Better Offline: "How To Argue With An AI Booster, Part Three"

Podcast: Better Offline (Cool Zone Media & iHeartPodcasts)
Host: Ed Zitron
Date: September 12, 2025

Episode Overview

In this final installment of the "How To Argue With An AI Booster" trilogy, tech industry veteran Ed Zitron dismantles the last—and worst—arguments made by AI boosters, especially around the economic reality, business models, technological promises, and societal impact of generative AI. With his trademark mix of irreverent humor, sharp skepticism, and unfiltered language, Ed exposes how many of these arguments don’t hold up under scrutiny, are driven by hype, and are swallowed by a credulous tech and media ecosystem. He challenges both AI evangelists and the journalists covering them to do better, highlighting the need for more honest and rigorous discussion of technology's impact on society.

Key Discussion Points & Insights

1. The Weak Foundations of AI Booster Arguments

• Ed opens with satirical dramatics, poking fun at the over-inflation of his own podcast and cutting straight to the real issue: AI booster arguments "crumble under the slightest bit of scrutiny." (03:00)
• Core message: These arguments exploit public ignorance about AI economics or try to overwhelm rational debate with hype and speculation.

2. Comparing AI to Amazon Web Services: A False Analogy

• Booster Claim: "AI is just like AWS—a massive initial investment that everyone doubted, but became necessary and profitable." (04:19, [Kevin Roose])
• Ed’s Critique:
  • AWS was a response to a clear, existing market need and became break-even within three years.
  • OpenAI, founded in 2015, hasn't proven sustainable demand even years after its launch; burning far more capital.
  • “Amazon Web Services took something that people already did and made it better and scaled it. Generative AI doesn’t fit that pattern.” (05:17)
  • Memorable quote: “Amazon Web Services was break even within three years and OpenAI was founded in fucking twenty fifteen...” (04:25)

3. Misleading AI Revenue Metrics

• Booster Tactic: Citing “annualized revenues” to imply greater success.

• Ed’s Take:

  • Annualized revenue is misleading—just monthly revenue times twelve, and ignores churn and actual profit.
  • “It’s a number intentionally used to make a company sound more successful... You heard $200 million annualized, but your mind did that.” (05:49)
  • Always press for profit and burn rate, not vague forward-looking numbers.

• Booster Quip: “AI companies are in growth mode... they’ll pull the profit lever when it’s time.”

• Ed’s Response:

  • “Why have none of them done this? None. Not one. Not one of them.” (06:57)

4. AGI Fantasies and Industry Hype

• Recurring Theme: Obsession with “AGI” (Artificial General Intelligence).

• Ed scorches the rhetoric:

  • “We do not know how thinking works in humans and thus cannot extrapolate it to a machine... anyone claiming they do is lying.” (07:11)
  • Meta’s chief AI scientist Yann LeCun: AGI isn’t possible with current models.

• Satirical advice: “If someone brings up AGI, ask them about the romantic relationship between Banjo and Kazooie. It’s all fan fiction.” (08:05)

• “Secret” Powerful Models:

  • “I’m hearing from people deep within the AI industry...” (08:39, [Kevin Roose])
  • Ed: This is just playground gossip—if something was real, people would show specifics.
  • “If they say ‘I couldn’t possibly tell you,’ then they’re full of shit... full of crap... full of doo doo.” (08:44)

5. ChatGPT’s Popularity as Proof of Value

• Booster Line: “ChatGPT is hugely popular—700 million weekly active users! Its popularity proves its utility.” (09:36, [Kevin Roose])

• Ed’s Dissection:

  • Ed questions what “active” means—could be as little as visiting once a week.
  • Points out that a massive PR blitz (constant news, features, and media mentions) manufactured this popularity.
  • On OpenAI’s metrics: “Even OpenAI hasn’t clearly defined what an ‘active user’ is, and their conversion rate is mediocre at best.” (19:35)

• Business Customers:

  • Ed recounts cheap trial offers and questionable accounting on business customers—“suspicious” revenue definitions and aggressive user counting.
  • Conversion rates appear poor (“less than 3 percent”), especially for a supposedly world-changing product.

6. OpenAI Revenue Claims vs. Reality

• Booster Claim: “OpenAI is making tons of money—that proves they’re successful and you’re wrong.” (25:01, [Kevin Roose])
• Ed’s Take:
  • The revenue is real, but transformative enterprise adoption is not.
  • No clear, documented business wins; CEO boasts aren’t backed by plain-English results.
  • “If there were [real wins], we’d have one article pointing to a ChatGPT integration that scaled a company or saved tons of money—written in plain English.” (25:04)
  • AI product features and business models lack meaningful differentiation from competitors.

