The AI Sandwich: Where Humans Excel in an AI World (29 min)

← Back to week overview

• Release date: 2026-04-22
• Listen on Spotify: Open episode
• Episode description:

  Most frameworks for working with AI agents assume humans should stay in the loop at every phase. That’s the wrong approach, says Cora general manager Kieran Klaassen.Kieran is the creator of Every's AI-native engineering methodology, compound engineering. His four-step framework—plan, work, review, compound—rebuilds how engineers work with agents. The insight, worked out with collaborator Trevin Chow, is about when to be in the loop and when to step away and let the model handle it. "LLMs are very good at just following steps, doing deep work, working for hours—days even now," Kieran says. "That thing is kind of solved."Kieran and Trevin describe an AI workflow as a sandwich. Agents are the workhorse filling, and humans are the bread, responsible for framing the problem at the start and reviewing the outputs at the end. Every CEO Dan Shipper talked with Kieran for AI & I about why setting the frame of a problem is still hard for agents, why simulated personas won't replace human judgment, Dan's bar for AGI—an agent worth running 24/7 with no off switch—and what Kieran's background as a classical composer taught him about performance, polish, and finding the parts of work that bring you joy.If you found this episode interesting, please like, subscribe, comment, and share!Head to http://granola.ai/every and get 3 months free with the code EVERYTo hear more from Dan Shipper:Subscribe to Every: https://every.to/subscribe Follow him on X: https://twitter.com/danshipper Compound engineering plugin: https://github.com/EveryInc/compound-engineering-pluginCompound engineering guide: https://every.to/source-code/compound-engineering-the-definitive-guideCompound engineering camp: https://every.to/source-code/compound-engineering-camp-every-step-from-scratchDiscover more resources in the episodeTimestamps:   00:00:00 – Introduction and the AI sandwich metaphor 00:02:33 – What compound engineering is and how it’s evolved 00:04:27 – The "work" phase of agentic coding is essentially solved 00:06:27 – Why humans belong at the beginning and the end of an AI workflow 00:11:06 – Dan's argument for why agents can't change frames—and how this will keep us employed 00:16:51 – Full automation is a moving target 00:23:21 – Musical composition as a model for human-AI collaboration 00:26:39 – Find your place in an AI-accelerated world by leaning into what brings you joy

Summary

• 🥪 The AI Sandwich Model: Humans act as the ‘bread’ in workflows, focusing on ideation and polishing, while AI fills the ‘middle’ with execution, creating a balanced human-AI collaboration that enhances creativity and efficiency.
• 🧠 Strategic Human Intervention: Key moments for human input occur at the start for framing problems and at the end for refining outputs with personal taste, preventing generic results and ensuring authentic, joyful work.
• 🔄 Compound Knowledge Loop: The review and compounding steps store learnings in repositories, allowing AI to avoid past mistakes and improve over time, making workflows more reliable without constant human oversight.
• 🎨 Embracing the Artistic Spectrum: Work spans from rote tasks (AI-automated) to creative art (human-led), encouraging professionals to shift toward fulfilling roles that leverage intuition, performance, and what brings personal energy.
• 🚀 Future-Proofing Careers: As AI advances, roles evolve toward product management and beauty-focused refinement, with humans providing irreplaceable frame-shifting and rare expertise, turning potential disruption into opportunity.

Insights

• How can the ‘AI sandwich’ model redefine human roles in knowledge work?
  • Time: 0:00 – 11:02
  • Answer: The ‘AI sandwich’ metaphor positions humans as the bread—essential at the ideation and polishing stages—while AI handles the middle execution, allowing humans to focus on creative framing and personal refinement. This approach prevents full automation by emphasizing human taste, joy, and frame-shifting, ensuring outputs feel authentic and elevated. It matters because it alleviates job loss fears, shifting roles toward more fulfilling, high-impact activities in engineering, design, and beyond.
• Why is human involvement crucial at the start and end of AI-assisted workflows?
  • Time: 2:30 – 8:52
  • Answer: In compound engineering, humans drive brainstorming and ideation to set the right problem frame, while at the end, they validate and polish for beauty and feel, drawing from personal experience and intuition that AI lacks. This prevents generic ‘slop’ and incorporates rare expertise or emotional resonance, making the work uniquely human. It’s vital for maintaining high standards as AI automates routine middle tasks, fostering innovation over rote execution.
• How does compound engineering evolve beyond coding to broader knowledge work?
• What makes full automation of creative work impossible despite AI advancements?

Want the full treatment? Download expanded insights—timestamps, complete answers, and course materials from AI Weekly Insights.