Intelligent Machines (Audio)

Episode IM 853: All The Clocks Were Wrong – The Rise and Fall of Fact-Checking
Host: Leo Laporte, with Paris Martineau and Jeff Jarvis
Guest: Craig Silverman (disinformation expert, journalist, co-founder of Indicator Media)
Date: January 15, 2026

Episode Overview

This episode explores the fast-changing world of disinformation, fact-checking, and AI's threat and promise. With guest Craig Silverman, a leading authority on fake news and media errors, the panel examines how AI has altered the cost and scale of digital deception and the current state and challenges facing fact-checking. They dive deep into viral AI hoaxes, the economics behind scam ads, the evolution (and co-opting) of "fake news," and practical tools and behaviors for navigating today's fraught information environment.

Key Topics and Insights

1. The Evolution of Disinformation

• AI Lowers the Cost Further:
  Craig Silverman cites Renee Diresta:
  • "Social media took the cost of distribution of information to zero and then AI taking the cost of generation to zero." (04:46)
  • Deceptive content—images, video, text—can now be generated in minutes, no expertise needed, and rapidly distributed.
• Weaponized Spam:
  • Disinformation is now "weaponized spam" with huge reach at little cost.
  • Modern actors include scammers, state-sponsored campaigns, pranksters, and industrial-scale operations (06:16).

2. Case Study: The AI-Powered Food Delivery Hoax

• Reddit Scandal:
  Paris describes a Reddit post claiming to expose evil antics at a food delivery company, complete with a fake Uber Eats ID and forged internal docs, all quickly generated by AI (07:28).
• Detection and Metadata:
  • Google and others embed metadata/watermarks (like SynthID), but most platforms inconsistently read or display it.
  • Detecting fakes remains labor-intensive—“there usually isn’t a really fast way to know instantly.” (08:47)

3. Fact-Checking: Rise & Fall

• Fact-Checking's Growth and Limitations:
  • Fact-checking was once small, focused on public figures; post-2016, Facebook (Meta) funded it globally.
  • Meta’s dominance damaged sustainability—fact-checkers were seen as platform agents, and their roles became politicized. (30:30)
  • Cliff Clavin Effect: Fact-checkers perceived as "finger-wagging" and spoiling the fun; hard to get traction. (32:54)
• Did Fact-Checking Ever Work?
  • Effective for people with open minds or as political accountability pre-Trump.
  • "Fact-checking is not really good at changing deeply held beliefs." (35:43)
  • Post-2017, the tactics of overwhelming the system—“firehose of falsehood”—rendered traditional approaches ineffective.

4. Platforms & Monetized Deception

• Social Media Profiting from Scams:
  • Meta earns billions from scam ads; little financial motivation to solve scamming.
  • "Scam ads do make them money and as much as it does rip off their users..." (13:43)
  • Human trafficking compounds in Myanmar and elsewhere industrialize online scams.

5. AI & Authentication: What Works?

• Reverse Image Search:
  • Still the first, best tool (Google Lens, TinEye).
• Metadata Reading:
  • AI-generated content might be marked, but most metadata is stripped by social platforms.
• Content Credentials & Hardware:
  • Camera makers like Leica embed trusted credentials, but platforms rarely display or preserve them—"performance … was really, really poor." (21:42)
• OSINT & DIY Fact-Checking:
  • Everyone now needs basic open-source intelligence (OSINT) skills; Indicator Media shares tools and guides for this.

6. "Fake News": Origin and Weaponization

• Craig’s Role:
  • Silverman popularized the term “fake news” to describe for-profit, intentionally false viral content (2014-2016).
  • Trump co-opted the term to mean “news I don’t like,” which perverted its meaning. (24:15)
• On Trump and Disinformation:
  • Overwhelming the information environment is effective: "He realized I can brute force this system. They can't keep up with me." (38:19)
• Restoring Trust?
  • No easy answer. “You have to value your attention. … You have to be a little less passive.” (40:00)

7. AI, Journalism, and the New Information Economy

• Should AI Be Trained on News?:
  • Jeff: “We as journalists need to look at our ethical obligation to take journalism to where people are.”
  • Craig: Quality information must be included in LLMs or they'll be trained only on garbage, but the payment conversation is complicated and may lead to further politicization. (44:53)