7. The Media’s Role in AI Hype

• Ed lambasts journalists for failing to report accurately or challenge overblown claims, e.g., parroting that GPT-4 completed agentic tasks (like hiring a TaskRabbit):

  • He traces a widely reported claim—GPT-4 hiring a human to solve a CAPTCHA—back to its dubious origins and finds the original researchers manually prompting the model, not true autonomy. (13:34-16:58)
  • Quote: “It is transparently, blatantly obvious that GPT-4 did not hire a TaskRabbit or indeed make any actions it was prompted to.” (16:58)

• On the GPT-4 “reasoning” myth:

  • “OpenAI should have gone out of their way to correct people. Instead they sat back and let people publish outright misinformation.” (16:58)

• Journalists’ Responsibility:

  • Media’s constant repetition and amplification of AI marketing claims, without technical scrutiny, amplifies the hype and perpetuates confusion.
  • “Media pressure, societal pressure, and pressure from bosses pushes hundreds of millions of people to try a product that even its creators can’t really describe.” (19:35)

8. Debunking AI Displacement, “Vibe Coding,” and Research Hype

Layoffs and Job Displacement:

• Claim: AI is destroying young people’s job prospects.
• Ed: Cites weak evidence—hiring issues in professions like accountancy have unrelated causes. “Anyone making this point is grasping at straws.” (27:00)
• Some lower-wage roles (translation, transcription) have been impacted, but “AI is not replacing people at the knowledge worker scale nor at the coder scale.” (31:00)

Science & Research:

• Claim: AI is revolutionizing scientific research.
• Ed: Most claims are exaggerated or based on “insufficient” results—if not outright retracted (MIT’s materials “discovery” paper). (30:30)
• Foundation models may help, but are not transformational as claimed.

“Vibe Coding” and AI for Non-Programmers:

• Claim: AI is letting anyone build software with plain English.
• Ed:
  • “AI can write code, sure, but that doesn’t make it good code.”
  • No actual, functioning, secure, successful company was ever built by someone “vibe coding” with AI.
  • “Vibe coding is a marketing term based on lies.” (34:15)
  • AI coding environments can make engineers slower ("recent study suggested that they actually make software engineers 19% slower"). (36:00)

9. Ed’s Closing Reflections on AI Skepticism and Media Failure

• The burden of proof unfairly falls on skeptics, who are constantly scrutinized, while boosters are celebrated.
• “Skeptic and critic are words said with a sneer... AI boosters are the first to get TV appearances and panel offers.” (41:00)
• On Sam Altman and media sycophancy:
  • “Sam Altman does not respect you. He is not your friend. Clammy Sam Altman thinks you are stupid and easily manipulated.” (43:00)
  • The tech media’s fear of missing out or “seeming stupid” is being weaponized against honest reporting.

Notable Quotes & Memorable Moments

• On AGI Hype:

  • “Anyone claiming they [know how to get to AGI] is lying. Anyone who’s talking about AGI is talking about fan fiction...” (07:11)
  • “Next time someone brings up AGI, bring up Banjo and Kazooie and their romantic involvement...” (08:05)

• On AI “Secret Models” Myth:

  • “If they say, ‘I couldn’t possibly tell you,’ then they’re full of shit. They are full of doo doo.” (08:44)

• On AI's Supposed Utility:

  • “If you use it for research, you do not respect natural research—you want a quick answer, it’s that simple.” (33:07)
  • “I have yet to hear one use case that truly impresses me or even one thing that feels possible now that wasn’t possible before.” (33:18)

• On the Media’s Role:

  • “This has gone beyond simple objectivity into the realm of an outright failure of journalism. I have never seen more misinformation about the capabilities of a product in my entire career.” (40:05)

• On Personal Motivation:

  • “I didn’t become a hater because I’m a contrarian. I became a hater because the shit that these fucking oafs have done to the computer pisses me off.” (44:30)

Key Segment Timestamps

• AI Booster Arguments Satire/Opening: 03:00
• AWS vs. AI Business Models: 04:19–05:47
• Annualized Revenue Metrics Critique: 05:49–06:55
• AGI and Industry Hype: 07:07–08:39
• Popular AI Myths ("Secret Models"): 08:39–09:36
• ChatGPT Popularity Dissected: 09:45–19:35
• TaskRabbit "Agentic AI" Incident/Media Critique: 13:34–16:58
• Media’s AI Hype Role: 19:35–25:00
• OpenAI Revenue Claims: 25:01–27:00
• Job Displacement Narratives Debunked: 27:00–31:00
• “Vibe Coding” Gets Shredded: 34:11–36:29
• Ed's Reflection on Skeptics vs. Boosters: 40:05–44:30

Final Thoughts

Ed Zitron ends the series with a passionate, at times scathing, but ultimately earnest plea both to the public and to the media: Ask more of those making big claims for AI. Skepticism is not cynicism—it's necessary, especially when so much money and cultural attention is at stake. The episode calls out boosterism, marketing sleight-of-hand, and shallow media coverage for their roles in inflating expectations beyond reality, and finishes with Ed reaffirming his honest, curiosity-driven outlook—“my shit fucking bangs... I research it very well... Everything I do comes from a genuine curiosity and an overwhelming frustration with the state of technology." (44:30)

For Further Engagement

• Ed Zitron’s newsletter and Discord (betteroffline.com)
• “Sincerity Wins The War” newsletter (linked in show notes)
• Colton Voge and Nick Suresh’s pieces on AI coding, referenced in-show
• Email: ezetteroffline.com

Episode concludes with gratitude to listeners: “I really do love you all for listening... I do not have it in me to pretend to be anything other than what I am and if that’s strange to you, well I’m a strange man, but at least I’m an honest one.” (44:30)

For a thorough debunking of AI booster myths and a passionate defense of tech industry skepticism, this is a must-listen episode—funny, raw, and meticulously argued.