--- schema: 1 type: video title: 'The AI Sandwich: Where Humans Excel in an AI World' 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 EVERY To hear more from Dan Shipper: Subscribe to Every: https://every.to/subscribe Follow him on X: https://twitter.com/danshipper Discover more resources in the episode Compound engineering plugin: https://github.com/EveryInc/compound-engineering-plugin Compound engineering guide: https://every.to/source-code/compound-engineering-the-definitive-guide Compound engineering camp: https://every.to/source-code/compound-engineering-camp-every-step-from-scratch Timestamps: 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 canonical: >- https://liquidgold.cc/channel/everyinc/explore/video/the-ai-sandwich-where-humans-excel-in-an-ai-world-1b2fe3003727 markdown: >- https://liquidgold.cc/channel/everyinc/video/the-ai-sandwich-where-humans-excel-in-an-ai-world-1b2fe3003727.md llms: null parent: 'https://liquidgold.cc/channel/everyinc.md' channel: everyinc updated: '2026-10-24T01:04:29.448Z' language: null children: [] videoId: G0LTv8hQ5Cs durationSeconds: 1711 publishedAt: '2026-04-22T18:51:26.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 0 endSeconds: 11 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=0' heldReason: rights - ordinal: 1 text: null startSeconds: 172 endSeconds: 273 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=172' heldReason: rights - ordinal: 2 text: null startSeconds: 172 endSeconds: 273 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=172' heldReason: rights - ordinal: 3 text: null startSeconds: 285 endSeconds: 388 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=285' heldReason: rights - ordinal: 4 text: null startSeconds: 285 endSeconds: 388 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=285' heldReason: rights - ordinal: 5 text: null startSeconds: 388 endSeconds: 399 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=388' heldReason: rights - ordinal: 6 text: null startSeconds: 399 endSeconds: 674 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=399' heldReason: rights - ordinal: 7 text: null startSeconds: 399 endSeconds: 674 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=399' heldReason: rights - ordinal: 8 text: null startSeconds: 674 endSeconds: 686 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=674' heldReason: rights - ordinal: 9 text: null startSeconds: 779 endSeconds: 871 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=779' heldReason: rights - ordinal: 10 text: null startSeconds: 779 endSeconds: 871 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=779' heldReason: rights - ordinal: 11 text: null startSeconds: 882 endSeconds: 1011 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=882' heldReason: rights - ordinal: 12 text: null startSeconds: 1011 endSeconds: 1022 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1011' heldReason: rights - ordinal: 13 text: null startSeconds: 1022 endSeconds: 1113 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1022' heldReason: rights - ordinal: 14 text: null startSeconds: 1113 endSeconds: 1123 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1113' heldReason: rights - ordinal: 15 text: null startSeconds: 1123 endSeconds: 1232 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1123' heldReason: rights - ordinal: 16 text: null startSeconds: 1232 endSeconds: 1243 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1232' heldReason: rights - ordinal: 17 text: null startSeconds: 1243 endSeconds: 1343 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1243' heldReason: rights - ordinal: 18 text: null startSeconds: 1343 endSeconds: 1353 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1343' heldReason: rights - ordinal: 19 text: null startSeconds: 1408 endSeconds: 1420 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1408' heldReason: rights - ordinal: 20 text: null startSeconds: 1420 endSeconds: 1609 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1420' heldReason: rights - ordinal: 21 text: null startSeconds: 1609 endSeconds: 1711 watch: 'https://liquidgold.cc/watch/G0LTv8hQ5Cs?t=1609' heldReason: rights --- # The AI Sandwich: Where Humans Excel in an AI World 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 EVERY To hear more from Dan Shipper: Subscribe to Every: https://every.to/subscribe Follow him on X: https://twitter.com/danshipper Discover more resources in the episode Compound engineering plugin: https://github.com/EveryInc/compound-engineering-plugin Compound engineering guide: https://every.to/source-code/compound-engineering-the-definitive-guide Compound engineering camp: https://every.to/source-code/compound-engineering-camp-every-step-from-scratch Timestamps: 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 Canonical: https://liquidgold.cc/channel/everyinc/explore/video/the-ai-sandwich-where-humans-excel-in-an-ai-world-1b2fe3003727 · Channel: [Every](https://liquidgold.cc/channel/everyinc.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.