Odd Lots Podcast – Episode Summary

Episode Title: Henry Blodget on the Software Selloff Hysteria and the Problem for OpenAI
Hosts: Joe Weisenthal and Tracy Alloway
Guest: Henry Blodgett (CEO, Regenerator; former Business Insider CEO)
Date: March 7, 2026
Location: Live at On Air Fest 2026

Main Theme

This live Odd Lots episode dives into the market's evolving AI narrative, the ongoing tech/software selloff, the precarious business model of OpenAI and other leading AI labs, and the existential questions AI poses for media and employment. Henry Blodgett, with a career straddling both Wall Street and digital media, offers historical context, market analysis, and media industry insight amid the AI "hysteria."

Key Discussion Points and Insights

1. The Flipping AI Market Narrative

Early Euphoria to Present-Day Panic

• The AI market’s sentiment has rapidly shifted from exuberant over high-flying AI company valuations to deep worries about AI’s potential to disrupt or destroy the software industry.

• Henry Blodgett: “This looks a lot like the Internet in the 1990s. … One day, it's catastrophe. None of us are ever going to have a job again. … Other days, hey, it's just a fancy word processor.” (05:25)

• The OpenAI valuation, once seen as excessive at $300 billion, is now $800 billion; parallels are drawn with 1990s dot-com optimism and hype cycles.

Memorable Quote:

“One of the first sentences of [the Citrini] report says 10% unemployment, which means 90% employment… not Armageddon.”
—Henry Blodgett [07:14]

2. The Citrini Substack Contagion

Newsletter Doomerism Moves the Market

• A scenario piece by James Van Gielen ("Citrini") about a dystopian, AI-dominated 2028 triggered a steep selloff in software stocks.
• Blodgett and the hosts question the market’s twitchiness—how a single Substack newsletter could jolt stocks.
• Tracy Alloway: “It speaks to what you just said… when we're in this really uncertain time, you can go from sort of euphoria to doomerism very quickly.” (07:14)
• Henry Blodgett: “Everybody is twitchy. Valuations were high... it's people saying, okay, there may be some change here.” (08:48)

Notable Moment

• Joe Weisenthal notes market's recycling fears, asking if AI agents mean "we don't need to buy regular business software anymore."
  [08:11–10:31]

3. Will AI Destroy Human Employment?

Historical Perspective: Tech Disruption and Jobs

• AI stirs fears of mass unemployment, but Blodgett recalls similar anxieties during past tech leaps—agriculture to industry, steam engines, telephone operators:

• “All through that the number of jobs grew. … The economy overall so far in history has gone on to create a lot more jobs.” (10:47)

• Tracy Alloway challenges this with: “With AI, it's hard to see what those new jobs… are going to be. What are we better at than a supercomputer?” (11:58)

• Blodgett responds that so far, fears haven’t materialized: “These job doom predictions started 20 minutes after ChatGPT came out. Where are we?” (12:20)

• Media example: “You guys continue to grow… You didn't just queue up LLMs.” (13:03)

Key Point

• The “stickiness” or lack thereof in AI platforms is discussed, with Blodgett highlighting that switching between leading models (ChatGPT, Claude, Gemini) is simple—unlike previous tech products (ex: Google search dominance).
  (13:45–14:54)

4. OpenAI’s Business Model: Is It Sustainable?

Funding, Cost Structure, and Strategic Risks

• Tracy Alloway: “OpenAI… is still losing money on every customer that's like a power user. So how does that actually translate into a working business model?” (19:47)
• Blodgett: “Bulls would say the cost… is going to plummet. … If they own the whole market, suddenly… they start to become incredibly profitable. We are not there yet.” (20:22)
• Highlights:
  • Unlike tech’s last cycle, today’s capital expenditure budgets are “eye watering”; OpenAI must constantly raise money, whereas Google/Facebook/MSFT fund with operating cash.
  • True-believer founders believe they’re racing to build AGI (“build the digital God”), justifying infinite R&D spending… until capital dries up.

Notable Quote:

“At some point you run out of money. … OpenAI has to raise it in the open market.”
—Henry Blodgett [22:05]

5. Bubbles, Fads, and Private Markets

• Cautionary tales about random companies pivoting to “AI,” echoing crypto and dot-com eras.
  • Example: the karaoke machine company’s “AI pivot” and the Japanese toilet company (23:17–23:38).
• “We are in a euphoric bubble period.” –Blodgett (23:53)
• Private secondary markets now allow more liquidity for late-stage private companies, but ordinary investors face high fees and scant transparency (no financials).
  • “Big, big warning signs going on over there. … It is the market trying to solve the problem that it's much harder to make money in OpenAI at 800 billion… than it is in the seed round.” (25:30)

