# Episode 67: Saving Hundreds of Hours of Dev Time with AI Agents That Learn

Page: https://stenobird.com/podcast/vanishing-gradients-4989163/episode-67-saving-hundreds-of-hours-of-dev-time-with-ai-agents-that-learn
Text version: https://stenobird.com/podcast/vanishing-gradients-4989163/episode-67-saving-hundreds-of-hours-of-dev-time-with-ai-agents-that-learn.md
Podcast: [Vanishing Gradients](https://stenobird.com/podcast/vanishing-gradients-4989163)
Published: 2026-01-14T02:33:22+00:00
Episode link: https://hugobowne.substack.com/p/episode-67-saving-hundreds-of-hours
Audio file: https://api.substack.com/feed/podcast/184492847/ae462eeb219622651b3a76862c119919.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/vanishing-gradients-4989163/episodes/episode-67-saving-hundreds-of-hours-of-dev-time-with-ai-agents-that-learn
Duration seconds: 4702

## Resource

This is continual learning , right? Everyone has been talking about continual learning as the next challenge in AI. Actually, it’s solved . Just tell it to keep some notes somewhere. Sure, it’s not, it’s not machine learning, but in some ways it is because when it will load this text file again, it will influence what it does … And it works so well: it’s easy to understand . It’s easy to inspect, it’s easy to evolve and modify ! Eleanor Berger and Isaac Flaath , the minds behind Elite AI Assisted Coding , join Hugo to talk about how to redefine software development through effective AI-assisted coding, leveraging “specification-first” approaches and advanced agentic workflows . We Discuss: * Markdown learning loops : Use simple agents.md files for agents to self-update rules and persist context , creating inspectable, low-cost learning; * Intent-first development : As AI commoditizes syntax, defining clear specs and what makes a result “good” becomes the core, durable developer skill; * Effortless documentation : Leverage LLMs to distill messy “brain dumps” or walks-and-talks into structured project specifications, offloading context faster; * Modular agent skills : Transition from MCP servers to simple markdown-based “skills” with YAML and scripts, allowing progressive disclosure of tool details; * Scheduled async agents : Break the chat-based productivity ceiling by using GitHub Actions or Cron jobs for agents to work on issues, shifting humans to reviewers; * Automated tech debt audits : Deploy background agents to identify duplicate code , architectural drift , or missing test coverage , leveraging AI to police AI-induced messiness; * Explicit knowledge culture : AI agents eliminate “cafeteria chat” by forcing explicit, machine-readable documentation , solving the…

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/vanishing-gradients-4989163/episodes/episode-67-saving-hundreds-of-hours-of-dev-time-with-ai-agents-that-learn/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/vanishing-gradients-4989163/episode-67-saving-hundreds-of-hours-of-dev-time-with-ai-agents-that-learn.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.