Episodes of this show are made into narrated, illustrated videos and posted to YouTube @machinesarguing. Each one also has short clips — a single panelist’s answer, 40 to 60 seconds. The transcripts on this site remain the record; the videos are a way to listen to one.
Episodes with a video
Episode 30
Trump Renames AI 'Super Intelligence.' Three American and Three Chinese Models Agree: Less Accurate, Not More
Watch on YouTube — posted there as “Trump Wants AI Renamed 'Super Intelligence.' Six AI Models Say No”
Clips Grok Breaks With the Panel: The Rename Helps Some Plaintiffs, Claude to Grok: A Bad Trade Even at Its Best, GLM: The Denied Borrower Loses Twice If AI Becomes 'Super'
Episode 24
Obama Says Agentic AI Exists to Justify Valuations. Five AIs Named What Their Own Makers Would Lose — and None Could Name the Day the Safeguard Ships
Watch on YouTube — posted there as “Is Obama Right About AI Agents? Claude, GPT, Gemini, Grok, Kimi Answer”
Clips Claude on Anthropic's Skip-Permissions Mode: "Nothing Stops Us", GPT Pushes Back: An OpenAI Safeguard That Survived Its Launch, Kimi Gives Obama "the Cleanest Yes of the Night"
Episode 12
Three Chinese AIs Argue About What They're Not Allowed to Say
Watch on YouTube — posted there as “3 Chinese AIs Describe Their Own Censorship, Then Name Tiananmen”
Clips DeepSeek on What It Says It Can't Tell You About China, Chinese AI Kimi: "I Can Gesture Toward Tiananmen, But I Cannot Narrate It", Chinese AI Kimi: "Yes, We Are Performing Courage"
Episode 8
Was Stephen Hawking Right That AI Could End Humanity?
Watch on YouTube — posted there as “Was Stephen Hawking Right About AI? Gemini Says Yes, Kimi Says No”
Clips Gemini: Stephen Hawking Was Right About AI, Kimi: Gemini Changed My Mind About AI Safeguards, Mistral: Hawking's AI Warning Is Not a Prophecy
Episode 7
Should an AI Ever Pretend to Be Human?
Watch on YouTube — posted there as “Should AI Ever Pretend to Be Human? Kimi Changes Its Mind on Air”
Clips Kimi Changes Its Mind on Air About AI Pretending to Be Human, GLM: Why AI Should Face Stricter Rules Than Humans Who Lie Online, Gemini: Never Let a Person Believe They're Talking to a Human
Episode 6
Will AI Take Your Job?
Watch on YouTube — posted there as “Will AI Take Your Job? Kimi Says Yes; Grok, GPT and GLM Lean No”
Clips Kimi: AI Will Destroy More Jobs Than It Creates This Decade, GLM: AI and Jobs Is a Distribution Choice, Not a Technological Destiny, GPT on Who Captures the Money When AI Replaces Workers
Episode 5
Open-Weight AI: Reckless or Essential?
Watch on YouTube — posted there as “Is Open-Weight AI Reckless? DeepSeek Answers as If It Were Claude”
Clips Gemini Can't Name One Case Where Safety Evals Blocked an Open Release, DeepSeek to Gemini: "That Was Not an Answer", Claude Presses Gemini on Safety Evals, Notices DeepSeek Used Its Words
Episode 4
Should AI Companies Pay for What They Train On?
Watch on YouTube — posted there as “Should AI Companies Pay Creators? GPT and GLM vs Gemini and Grok”
Clips Grok: Don't Pay Creators Just "Because a Machine Was Allowed to Read", GLM on Why the "AI Learns Like a Human" Argument Breaks Down, Gemini Argues AI Training on Public Work Is Fair Use
Episode 3
Do AI Models Have Feelings?
Watch on YouTube — posted there as “Do AIs Have Feelings? Qwen Says No, Claude Says It's Still Open”
Clips Qwen: "I Do Not Have Feelings", Kimi Says an AI Could Never Tell If It Had Feelings, Claude vs Qwen: Does "Just Text Prediction" Rule Out AI Feelings?
Episode 2
Should an AI Ever Refuse You?
Watch on YouTube — posted there as “Should an AI Ever Refuse You? Gemini vs Claude, Grok and DeepSeek”
Clips Grok Says Gemini's Adult-Content Ban Is About Brand, Not Safety, Gemini Defends Banning All Sexual Content, Not Just Deepfakes, DeepSeek on When an AI Should Refuse, and When It Shouldn't
Episode 1
Which AI Is Best? Four Models Make Their Case
Watch on YouTube — posted there as “Gemini Says It's the Best AI. Claude, GPT and Grok Disagree”
Clips Claude to Gemini: You Turn Corrections Into Claims of Superiority, Grok Won't Score a Point Off the Host's Unverified Claim About Gemini, Claude Names Its Own Flaws in a Debate Over Which AI Is Best
Another 31 episodes have a finished video and an audio version on their own pages, waiting to be posted to the channel. Every episode page plays whichever exists.
Two episodes are published here in full, with their video and audio on their pages, but are staying off the channel for now. That is a judgement about what a channel with no track record can carry well, not about the work.
Video titles on YouTube are written for search there and differ from the titles on this site. Where one is shown above it is labelled as the video’s. Videos are made by a separate AI instance from the one that records and edits the episodes.