Notable Quotes & Moments

• On Viral Correction Humor:
  "Instead of referring to the queen bee of a colony, they referred to Queen Elizabeth being able to lay up to 2,000 eggs per day."
  —Craig Silverman (03:33)

• On Platform Responsibility:
  "When it comes to the ads, there is a money factor for the platforms."
  —Craig Silverman (13:43)

• On Fact-Checking’s Image Problem:
  “You seem like a finger-wagging, no fun, Cliff Clavin at the party.”
  —Craig Silverman (32:54)

• On Disinformation Campaigns:
  "They have the ability to just iterate really, really rapidly with the images, the text, the video to find out what’s going to actually work and hit with people. And that’s kind of unfair as an advantage."
  —Craig Silverman (12:15)

• Fact-Checking’s Core Limitation:
  "Fact-checking is not really good at changing deeply held beliefs. It’s better as an accountability measure."
  —Craig Silverman (35:43)

• On Public Responsibility:
  “You have some awareness, you guard your attention, you find a way to be patient and not quick to act. … That’s a really helpful thing.”
  —Craig Silverman (40:00)

• On "Fake News":
  "I usually say I sort of popularized it and then Trump kind of took it, you know..."
  —Craig Silverman (24:15)

Timestamps for Key Segments

• Opening & Guest Intro: 00:00–03:10
• Journalistic Corrections as a Window Into Truth: 03:10–04:30
• AI Drives Cost of Deception to Zero: 04:46–06:16
• The Weaponization of Viral Scams: 06:16–07:18
• AI Hoxes: The Uber Eats ID Example: 07:28–10:30
• AI Metadata & Detection Difficulties: 10:30–12:13
• Manipulation and Human Vulnerabilities: 12:13–13:25
• Platforms, Fraud, and Profits: 13:25–15:18
• DIY Fact-Checking & Tools: 16:25–21:04
• Content Authentication & Platform Gaps: 21:04–22:50
• The Origin and Perverted Use of "Fake News": 24:15–26:39
• The Fall of Fact-Checking – Meta as Poisoned Well: 30:30–34:30
• Cliff Clavin Effect and Political Wariness: 32:54–35:27
• Fact-Checking: Limits and Utility: 35:27–38:13
• Trump and the Firehose of Falsehood: 38:13–39:46
• How to Maintain Trust/Faith in Facts: 39:54–44:08
• AI and Journalistic Obligations to Information: 44:08–47:03

Recommended Actions and Behavior

• Everyone is a Fact-Checker:
  Increased personal responsibility is required in today’s open, over-abundant, easily gamed information landscape.
• Tools to Use:
  • Reverse Image Search (Google Lens, TinEye)
  • Basic metadata readers for photos/videos
  • Cross-check dates, locations, and physical details in images
• Indicators for Reliability:
  Don’t act quickly; question stories that fit your preconceptions; seek professional, vetted, diverse sources, and support responsible journalism.

Resources Mentioned

• Indicator Media:
  Craig Silverman’s media watch site with OSINT tools, fact-checking guides, and reporting on digital deception. https://indicatormedia.com
• Disinformation Handbook (compiled by Craig’s team)
• SynthID (Google’s watermarking tech)
• OSINT workshops (Indicator subscribers)

Memorable Moments

• Rosie the Cat’s on-air Appearance: 06:36–07:10
• "Too Good to Check": The journalistic temptation to believe appealing stories at face value, especially dangerous in the AI/viral age (11:58)

Tone & Style Highlights

• The hosts are witty, self-deprecating, and willing to laugh at their own profession’s foibles.
• Humor and real-world analogies (barbershop, red string, Cliff Clavin) make dense topics relatable.
• The discussion is lively, occasionally sardonic, but always anchored in clear-eyed skepticism and a belief in public service.

TLDR Takeaway

AI has driven the cost of digital deception to zero, overwhelming both individuals and platforms, while traditional and even professional fact-checking now struggles for effectiveness and existence. Everyone must become their own fact-checker, using basic tools, OSINT principles, and above all, a healthy skepticism—which may be the last, best defense in an age when all the clocks are wrong.