Gutenberg Lab — Morning Paper

Meta Earned $16 Billion from Scam Ads While Americans Lost $148 Billion — Platform Profit Structures Enable the Scam Economy

Featuring

Joe Rogan × Mariana van Zeller (investigative journalist, National Geographic, creator of "Trafficked" and "Super Scam Me")

Bottom line

Scam victimization is not about individual stupidity but structural complicity — platforms profit from fraud while victims face shame and double victimization.

3-Line Summary

• National Geographic investigative journalist Mariana van Zeller exposes the $148 billion annual scam economy, tracing it from romance scams to forced labor compounds in Myanmar
• Reuters obtained internal Meta documents showing the company projected $16 billion in 2024 revenue (10% of total profit) from scam ads, charging suspected scammers *higher rates* rather than banning them
• Hundreds of thousands of people kidnapped into Myanmar/Cambodia scam compounds are tortured and forced to scam Americans, with profits funding Chinese criminal gangs and rebel groups in war zones

3 Key Points

1. Platform profit structures amplify scams

Internal Meta documents (obtained by Reuters) show the company estimated users were exposed to 15 billion scam ads per day and projected $16 billion in 2024 revenue from scams and banned goods (10% of total profit). Meta's "penalty bid" system charges suspected scammers *higher rates* if automated confidence is below 95%, but allows them to stay on the platform. If a user clicks a scam ad, Meta's personalized algorithm feeds them *more* scam ads. The company receives 100,000 user reports per week; most scams are not removed.

2. Scam compounds are economic engines for conflict zones

Van Zeller visited a scam city in Myanmar built in months by Chinese criminal gangs — casinos, karaoke bars, go-kart tracks, water parks, all funded by scam profits. Hundreds of thousands of people from Brazil, India, and the Philippines are kidnapped, beaten, tortured, raped, and forced to scam Americans. One victim was a "blood slave" — blood drained 7 times in 6 months and sold. Chinese gangs pay rebel groups to operate in war-torn regions governments cannot control.

3. Sextortion drives teenage suicide; platforms allow repeat offenders

Gavin Guffey, teenage son of South Carolina Congressman Brandon Guffey, killed himself after being sextorted on Facebook. The platform removed the scammer's profile but allowed the *same email address* to create new profiles. Weeks after Gavin's burial, family members were contacted by someone with his photos, demanding money. Congressman Guffey became an advocate for regulation, helping pass the "Take It Down" act under the Trump administration.

Editorial Perspective

This is the opioid crisis playbook. Purdue Pharma profited from prescriptions while 1 million Americans died; Meta profits from scam ads while Americans lose $400 million per day. The conversation must shift from victim-blaming to platform accountability. The Safe Platforms Act (requiring advertiser verification, easier reporting, blocking repeat offenders, and giving victims legal standing to sue) is the regulatory minimum. But as with the opioid crisis, enforcement will lag until the body count becomes politically unbearable. The difference: scam victims don't die visibly in emergency rooms — they die quietly, by suicide, after being shamed by the people who should have protected them.

Source: The Joe Rogan Experience "Mariana van Zeller" Episode 2567 (October 16, 2026)

(unknown)

AI Disclosure: Produced with AI assistance; facts and analysis reviewed by our editorial team.

• --

Why Circle Built ARC — A Regulated Blockchain for the AI Agent Economy

Featuring

Robert Hackett × Jeremy Allaire (Co-founder & CEO, Circle), Chris Dixon (Founder & Managing Partner, a16z Crypto)

Bottom line

Blockchain is in its pre-iPhone moment — 15 years into a trial-and-error journey, just before the right convergence happens.

3-Line Summary

Circle launched ARC, a new blockchain designed as a regulated "economic operating system" for an era when AI agents become the primary economic actors. After a decade building USDC (one of the world's dominant stablecoins), Allaire concluded existing blockchains are too slow, expensive, and unregulated for mainstream adoption. ARC uses known, elected validators (not anonymous ones) and achieves sub-second settlement finality.

3 Key Points

1. Designed for AI agents as primary users

Allaire: "The primary actors that were going to be using these new types of operating systems were increasingly going to be machines. The primary deployers of software, and the primary entities that were going to be running on this was going to very likely be AI." Cognitive labor is economic work, and its coordination will happen on these networks.

