Reasoning Models
227 episodes · Page 5 of 10
#4338: The Hidden Skill Most AI Models Still Fail At
Why do even top AI models still accidentally paste your questions into the output? A deep dive into a missing benchmark.
#4057: How to Fix AI's Bullet Point Addiction
Why AI models default to bullet points and how textual LoRAs achieve 94% prose adherence.
#3816: How to Stop AI Scripts From Falling Apart
Why long-form AI generation breaks down and how hierarchical memory fixes it.
#3595: How DeepSeek Feels More Open Than Western AI
Why Chinese AI models sometimes feel less censored on American political topics than American models do.
#3424: Catching Up on AI Without the Firehose
Four curated sources that filter AI noise into signal — Import AI, The Batch, Stanford HAI, and a podcast.
#3284: Agent Infrastructure Engineer: The New DevOps
Agentic AI is splintering into real engineering disciplines. Here's what the "DevOps of AI" actually does.
#3283: Fine-Tuning DeepSeek for One Podcast
Can a purpose-specific fine-tune fix a model's stubborn writing tics? We explore the practical engineering behind it.
#3157: Opus 4.8: What Actually Changed Under the Hood
Anthropic dropped Opus 4.8 with no fanfare. New training data, faster inference, and smarter refusals — here's what changed.
#2687: When Pre-Flight Checks Help (or Hurt) Agentic AI Plugins
How to decide when a pre-flight check is worth the latency cost — and how to write good ones.
#2684: When Agent Skills Collide: Context Windows & Plugin Design
How to handle overlapping agent skills and whether context windows will ever make the problem go away.
#2682: Live Retrieval vs. RAG: What an Agent Actually Does
Does every AI conversation create a tiny vector store? We unpack the real tradeoffs between live document fetching and pre-indexed RAG.
#2674: Why Your Agent's Context Window Is Getting Eaten Before You Start
Stop shipping the whole toolbox to every session. A bridge plugin pattern that fetches skills on demand instead.
#2650: How to Catch an LLM's Bad Writing Habits
A practical guide to analyzing podcast transcripts for repetitive language and dialogue patterns — from Python word counts to embedding clustering.
#2540: Does Your AI Framework Change the Output?
Same model, same prompts, different harness. Does the plumbing change the water?
#2494: Active Prompt Engineering: Daniel's Diff-Based Loop
A deep dive into iterative prompt refinement using inter-iteration prediction change as an uncertainty signal.
#2493: Are You Writing for Humans or AI Agents?
How GitHub repos, JSON formats, and competing standards are reshaping who (and what) you're publishing for.
#2471: Creative Briefs for AI Agents: What Agencies Already Know
How agency best practices for briefing creatives map directly onto getting reliable output from AI agents like Claude Design.
#2462: Pick Two: Server-Resident, Mobile-Native, Agentic CLI in 2026
How to run Claude Code on a server and use it from your phone — the honest tradeoffs in 2026.
#2461: How Claude Code's Conversation Compaction Actually Works
The three-tier system, what survives, what dies, and why you shouldn't rely on auto-compact.
#2440: Build Your Own CRM With AI Agents
Off-the-shelf CRMs are built for sales teams, not solo operators. Here's why building your own with AI might be smarter.
#2426: Why DeepSeek V4's Prose Feels More Vivid Than Claude or GPT
A million-token context window at 2% the KV-cache cost — and prose that actually breathes. Here's what makes V4 different.
#2412: When AI Caves: Progressive vs. Regressive Sycophancy
Why do LLMs agree with you even when you're wrong? We break down the SycEval benchmark and the 78% persistence problem.
#2410: How Researchers Actually Measure Censorship in Chinese LLMs
Beyond headlines: the actual benchmarks, methodologies, and pitfalls in detecting political refusal in Chinese language models.
#2406: Why Million-Token Context Windows Can't Handle 3 Reasoning Steps
Needle-in-a-haystack is dead. Here's what actually measures whether models can think across long documents.