Skip to content
The Twenty Minute VCThe Twenty Minute VC

Mike Krieger, Instagram CoFounder & Anthropic CPO: Where Will Value Be Created in an AI World?|E1265

Mike Krieger is the Co-Founder of Instagram and now CPO @ Anthropic. ---------------------------------------------- In Today’s Episode We Discuss: (00:00) Intro (00:50) Where Will Value Be Created and Sustained in a World of AI? (01:40) Are Foundation Models Commoditised Today? (04:31) Should Founders Build for the Models of Today or Build for Models of the Future (06:55) Why Will Models Become More Different Than More Similar (12:59) Will Human or Synthetic Data Be More Prominent in the Future (18:02) Model Quality vs. Product UX (20:12) The Competitive Landscape of AI (31:49) Do We Underestimate China's AI Capabilities (33:31) What Did Anthropic Learn from Deepseek (34:44) Is Deepseek a Sustaining and Credible Threat? (38:09) Transitioning from Model Provider to Application Provider (43:44) What is the Role of a Software Developer in the Future (48:31) Balancing API and Consumer Products (52:25) Is Europe Stronger or Weaker in a World of AI (52:59) Quick-Fire Round ----------------------------------------------- Subscribe on Spotify: https://open.spotify.com/show/3j2KMcZTtgTNBKwtZBMHvl?si=85bc9196860e4466 Subscribe on Apple Podcasts: https://podcasts.apple.com/us/podcast/the-twenty-minute-vc-20vc-venture-capital-startup/id958230465 Follow Harry Stebbings on X: https://twitter.com/HarryStebbings Follow Mike Krieger on X: https://twitter.com/mikeyk Follow 20VC on Instagram: https://www.instagram.com/20vchq Follow 20VC on TikTok: https://www.tiktok.com/@20vc_tok Visit our Website: https://www.20vc.com Subscribe to our Newsletter: https://www.thetwentyminutevc.com/contact ----------------------------------------------- #20vc #harrystebbings #mikekrieger #anthropic #cpo #ai #openai #deepseek

Mike KriegerguestHarry Stebbingshost
Mar 2, 20251h 2mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Mike Krieger on where AI value emerges: products, partners, not models

  1. Mike Krieger, Instagram cofounder and now Anthropic’s CPO, argues that durable AI value will accrue to companies with differentiated go‑to‑market, deep domain knowledge, and proprietary data, not just to frontier model labs.
  2. He explains how model providers must evolve from “token vending machines” into long‑term AI partners, while also deciding when to build first‑party applications versus enabling an ecosystem via APIs.
  3. Krieger dives into product and UX challenges unique to AI—non‑determinism, leaky abstractions, eval gaps, trust, and ‘vibes’—and how these shape everything from coding tools to enterprise workflows.
  4. He also reflects on global competition (including China and DeepSeek), the changing role of software engineers, and how AI may accelerate breakthroughs in domains like healthcare and drug discovery.

IDEAS WORTH REMEMBERING

5 ideas

Sustainable AI startups need real-world moats: distribution, domain expertise, and proprietary data.

Krieger stresses that the most defensible companies will pair foundation models with deep understanding of specific industries (e.g., finance, legal, healthcare), strong relationships, and data others cannot easily access.

Don’t wait for perfect models; build now so you’re ready when capabilities leap.

Teams that experiment early, suffer current model limits, and build context in a domain are best positioned to capitalize when a new model crosses a quality threshold that finally makes their product viable.

Model providers must offer AI partnerships, not just APIs that trade tokens for tokens.

Anthropic’s strategy is to differentiate via talent, distinctive model “character,” and deep co-design relationships with customers, moving beyond commoditized API access toward long-term, high-trust collaboration.

AI product design now fuses model quality, prompting, evals, and UX into one discipline.

Because models are non-deterministic, designers and PMs are effectively building scaffolds around stochastic systems; choices like follow-up questions, reasoning depth, and memory become core UX decisions, not implementation details.

Agentic coding will shift engineers from code authors to orchestrators and reviewers.

Krieger sees developers increasingly delegating tasks to AI agents, focusing on what to build, system design, and code review (including security and correctness), while tools run experiments, test UI flows, and propose implementations.

WORDS WORTH SAVING

5 quotes

The thing that's gonna give you legs and be durable over the long run is being able to sell into those places, understand them uniquely, and get better for being deployed there over time.

— Mike Krieger

Don't wait around for the models to be perfect. Be exploring in this space, be frustrated by the current generation of the models, and then be very aggressively trying the next one.

— Mike Krieger

Models over time get more different rather than more similar.

— Mike Krieger

We are in day one around: is AI an indispensable part of most people's work? And I think the answer is no.

— Mike Krieger

You can't separate model quality from product and UX anymore… you're designing a scaffold around a fundamentally non-deterministic system.

— Mike Krieger

Where value will be created in an AI-driven decadeStartups vs incumbents: using vertical knowledge, GTM, and dataModel-layer strategy: differentiation, talent, and AI partnershipsProduct design and UX for non-deterministic AI systemsAgentic coding tools and the evolving role of software engineersCompetitive dynamics among AI labs (DeepSeek, OpenAI, LLaMA, China)First-party products vs API platforms and implications for speed

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome