El Podcast

E166: Is the Internet Too Big to Moderate? — John Wihbey

Episode Summary

John Wihbey is a professor at Northeastern University who studies how social media and AI shape what we see and talk about online. In this episode, we break down why certain voices get boosted or buried, why platforms struggle with moderation, and how AI is now creating both the problem and the tools to fix it. We also talk about the shaky business of news and content creation today—and what skills young people will need to actually stand out in an AI-heavy world.

Episode Notes

A wide-ranging conversation with Northeastern’s John Wihbey on how algorithms, laws, and business models shape speech online—and what smarter, lighter regulation could look like.

Guest bio: John Wihbey is a professor of media & technology at Northeastern University and director of the AI Media Strategies Lab. Author of Governing Babel (MIT Press). He has advised foundations, governments, and tech firms (incl. pre-X Twitter) and consulted for the U.S. Navy.

Topics discussed:

• Section 230’s 1996 logic vs. the algorithmic era
• EU DSA, Brazil/India, authoritarian models
• AI vs. AI moderation (deepfakes, scams, NCII)
• Hate/abuse, doxxing, and speech “crowd-out”
• Platform opacity; case for transparency/data access
• Creator-economy economics; downranking/shadow bans
• Dead Internet Theory, bots, engagement gaming
• Sports, betting, and integrity (NBA/NFL)
• Gen Z jobs; becoming AI-literate change agents
• Teaching with AI: simulations, human-in-loop assessment

Main points & takeaways:

• Keep Section 230 but add obligations (transparency, appeals, researcher access).
• Europe’s DSA has exportable principles, adapted to U.S. free-speech norms.
• States lead on deepfake/NCII and youth-harm laws.
• AI offense currently ahead; detection/provenance + humans will narrow the gap.
• Lawful hate/abuse can practically silence others’ participation.
• CSAM detection is harder with synthetics; needs better tooling/cooperation.
• News/creator models are fragile; ad dollars shifted to platforms.
• Opaque ranking punishes small creators; clearer recourse is needed.
• Engagement metrics are Goodharted; bots inflate signals.
• Live sports thrive on synchronization; gambling risks long-term integrity.
• Students should aim to be the person who uses AI well, not fear AI.

Top 3 quotes:

• “Keep 230, but add transparency and obligations—we don’t need censorship; we need visibility into how platforms actually govern speech.”
• “AI versus AI is the new reality—offense is ahead today, but defense will catch up with detection, provenance, and human oversight.”
• “The platform is king—monetization and discoverability are controlled by opaque algorithms, and that unpredictability crushes small creators.”