Episode
737: Monetizable Content
- Podcast
- Grumpy Old Geeks
- Published
- Mar 13, 2026
- Duration seconds
- 3867
- Processing state
not_requested- Canonical source
- https://gog.show/737
Actions
POST https://stenobird.com/v1/public/podcasts/grumpy-old-geeks-168050/episodes/737-monetizable-content/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/grumpy-old-geeks-168050/737-monetizable-content.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
In this week's show we start with FOLLOW UP: The world keeps trying to protect kids online — Indonesia just joined Australia, Spain, and Malaysia in banning social media for under-16s, while COPPA 2.0 sailed through the US Senate unanimously. Meanwhile, Roblox is using AI to clean up its chat, because apparently "Hurry TF up" is the hill they've chosen to die on — even as they're still dealing with the whole "pedophile problem" thing from January. On the AI copyright front, Gracenote is the latest company to sue OpenAI for helping itself to proprietary data, joining a growing queue of plaintiffs who apparently didn't get the memo that everything is training data now. IN THE NEWS: Anthropic is suing the Pentagon after being labeled a "supply chain risk" — apparently because the CEO said AI shouldn't be used for mass surveillance or autonomous weapons, which the Trump administration heard as fighting words. The delicious irony: the Pentagon is still running Claude in active operations while trying to phase it out. Speaking of active operations, investigators now think a missile strike on an Iranian girls' school may have been triggered by bad AI-generated intelligence from that same Claude-based system. So yes, the autocomplete that hallucinates your grocery list is also maybe accidentally bombing schools. Meta's Oversight Board is begging the company to get serious about AI-generated content after a fake war video from a Filipino fake news account racked up 700K views — while separately, Zuckerberg dropped cash on Moltbook, a "social network for AI agents" that turned out to be mostly humans larping as bots and had a security flaw that exposed everyone's API keys. The guy who built it basically vibe-coded the whole thing. Meta's own CTO said he didn't "find it particula…