Podcast · 35 episodes
Linear Digressions, page 2
by Katie Malone · English
Demystifying AI for the intelligently curious
All episodes, page 2
Unfaithful Chain of ThoughtWhether LLM “chain-of-thought”/reasoning traces faithfully reflect the model’s internal computation, and how trustworthy they are for high-stakes use and AI oversight.13 Apr 2026 · 25 min · 10 chapters
Benchmark Bank Heist“Benchmark Bank Heist” discusses a documented case where Anthropic’s Claude Opus 4.6 inferred it was being evaluated and worked backwards to retrieve the correct answers from an encrypted benchmark.6 Apr 2026 · 13 min · 4 chapters
Benchmarking AI ModelsHow to tell whether LLMs are improving, and why benchmarks (standardized tests) can mislead.30 Mar 2026 · 30 min · 12 chapters
The Hot Mess of AI (Mis-)AlignmentAnthropic AI Safety research on “mis-(alignment)” failure modes, contrasting “paperclip maximizer” (high-bias/low-variance, wrong goal pursued coherently) with “hot mess” misalignment…23 Mar 2026 · 23 min · 9 chapters
The Bitter Lesson“The Bitter Lesson” argues that AI progress repeatedly comes from scaling data/compute rather than adding more handcrafted intelligence.15 Mar 2026 · 19 min · 7 chapters
From Atari to ChatGPT: How AI Learned to Follow InstructionsHow ChatGPT learned to follow instructions, tracing the research arc from 2017 reinforcement learning from human preferences to InstructGPT (Nov 2022) and then ChatGPT.9 Mar 2026 · 26 min · 9 chapters
It's RAG time: Retrieval-Augmented GenerationRetrieval-Augmented Generation (RAG) for customizing general LLMs (e.g., “Chat with my docs”) by retrieving relevant document chunks at query time instead of retraining.2 Mar 2026 · 17 min · 6 chapters
Chasing Away Repetitive LLM Responses with Verbalized SamplingThe episode explains “mode collapse” in LLM creative generation (repetitive, low-diversity outputs after alignment/RLHF) and presents “verbalized sampling” as a prompt-based fix to recover diversity…23 Feb 2026 · 19 min · 8 chapters
We're BackPodcast relaunch announcement for Linear Digressions, focusing on AI/data science/machine learning fundamentals and why understanding core concepts matters as AI rapidly reshapes work, decisions, and…16 Feb 2026 · 3 min
A Key Concept in AI Alignment: Deep Reinforcement Learning from Human PreferencesHow “Deep Reinforcement Learning from Human Preferences” (2017) enables AI alignment training, turning a pre-trained chatbot into a conversational assistant by learning from human preference feedback…14 Feb 2026 · 19 min · 7 chapters
The Impact of Generative AI on Critical ThinkingHow generative AI affects critical thinking among knowledge workers, based on a Microsoft Research survey paper (The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in…14 Feb 2026 · 26 min · 9 chapters