Better Offline – "Working In The Dot Com Bubble" ft. Matt Rosoff
Podcast: Better Offline (Cool Zone Media & iHeartPodcasts)
Host: Ed Zitron
Guest: Matt Rosoff, Editor-in-Chief of The Register
Date: January 26, 2026
Episode Overview
In this incisive episode, Ed Zitron sits down with Matt Rosoff—a Silicon Valley veteran and media editor who’s witnessed the rise and fall of tech cycles since the early 1990s. Together, they dissect the atmosphere of the original dot com boom, what it was like to work in the industry during that era, and draw sharp parallels (and contrasts) with today's speculative AI bubble. The discussion blends first-person anecdotes with sharp industry critique, tackling topics such as the intoxicating optimism of 90s tech, retail investor risk, and today's echoing cycle of tech hype. If you've ever wondered what it actually felt like inside the bubble—or whether the new AI gold rush is just another mirage—this episode is for you.
Key Discussion Points & Insights
1. Life Inside the Dot Com Bubble ([03:54]–[07:16])
- Matt’s Beginnings in Tech:
- Moved to San Francisco in 1992 during a post-recession lull, started with internships and odd jobs.
- Joined CNET in 1995 as employee #42, back when "content" was called an "online magazine."
- Helped launch "Digital Dispatch," one of the first email newsletters—combining web updates with “occasional snarky commentary.”
- Early tech skepticism: doubted MP3 players ("can only hold 12 songs, complete waste of time"), reflecting the era’s uncertainty about what would last.
"I think when I left [CNET] there were over 500 people working there...it was fun being at a company growing fast." —Matt Rosoff [05:03]
2. Wild Ideas, Startups, and Mass Hysteria ([07:16]–[14:06])
- Startup Mania:
- Matt briefly worked for "the most incompetently run company" (online education startup: Catio), which had “no way to put the content on the Internet” and no basic digital infrastructure.
- Venture capital was flowing to companies on the basis of PowerPoint decks and hype, not real products.
"They didn't even know what a CMS was...We were using Visual Source Safe, which is a Microsoft product for code." —Matt Rosoff [09:14]
- First Signs of Trouble:
- Noticed companies like pets.com, Cosmo.com, and lackey.com ("it was kind of like TaskRabbit") burning cash on unsustainable business models.
"It was obvious that a lot of the specific individual companies...were completely hot air...they were selling dollars for 75 cents." —Matt Rosoff [10:29]
3. The Bubble Bursts: Early Warnings and Fallout ([14:39]–[19:55])
- Moments of Realization:
- Awareness hit when startups had funding but nothing to show for it; employees were "play acting."
- Company excesses: CNET's extravagant holiday parties, realizing post-IPO that “we’re losing money.”
"I didn't know you could stay in business and lose money. How naive I was." —Matt Rosoff [16:23]
- The Broader Landscape:
- Infrastructure companies (like Lucent, WorldCom) collapsed as circular, “deal-for-deal’s sake” strategies unraveled.
- Bubble felt real: web was universally exciting and transformative, even as the economics cracked.
4. Dot Com vs. AI Bubble: Parallels and Contrasts ([17:21]–[31:38])
- Skepticism Over AI Hype:
- Matt is unimpressed by AI’s claimed “revolutionary” status, calls it more “autocomplete” than intelligence.
- Compares immediate, obvious impact of the Internet/iPhone to the more opaque value proposition of today’s LLM-based AI.
"It was a lot more immediately obviously huge...When I look at AI, to me it doesn't seem nearly as revolutionary...it doesn't feel immediately massively obvious that this is a huge change." —Matt Rosoff [17:21]
- Cultural Context:
- The dot com boom centered around public participation; everyone was gung-ho, and large-scale retail investment fueled the mania.
- Today, AI’s speculative surge is driven more by venture capital and tech giants, with limited direct retail exposure.
5. Big Tech, Retail Investors & Consequences of Collapse ([31:38]–[47:36])
- Aftermath of the Dot Com Crash:
- “People in tech weren’t permanently hurt” —many bounced back as the industry kept evolving.
- Non-tech retail investors were the biggest losers, often wiped out after putting savings into risky IPOs.
"My attitude toward investing has always been caveat emptor. You shouldn't be betting if you can't afford it." —Matt Rosoff [34:42]
- Cautions for the AI Bubble:
- AI risks are different: "mostly really large tech companies" could weather a drawdown, but massive credit deals and VC inflows are new.
