Anthropic’s CPO on what comes next | Mike Krieger (co-founder of Instagram) · Lenny's Podcast: Product | Career | Growth summary

Key takeaways

AI is revolutionizing code generation, with 90-95% of Anthropic's code now being AI-generated, leading to new challenges in product development.
Embedding product managers with engineering teams significantly enhances product impact and accelerates innovation.
Recognizing when to pivot or shut down a product is crucial for focusing resources on more promising ventures.
Effective prompting techniques are essential for maximizing AI's capabilities and eliciting meaningful responses from models.
AI startups should specialize in niche markets while leveraging deep industry knowledge and customer relationships to thrive.
The Model Context Protocol (MCP) represents a significant advancement in how AI products can integrate, enhancing user experience and functionality.

Summary

Transforming Product Development with AI

Mike Krieger discusses how 90-95% of code at Anthropic is now generated by AI, resulting in both efficiencies and new bottlenecks. The rapid increase in AI-generated code has forced the company to rethink its development strategies, particularly around pull requests and the engineering review process. Fully embracing AI has changed the traditional PM-Engineering dynamic, leading to earlier prototyping and the necessity for engineers to adjust their workflows to cater to AI's capabilities.

The Evolving Role of Product Managers and Teams

Krieger emphasizes the value of embedding product managers within engineering and research teams, allowing for more significant impact and collaboration. This integrated approach not only fosters innovation but also helps navigate the complexities brought on by rapid AI advancements. He outlines how product teams can still bring meaningful contributions to AI products, focusing on strategy, user comprehension, and exploring the potential of new functionalities.

Insights from Shutdown of Artifact

Krieger candidly shares the journey behind shutting down Artifact, despite its initial promise. He reflects on the challenges of fostering organic growth in a niche market and the impact of external factors on user experience. The decision stemmed from a lack of sustainable growth and resource efficiency, revealing the importance of recognizing when to pivot and focus energy on more promising ventures.

Prompting Techniques for Maximal AI Utilization

Listeners are introduced to specific techniques for querying Claude effectively, including the importance of pushing the boundaries of the AI's reasoning capabilities. Krieger stresses the significance of crafting prompts that leverage the model's strength, suggesting users engage in more robust dialogues that elicit in-depth responses rather than surface-level answers. This empowers users to maximize Claude's utility in their projects.

Market Positioning as an AI Startup

Krieger shares insights for AI startups on how to carve out a niche amid the competition from established players like OpenAI. He urges founders to focus on specific market needs, deepen their understanding of customer dynamics, and to innovate beyond the conventional interfaces of AI tools. The future of AI lies in leveraging industry expertise and creating solutions that address specific user challenges.

The Role of MCP (Model Context Protocol)

MCP is positioned as a game-changer in AI product integration. Krieger discusses how MCP enables AI models to utilize contextual information and memory effectively, creating a richer interaction experience. This framework encourages innovation by allowing different applications and systems to leverage AI more seamlessly, enhancing product functionality across the board.

Show notes

Mike Krieger is the chief product officer of Anthropic and the co-founder of Instagram. After leaving Meta, he co-founded Artifact, an AI-powered news app that I absolutely loved, and joined Anthropic to lead product in 2024.

In this episode, you'll learn:

• How Anthropic uses AI to write 90-95% of code for some products and the surprising new bottlenecks this creates

• Why embedding product managers with AI researchers yields 10x the impact of traditional product development

• The three areas where product teams can still add massive value as AI gets smarter

• How Anthropic plans to compete with OpenAI long-term

• How to use Claude as your product strategy partner (with specific prompting techniques)

• Why Mike shut down Artifact despite loving the product, and what founders can learn from it

• Where AI startups should build to avoid getting killed by OpenAI, Anthropic, and Google

• Why MCP (Model Context Protocol) might reshape how all software works

• The counterintuitive product metrics that matter for AI

• How to evaluate whether your company is maximizing AI’s potential or just scratching the surface

—

Brought to you by:

Productboard—Make products that matter

Stripe—Helping companies of all sizes grow revenue

OneSchema—Import CSV data 10x faster

—

Where to find Mike Krieger:

• X: https://x.com/mikeyk

• LinkedIn: https://www.linkedin.com/in/mikekrieger/

—

Where to find Lenny:

• Newsletter: https://www.lennysnewsletter.com

• X: https://twitter.com/lennysan

• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/

—

In this episode, we cover:

