Shopify CEO: This Is Your Edge Against AI
Tobi LütkeSpeaker- Shane ParrishHost
Key ideas
Lütke describes how Shopify works with AI agents, argues that responsibility and judgment stay human, and says the best path often lacks quick feedback.
From a machine transcript of the audio. Quotes link to the moment they are said.
- 1
Make AI use visible to everyone
Lütke says Shopify's Slack agent, River, works only in open channels so people can watch colleagues use it, his attempt to reproduce what offices once taught by osmosis. He estimates that up to about 50% of pull requests now come out of such chat conversations rather than traditional engineering work.
- 2
Let AI prepare the call, not own it
Lütke says he uses AI to create the right environment for judgment rather than to supply it, because machines can't take responsibility. For major decisions his agent runs a small council of expert roles across several models and synthesizes the results, which he puts at about half an hour and $15 to $20 in tokens.
- 3
Watch for over-output from lazy work
Asked what AI has made worse, Lütke says lazy work now shows up as too much output rather than too little. His examples are AI pull requests the author never really read, passed to colleagues to review, and long emails the recipient has to compress again with AI.
- 4
Distrust the option with fast feedback
Lütke argues that finding a good choice is not the hard part; choosing among several good ones is. He says he starts out suspicious of the option that shows results this quarter and of the orthodox solution, while allowing that orthodoxy can be correct in regulated areas such as payments. He expects taste and judgment, built through many reps, to matter most in ten years.
- 5
Prune and refound instead of adding
Using SpaceX's Raptor engine as his example, Lütke says things improve through removal and rebuilding, not addition. He suggests a department sometimes needs a refounding event, a 2.0 version started from what has been learned, and says failure is never a problem unless it is catastrophic.
In Tobi’s words
“Here's the thing that LLMs and machines cannot do. Machines can't take responsibility. And I think this is actually probably the most overlooked thing in the entire stack.”
Tobi Lütke·Shopify CEO: This Is Your Edge Against AI·Watch at 14:12(opens the original in a new tab)“I find a very high correlation between the right path and the ones that don't have feedback loops attached.”
Tobi Lütke·Shopify CEO: This Is Your Edge Against AI·Watch at 37:59(opens the original in a new tab)“Things need to be pruned. You cannot make things better and better by adding stuff. You can't. You must prune. You must take steps. You must rebuild. You must create an end for things.”
Tobi Lütke·Shopify CEO: This Is Your Edge Against AI·Watch at 59:04(opens the original in a new tab)About this entry
- Published
- Sep 15, 2026
- Channel
- The Knowledge Project Podcast
- Type
- Video
- Length
- 1:04:49
- Language
- English
- Added
- Nov 2, 2026
- Source last checked
- Nov 2, 2026
What’s available here
Original source
Linked. Opens on YouTube.
Player on this page
Available. YouTube’s own player, loaded only when you press play.
Full text or transcript
Not reviewed. No license to show a text version here.
Our summary
Draft. Written by us with AI from a machine transcript of the video’s audio (Whisper) and checked against it. Machine transcripts can mishear; each quote links to the moment so you can hear it. Not yet rights-reviewed.
Ask questions with AI
Not reviewed. Off until there is licensed text to work from.