2. CCTP as Circle's infrastructure moat

Circle Cross-Chain Transfer Protocol (CCTP) is "the most widely used interoperability protocol in the world right now," handling 50-70% of cross-chain value transfers. Circle mints and redeems $150 billion per quarter. Now CCTP is open to all ARC-issued assets, not just USDC. "We built the highways to move value around all these networks, and the only car that could drive on the highway was USDC until now."

3. Stablecoins are winner-take-most, not winner-take-all

Allaire predicts: "The stablecoin market's sort of a winner take most market structure, not a winner take all market structure. As this goes from several hundred billion to hopefully several trillion, there's going to be multiple networks. We don't assume this is a 90% market."

Editorial Perspective

Allaire's mobile analogy is precise. In the early 2000s, PalmPilot, BlackBerry, and Windows Phone all received massive investment but "all terrible." Convergence happened only when iPhone and Android aligned device, OS, developer incentives, and economic incentives simultaneously in 2007. Blockchain is 15 years into that same journey, and "very hard to say today, oh, this is the iOS of that or the Android of that." Yet while Allaire asserts "no one has a moat," Circle itself already holds stablecoin network effects and CCTP's interoperability infrastructure. This paradox reveals both the market's immaturity and Circle's first-mover advantage within it.

Source: Web3 with a16z, "What Comes After Stablecoins? (ft. Circle CEO Jeremy Allaire and Chris Dixon)" (October 16, 2026)

https://web3-with-a16z.simplecast.com/episodes/after-stablecoins-vHN5EY2_

AI Disclosure: Produced with AI assistance; facts and analysis reviewed by our editorial team.

• --

Healthcare Access Infrastructure for the AI Era: How Zocdoc Solved the 20-Year Provider Directory Problem

Featuring

Julie Yoo (General Partner, a16z; former Kyruus founder) × Oliver Kharraz (Founder & CEO, Zocdoc)

Bottom line

The future of healthcare access isn't owning the front door—it's opening every door the patient knocks on.

3-Line Summary

• Zocdoc solved the decades-old "provider directory problem" (accurate data on which doctors accept which insurance, where they practice, and when they're available) by making data transactional rather than static—every booking validates the information, achieving over 99% accuracy vs. typical insurance directories
• The solution took 20 years because it required solving "literally 1,000+ edge cases" including hundreds of EHR systems, tens of thousands of insurance plans, and real-world workflows like office managers using Post-it notes instead of updating the EHR
• The company is now positioning as "healthcare access infrastructure" rather than defending its marketplace—opening its platform to payers, AI agents (Gemini partnership), and third-party platforms to handle the "explosion of front doors" in the AI era

3 Key Points

1. Transaction-based data beats static directories: Insurance provider directories are "Encyclopedia Britannica"—outdated by publication. Zocdoc built a "Wikipedia model" where actual bookings validate information in real-time, achieving over 99% data accuracy. "You make any mistake exactly once" because each transaction self-corrects the system

2. Best doctors have worst utilization: Counterintuitive Zocdoc finding—top doctors tend to have lowest utilization because they treat patients with busy lives who cancel more often, and they don't double-book. Example: NYC academic medical center dean booked out 3 months still filled approximately 40 last-minute slots via Zocdoc in 30 days

3. National wait times worsening, Zocdoc improving: Average wait time to see a primary care doctor nationally went from 17 days (when Zocdoc started) to 31 days now—nearly doubled. Zocdoc users see doctors within 3 days or less (72 hours) across broad specialties. Mechanism: surfacing last-minute cancellations, no-shows, reschedules. Doctors "in the wild" have approximately 70% load factors (like 1970s airlines); Zocdoc fills the gaps

Editorial Perspective

Oliver's metaphor—healthcare is "a city with multiple doors" not "a building where you trap patients"—captures a fundamental platform shift. The App Store economy succeeded with a single front door; the AI agent era assumes an explosion of entry points. Zocdoc's strategy of opening its infrastructure rather than defending its marketplace is an adaptation to this structural change. The open question: can "opening all doors" coexist with care coordination and quality assurance? The data showing 75-80% of patients rebooking with the same provider suggests openness and continuity can coexist, but whether this holds when AI agents mediate the relationship remains unproven.

Source: a16z Podcast "Why AI Could Finally Fix Healthcare Access | Zocdoc CEO Oliver Kharraz" (October 17, 2026)

https://a16z.simplecast.com/episodes/why-ai-could-finally-fix-healthcare-access-zocdoc-ceo-oliver-kharraz-9n_zG60B

AI Disclosure: Produced with AI assistance; facts and analysis reviewed by our editorial team.