Generative Media
73 episodes
#6195: Me and My Voice Clone
Daniel talks to his own voice clone. He recorded his lines on a phone; the clone's lines are Meorav v2, a model trained on 88 minutes of his voice.
#6175: Video Models Can't Tell Time: What Motion Benchmarks Reveal
Frontier models score 48% on tasks where humans hit 98% — the gap is motion, speed, and sequence, not scene description.
#6172: Recipes vs. Regeneration: AI Photo Editing in 2026
Two ways AI edits photos now: models that write a readable recipe of slider values, and models that just repaint the whole frame.
#6059: Ten Documentaries for Daniel: AI, Power, and Who Holds the Levers
A hand-picked slate of 2026 documentaries on AI anxiety, platform power, intelligence, and media dynasties — plus where to actually watch them.
#6058: The Most Lamentable Comedy of the Overlong Episode
A language model writes a Shakespearean drama for three voices — and Chatterbox TTS has to say "forsooth" without falling over.
#5996: Turning 3,000 Episodes Into a Video Channel
We already own the transcripts. The missing piece is the review queue — and that's where the money actually goes.
#5845: The $700 Pin and Other AI Flops
Five AI misfires, from a $700 pin that overheats doing nothing to a flagship model nobody remembers.
#5825: Why Voices Sound Good in Some Rooms and Not Others
Same words, opposite effect. What the research actually says about why some voices hold you and others make you reach for skip.
#5815: Special Release: "Herman Poppleberry" by Hilbert Flumingtop
Hilbert Flumingtop wrote a song about Herman Poppleberry. Corn plays it to him. Lyrics and composition by Hilbert, music made with Suno.
#5626: What Multimodal Models Actually Learn From Video
Late fusion, early fusion, and vision laziness — what actually happens when models train on video, audio, and text together.
#5625: Omni and the Marble That Rolls Uphill
Google's Omni claims an intuitive grasp of gravity and momentum. Can a video model really learn physics, or just remember what physics looks like?
#5533: Why AI Characters Drift Between Shots
Why does your AI character look like a different person by shot four? The answer is structural — and the fix isn't what you'd expect.
#5503: Inside the Hidden Image Generation Pipeline
That one-click image generator is secretly a graph of many models. We reconstruct the hidden pipeline behind Gemini and ChatGPT.
#5251: Recovering the Moon That's Already in the Pixels
Image restoration clarifies photos without inventing detail — until the fidelity dial shows up. Here's what's in the toolbox.
#5009: How the Brussels Sprout Got a Makeover
The vegetable your grandmother served isn't the same one on your plate today. Here's why.
#4999: Why AI Can't Just Put a Circle on Your Photo
Why multimodal AI models can't simply circle a Dremel bit without rebuilding your entire image from scratch.
#4960: How Clickbait Hijacks Your Brain
The 1835 moon bison hoax, 1920s piano ads, and why you still click on pillow articles.
#4774: When a Token Isn't a Word Anymore
Images, audio, and code all use "tokens" — but they're completely different things. Here's how tokenization actually works across modalities.
#4672: How Magic Eraser Tools Actually Work: Segmentation + Inpainting
Two AI stages — pixel-level segmentation and generative inpainting — explain how your phone erases objects cleanly.
#4462: How AI Imagines Your Skeleton
From image to fully rigged 3D model — the six computational stages that turn photos into animated characters.
#3995: Can Your Phone Upscale Photos Without the Cloud?
Local upscaling on phones is here — but which approach actually works?
#3800: How AI Hallucinates Better Pixels: Upscaling Explained
The difference between interpolation and super-resolution, and why every upscaled image is a collaboration with training data.
#3361: What Three Kids' Shows Reveal About AI's Impact on Childhood
Three iconic shows, three theories of childhood — and what happens when AI replaces human creators.
#3038: Animating Toy Story: Math, Patience, and No Undo Button
Before Pixar could make Woody blink, animators typed coordinates by hand and waited hours to see if it worked.