scripod.com
How to Build an Agent-native Product | Mike Krieger

Overview

Shownote

Highlights

Transcript

Chapters

Pins

How to Build an Agent-native Product | Mike Krieger

AI & I

Mar 25
AI & I

AI & I

Mar 25
OverviewShownoteHighlightsTranscriptChaptersPins

Shownote

Mike Krieger built one of the most consequential consumer apps of the last two decades as cofounder of Instagram. He is now at the frontier of determining what makes a breakout AI-native product as co-lead of Anthropic Labs. Dan Shipper talked with Krieger...

Highlights

Mike Krieger, Instagram co-founder and now co-lead of Anthropic Labs, reflects on how AI is reshaping product development—drawing sharp contrasts between the disciplined, human-centered design ethos behind Instagram and today’s accelerated, model-driven landscape.
00:00
Today's models are good at adding features but not at deciding what to cut from products
01:39
Mike Krieger discusses evolution in product-building and reflects on Instagram Stories
02:37
Current AI models are good at adding features but struggle with determining what to cut
07:20
A simple shareable markdown link—inspired by a speech-text app—became popular after a failed vibe-coded editor
09:05
AI models make rewrites quicker and less painful, enabling faster product iteration
11:43
Agents should be able to do anything a user can in an app
24:32
Models like Claude still need human oversight in system architecture and prompting, as the natural tendency in prompting is to over-complicate instructions
36:50
Custom instructions and skill systems help avoid core product bloat in AI
38:57
Being willing to rewrite the stack is essential for modern product development
43:14
Prompting Claude to delegate work to sub-agents keeps the run-loop open, making it feel more like a conversation partner

Chapters

The Harder Problem in Software Design
00:00
Introduction
01:39
What's gotten easier—and what hasn't—about building products in the age of AI
02:33
Why vibe coding creates "indoor trees"
05:00
How rewrites have become a normal part of the development process
09:00
What "agent native" product design means
11:39
How Mike's labs team is structured and the cofounder model
24:27
The best signal for a product bet is someone with "break through walls" conviction
29:33
Navigating enterprise customers while keeping pace with rapid AI change
38:51
OpenClaw, personal agents, and the product question defining 2026
40:54

Transcript

Mike Krieger: The models today are good at adding features. They're not necessarily good about figuring out what to cut out of the product. You can get it to go zero, not just zero to one, but zero to end. Pretty quickly over the matter of hours. It's made...