-up internet traffic Email pressure campaign: 400,000 emails - Steve Case’s response reportedly sent to Congress within 48 hours Traffic growth claim: doubling every quarter - The market expectation cited during the telecom boom Traffic growth reality: doubling every year - Levin says this was the actual rate, and the difference mattered greatly FCC heating bill spike: $400,000 extra for one month - Anecdote illustrating the intensity of rulemaking after the 1996 Act World Trade Organization agreement: digital traffic traveled much more cheaply around the world - Referenced as part of the broader policy environment supporting global digital communications

Pivotal Quotes: "There was a bubble relating to internet applications. But there was also a bubble related to telecommunications infrastructure." — Blair Levin: Defines the late-1990s boom as two intertwined speculative waves rather than one "We lost a case at the Court of Appeals... But then the Supreme Court gave the power back to the FCC." — Blair Levin: Explains the legal and regulatory struggle over the FCC’s authority after the 1996 Act "The day after Time Warner made probably the worst deal of all time and bought AOL." — Blair Levin: Identifies the AOL-Time Warner merger as an early warning sign that the telecom/internet narrative was breaking down

Implications: For listeners and investors, the episode suggests that today’s AI surge may be less about whether a boom exists and more about where durable value will settle after the cycle. Even if some firms fail, infrastructure, cloud, and a few platform winners may endure.