Artificial Intelligence
Episodes whose main subject is AI itself: building with, training, evaluating, or debating AI systems. AI used only as a tool for another topic goes to that topic.
#5188: DeepSeek's Point Release That Isn't
DeepSeek shipped a whole new architecture and called it a point release. Here's what actually changed inside the model.
#5187: Translation Is Not a Pipe: What Machines Miss
Rubrics rate machine translation above human work — until readers pick side-by-side. What translators know that metrics can't see.
#5184: AI Is a Number Factory, Not a Wordsmith
Under the prose, every AI model is just matrices of floating-point math. So where does the randomness actually come from?
#5183: Where AI Actually Sits on the Hype Cycle
AI has been doing real work for decades. So why do vendors still shout it from the rooftops?
#5182: DeepSeek V4.1 Flash: 1M Context, 437x Smaller KV Cache
DeepSeek V4.1 Flash landed with a 1M-token window and a KV cache 437x smaller than V1. Here's what actually changed — and why the middle of your co...
#5175: Vanilla Sky: The Movie That Makes Sense on the Third Watch
A 41% critic score, a D- CinemaScore, and $203M at the box office. How Vanilla Sky became a movie people can't stop rewatching.
#5169: Fine-Tuning vs From-Scratch for Minor Languages
What 774 experiments and a new Armenian model reveal about the tradeoff between fluency and knowledge in low-resource languages.
#5165: Weighting Memory in a RAG Pipeline
Daniel's memory leg won't fire. Turns out system prompts have no inherent vector-space privilege — but retrieval does have a real knob.
#5154: DeepSeek's Two-Endpoint Philosophy
DeepSeek quietly routed Pro traffic to Flash — and that routing change says everything about its two-endpoint strategy.
#5140: Teletext: The One-Way Internet That Won't Die
Before dial-up, there was teletext — scores hidden in TV signals. Why it felt like the internet, and why the Dutch never turned it off.
#5136: How to Build a City Companion AI
A travel guide GPT that went viral reveals the blueprint for a voice-powered AI that knows your city—and when to speak up.
#5123: Who Gets to Call Themselves an Artist?
Creative rivalry in couples, engraver vs. painter status, and who decides what counts as art.
#5118: Model Identity Fraud: Distillation or Data Contamination?
When AI models lie about who they are, is it stolen knowledge or just contaminated training data?
#5117: How AI Training Data Gets Filtered (and Exploited)
Six stages of content filtering stand between raw web crawls and your AI model — here's where poisoning attacks slip through.
#5116: Beyond Common Crawl: The Datasets Shaping AI
Common Crawl is just the bulk flour. We explore the curated datasets that actually shape what AI models know and do.
#5113: Can Rival Labs Poison AI Training Data?
A few hundred crafted documents can shift a model's stance. Here's how data poisoning actually works.
#5108: How to Identify Anything in the Sky: A Jerusalem Drone Case Study
A reader saw a green strobe hovering over Jerusalem. Here's how to think like an educated guesser.
#5089: How Runway Overruns Happen and the EMAS Safety Net
A Prime Air 767 overran at Miami. We break down the physics, the safety engineering, and why EMAS beds matter.
#5088: What Does a Pilot's Gray Hair Actually Buy You?
Is flight seniority about skill or something harder to measure? We dig into what experience really buys in the cockpit.
#5056: How AI Writes a 30-Minute Podcast in One Pass
What does it take for a model to write a full episode without edits? The answer involves context windows, repetition, and a surprising architecture...