6. AI and the Future of Media

Newsroom Angst, Fragmentation, and Brands

• Tracy Alloway: “You did your own AI newsroom experiment... did you expect a backlash?" (29:29)
• Blodgett details experimenting with AI “staff”: “They all had different personalities, they had headshots, et cetera. I wrote this up… Not spectacular, but pretty good.” (30:12)
• He recounts how media executives and young journalists are anxious about jobs and AI experiments. (32:27)

Historical Media Context

• Media got vastly more competitive as distribution changed (rise and fall of social, return to direct subscriber relationships)—“the problem... is there's way too much media.” (32:27–35:38)
• In the AI/LLM era, possible binary future:
  • “Big chat platforms become the front page,” or
  • “Distribution and quality become even more important for trusted brands.” (35:38)
• Blodgett takes the optimistic view: trusted brands (Bloomberg, NYT, niche experts) will survive and even thrive.
  • “For the brands that make it, it is terrific. I trust Bloomberg. … Occasionally there's a mistake, but it is an honest mistake that will be fixed quickly.” (36:28–37:16)

7. How Newsrooms Should (and Shouldn’t) Use AI

• AI is “enhancement” for research, drafting, summarizing—freeing up journalists for judgment, interviewing, and storytelling.
• “We have conflated reporting and writing as journalism… There are opportunities now where the writing lift may be lightened by [AI].” (38:30)
• But investigative skills, judgment, and differentiation remain critical.
  • “It's very hard to protect news… But the organizations that produce it all the time are going to have a huge advantage.” (41:10)
• Niche and expert brands will survive, as will those with devoted audiences.

Quote:

“What is going to stay the same in media is that… we're always going to want to know what's happening and what it means and we're always going to want to be entertained.”
—Henry Blodgett [42:35]

8. The CEO/Business Environment in a Trump Era

• CEOs, Blodgett argues, are pragmatic—not “free speech crusaders”—sticking to company interests, avoiding controversy. (44:27)

9. Reflections on Modern Media, Direct Relationships, and the Value of Experimentation

• If selling Business Insider today, the pitch would be: Be relentlessly adaptive, serve passionate subscribers, and don't reflexively fear AI (46:09).
• Blodgett describes his experiments in subscriber-based media, and even writing fiction, confessing both the joy and futility given AI’s ability to mass-generate content (48:02–49:30).
  • “For me as an analyst and writer… the joy is the process. … I learned so much. … What are humans going to do? … Chess is bigger than it has ever been in person [even though computers beat us].” (49:58)
  • The future is a partnership: “We have AI research and write stuff we don’t really want to do, and we save our writing and orating… for what we’re passionate about.” (50:27)

Timeline of Important Segments

• Market Hysteria and AI Valuations — [05:25]
• Citrini Substack Doomer Report & Market Impact — [06:23–07:14]
• Will AI Destroy Jobs? | Historical Perspective — [10:47–12:20]
• AI Software Replacing Enterprises? — [08:11–10:31]
• AI Platform Switchability vs. 1990s Network Effects — [13:45–14:54]
• OpenAI’s Business Model Debated, Capital Needs — [19:47–22:56]
• Bubble Psychology, AI Fads & Private Markets — [23:53–27:19]
• AI in Newsrooms, Media Competition, and Brand Survival — [29:29–37:54]
• AI as Media Enhancement vs. Replacement — [38:30–42:35]
• Reflections on Subscriber Strategy, Direct Relationships, and Making Mistakes — [46:09–47:42]
• Process of Writing (Humans vs. AI), Chess Analogy — [48:02–50:32]

Notable Quotes & Moments

• “[AI hysteria] is predictions. ... One day it’s catastrophe. ... Other days, hey, it’s just a fancy word processor.” –Henry Blodgett [05:25]
• “So far from Armageddon, it’s people saying, OK, there may be some change here.” –Henry Blodgett [08:48]
• “There really will be no jobs for humans anymore… I think the bigger risk with AI is somebody invents an agentic system that gets out in the world and does a lot of damage.” –Henry Blodgett [10:47]
• “For the brands that make it, it is terrific. I trust Bloomberg. ... occasionally there's a mistake, but it is an honest mistake that will be fixed quickly.” –Henry Blodgett [36:28]
• “We have conflated reporting and writing as journalism… The writing lift may be lightened by [AI].” –Henry Blodgett [38:30]
• “Chess is bigger than it has ever been in person. … My hope is we have AI research and write stuff we don’t want to do, and we save our writing for what we are passionate about.” –Henry Blodgett [49:58–50:32]

Takeaway

This episode frames current AI mania as the latest in a string of historic tech cycles. Despite markets oscillating between euphoria and doom, most incumbents won't be wiped out overnight; big brands and niche experts in media are likely to thrive if they continuously adapt and deepen their relationships with passionate audiences. Meanwhile, AI will change workflows and business models, but as in chess, human drive for meaning and narrative may endure in new, surprising forms.