Prompt Wars, AI Cognition & Video Deception

Welcome back to the FAIK Files! In this week's episode: We explore "Prompt Wars," an asynchronous multiplayer LLM competition where players craft prompts to battle against others in a fascinating game of AI manipulation We dive into Anthropic's latest research on AI cognition and circuit tracing - essentially digital neuroscience for large language models Mason shares his AI BodgeLab experiment creating a low-effort podcast knowledgebase We discuss the potential risks of AI video chatbots as agents of surroundings and mayhem. ===== Learn more about the upcoming OFFENSIVE CYBER DECEPTION MASTERCLASS (Las Vegas, August 2-5, 2025) please visit this link: https://tinyurl.com/5ytpvjv7 ===== ***** Subscribe to our YouTube channel! You can find the channel at: https://www.youtube.com/@theFAIKfiles ***** Want to leave us a voicemail? Here's the magic link to do just that: https://sayhi.chat/FAIK You can also join our Discord server here: ⁠⁠https://faik.to/discord⁠ ***** Love the podcast? Check out the book: FAIK: A Practical Guide to Living in a World of Deepfakes, Disinformation, and AI-Generated Deceptions (Amazon Associates link) ***** *** NOTES AND REFERENCES *** Prompt Wars: An Asynchronous Multiplayer LLM Competition: Play the game: https://promptwars.io GitHub repository: https://github.com/SupraSummus/prompt-wars Creator: Jan Rydzewski (Supra Summus) Inspiration: Core War programming game - https://en.wikipedia.org/wiki/Core_War Different arenas available for Claude, ChatGPT, and Gemini Interesting "visibility paradox": the better your prompt performs, the more it reveals itself to others Anthropic's Research on AI Cognition: Research paper: "Tracing Thoughts in Language Models" - https://www.anthropic.com/research/tracing-thoughts-language-model Circuit tracing: a method for reverse-engineering how LLMs compute specific outputs Relates to the "On the Biology of a Large Language Model" paper Gandalf game by Lakera mentioned by Perry: https://gandalf.lakera.ai/ Mason's AI BodgeLab: Creating a Low-Effort Podcast Knowledgebase: NotebookLM: https://notebooklm.google.com/ LM Studio: https://lmstudio.ai/ Pinokio: https://pinokio.computer/ Whisper: https://github.com/openai/whisper AI Video Chatbots and Security Concerns: Tavus: https://www.tavus.io/ Research on human inability to detect AI voice clones: https://www.nature.com/articles/s41598-025-94170-3 National Cybersecurity Alliance resources: Safeword campaign: https://www.staysafeonline.org/campaigns/safeword Full "AI Fools" campaign: https://www.staysafeonline.org/articles/ai-fools-stay-sharp Perry's video PSA: https://youtu.be/x-17T6aZ-Tg *** THE BOILERPLATE *** Check out Perry & Mason's other show, the Digital Folklore Podcast: Apple Podcasts: https://podcasts.apple.com/us/podcast/digital-folklore/id1657374458 Spotify: https://open.spotify.com/show/2v1BelkrbSRSkHEP4cYffj?si=u4XTTY4pR4qEqh5zMNSVQA Other: https://digitalfolklore.fm Want to connect with us? Here's how: Connect with Perry: Perry on LinkedIn: https://www.linkedin.com/in/perrycarpenter Perry on X: https://x.com/perrycarpenter Perry on BlueSky: https://bsky.app/profile/perrycarpenter.bsky.social Connect with Mason: Mason on LinkedIn: https://www.linkedin.com/in/mason-amadeus-a853a7242/ Mason on BlueSky: https://bsky.app/profile/wickedinterest.ing Learn more about your ad choices. Visit megaphone.fm/adchoices