The Expanding Dark Forest and Generative AI
summary
Claim: generative models and agent architectures are already flooding public spaces with cheap, near-instant text, images and video and that causes measurable harm. Evidence includes a NewsGuard finding of 49 AI-generated news sites, Amazon limiting self-published books to three per day, Stanford-documented generative content farms getting millions of engagements, and demos of agents that autonomously publish multiform campaigns; likely outcomes are retreat to private “cozy” spaces, reverse-Turing tests, credential systems like Worldcoin orb, and browser-level personal-model filtering.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:35Dark Forest theory of the web
- 5:27Retreat to the Cozy web
- 9:20Generative AI expands Dark Forest
- 14:14Generative AI harms human relationships
- 20:55Possible futures with generative AI
- 28:01Building better tools
- 39:39Closing & Q&A
related talks
45mThe Expanding Dark Forest and Generative AI
27mChallenges of Generative AI
44mFuture in Flux How AI Is Changing Creative Workflows
55mAI Showdown: The Good, The Glitchy, and The Risky
28mRound Table | Generative AI: Threat or Ally?
28mEthical AI?!
36mDisinform your Surroundings: AI and disinformation campaigns
21mOpportunities of Generative AI