- Potential contagion due to concentration: “Nvidia has customers that make up 15–20% of their revenue each quarter...half to 80% of their revenue is going to disappear [if the bubble pops].” — Ed Zitron [47:36]
6. Tech Fads, Real Value, and End-of-Era Feelings ([49:38]–[56:43])
- The Nature of Bubbles and Tech Cycles:
- No historical parallel to the unique mixture of hype, vaporware, and scale seen with OpenAI and the current wave.
- AI's "killer apps" in everyday life remain elusive, especially compared to the Internet’s instant utility.
- Teenagers (i.e., the “future demographic”) are mostly indifferent or openly mock AI outputs.
"If you've got teenagers who are like making fun of calling bad stuff AI—I'm actually serious, that is legitimately lethal." —Ed Zitron [56:43]
"Young people love Tiktok...They're scrolling vertical video all day." —Matt Rosoff [57:15]
7. AI Psychosis & Executive FOMO ([58:14]–[62:05])
- Executives Driven by Hype:
- Ed posits that “AI psychosis” affects tech leaders, leading to mass overinvestment based on overestimating AI potential.
- Matt links this to "anti-expert" social trends and a desire to feel ahead of the curve.
"That's kind of what I'm getting at...using a large language model convinces some people that they are in the future and creates this kind of religious feeling around it." —Ed Zitron [60:09]
- AI as Continuation, Not Revolution:
- AI is seen as an extension of past tech cycles, not a bold new epoch.
"AI is just kind of this, 'hey, let's keep this playbook running.'" —Matt Rosoff [55:31]
Notable Quotes
-
On Early Tech Reporting:
"My very first task there was to come up with an email newsletter which we called Digital Dispatch...occasional snarky commentary."
—Matt Rosoff [03:54] -
On Realizing Company Instability:
"I quickly got out of there...found a job at Directions on Microsoft...they were out of business by, I believe, early 2001."
—Matt Rosoff [10:09] -
On Cultural Memory:
“It was just presumed that this is the future. If you don’t want to be part of the future, that’s fine, but why don’t you go be a survivalist and live off the grid in the woods kind of thing.”
—Matt Rosoff [30:14] -
On the Nature of Hype:
“I don't recall...large scale where everybody in the industry was saying this is going to revolutionize everything and end users are kind of going, huh. I don't know...”
—Matt Rosoff [51:47] -
On Tech Teen Skepticism:
"My son is 15 and he's very cynical...They make fun of phrases and terms that are so AI. And he'll see a poster...he's like, ha, that's so AI."
—Matt Rosoff [56:43]
Key Timestamps
- Matt’s entry into dot com era: [03:54]
- Excess and culture at CNET: [05:03]
- Dot com startup failures: [09:14]
- Early bubble warning signs: [10:29]
- Public investment & retail boom: [27:08]
- AI skepticism vs. dot com optimism: [17:21], [24:39], [33:28]
- Tech industry resilience post-crash: [37:17]
- Comparison to AI investing scale: [46:04]
- Teen & user sentiment today: [56:43]
- Hype, psychosis, and executive FOMO: [58:14], [60:09]
- Broader cultural/tech era shifts: [54:12]
Memorable Moments
- Ed and Matt riff on the absurdity and fun of dot com-era parties, including playing live music at the Fillmore—contrasted sharply with their retrospective financial naiveté. ([14:45])
- The frank admission of missing the immediacy of the Internet’s impact, versus the present-day murkiness around AI’s actual utility—especially among younger users. ([17:21], [56:43])
- Ed’s theory of “AI psychosis”: that being exposed to large language models tricks even industry leaders into believing hype, accelerating the cycle of overinvestment. ([58:14])
Tone and Language
- The conversation is fast-paced, deeply knowledgeable, and laced with humor and cynicism—especially toward the excesses and myths of both old and new tech cycles.
- Both speakers balance nostalgia for the “go-go” past with clear-eyed skepticism of present-day claims.
Conclusion
This episode offers a rare insider’s look at the dot com era, serving as a cautionary tale for today’s AI-chasing tech world. The hosts dismantle current hype, compare it to lessons from the past, and issue a warning about the social, personal, and financial risks when exuberance outpaces reality.
Summary prepared for Better Offline listeners and tech history enthusiasts seeking to understand the cyclical nature of tech bubbles—and what, if anything, will be different this time.