(00:00) Introduction to Mike Krieger

(04:20) What Mike has changed his mind about regarding AI capabilities

(07:38) How to avoid scary AI scenarios

(08:55) Skills kids will need in an AI world

(11:53) How product development changes when 90% of code is written by AI

(17:07) Claude helping with product strategy

(21:16) A new way of working

(23:55) The future value of product teams in an AI world

(27:18) Prompting tricks to get more out of Claude

(29:52) The Rick Rubin collaboration on “vibe coding”

(32:42) How Mike was recruited to Anthropic

(35:55) Why Mike shut down Artifact

(42:41) Anthropic vs. OpenAI

(47:11) Where AI founders should play to avoid getting squashed

(51:58) How companies can best leverage Anthropic’s models and APIs

(54:29) The role of MCPs (Model Context Protocols)

(58:25) Claude’s questions for Mike

(01:03:15) Claude’s heartfelt message to Mike

—

Referenced:

• Anthropic: https://www.anthropic.com/

• Claude Opus 4: https://www.anthropic.com/claude/opus

• Dario Amodei on X: https://x.com/darioamodei

• AI 2027: https://ai-2027.com/

• Tobi Lütke’s leadership playbook: Playing infinite games, operating from first principles, and maximizing human potential (founder and CEO of Shopify): https://www.lennysnewsletter.com/p/tobi-lutkes-leadership-playbook

• Claude Shannon: https://en.wikipedia.org/wiki/Claude_Shannon

• Information theory: https://en.wikipedia.org/wiki/Information_theory

• TypeScript: https://www.typescriptlang.org/

• Python: https://www.python.org/

• Rust: https://www.rust-lang.org/

• Bending the universe in your favor | Claire Vo (LaunchDarkly, Color, Optimizely, ChatPRD): https://www.lennysnewsletter.com/p/bending-the-universe-in-your-favor

• Announcing a brand-new podcast: “How I AI” with Claire Vo: https://www.lennysnewsletter.com/p/announcing-a-brand-new-podcast-how

• A conversation with OpenAI’s CPO Kevin Weil, Anthropic’s CPO Mike Krieger, and Sarah Guo: https://www.youtube.com/watch?v=IxkvVZua28k

• Jack Clark on LinkedIn: https://www.linkedin.com/in/jack-clark-5a320317/

• Artifact: https://en.wikipedia.org/wiki/Artifact_(app)

• Joel Lewenstein on LinkedIn: https://www.linkedin.com/in/joel-lewenstein/

• Daniela Amodei on LinkedIn: https://www.linkedin.com/in/daniela-amodei-790bb22a/

• Boris Cherny on LinkedIn: https://www.linkedin.com/in/bcherny/

• Gunnar Gray on LinkedIn: https://www.linkedin.com/in/gunnargray/

• The Model Context Protocol: https://www.anthropic.com/news/model-context-protocol

• The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO): https://www.lennysnewsletter.com/p/the-rise-of-cursor-michael-truell

• Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (CEO and co-founder): https://www.lennysnewsletter.com/p/building-lovable-anton-osika

• Inside Bolt: From near-death to ~$40m ARR in 5 months—one of the fastest-growing products in history | Eric Simons (founder and CEO of StackBlitz): https://www.lennysnewsletter.com/p/inside-bolt-eric-simons

• Jimmy Kimmel Live: https://www.youtube.com/user/JimmyKimmelLive

• ChatGPT: https://chatgpt.com/

• Gemini: https://gemini.google.com/app

• OpenAI’s CPO on how AI changes must-have skills, moats, coding, startup playbooks, more | Kevin Weil (CPO at OpenAI, ex-Instagram, Twitter): https://www.lennysnewsletter.com/p/kevin-weil-open-ai

• Windsurf: https://windsurf.com/

• Menlo Ventures: https://menlovc.com/

• Harvey: https://www.harvey.ai/

• Manus: https://manus.im/

• Bench: https://www.bench-ai.com/

• Strategy Letter V: https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

• Kevin Scott on LinkedIn: https://www.linkedin.com/in/jkevinscott/

—

Recommended books:

• The Goal: A Process of Ongoing Improvement: https://www.amazon.com/Goal-Process-Ongoing-Improvement/dp/0884271951

• The Way of the Code: The Timeless Art of Vibe Coding: https://www.thewayofcode.com/

• The Hard Thing About Hard Things: Building a Business when There Are No Easy Answers―Straight Talk on the Challenges of Entrepreneurship: https://www.amazon.com/Hard-Thing-About-Things-Building/dp/0062273205

—

Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.

—

Lenny may be an investor in the companies discussed.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.lennysnewsletter.com/subscribe

Anthropic’s CPO on what comes next | Mike Krieger (co-founder of Instagram)