Lenny's Podcast• Apr 10, 2025• 1:31:40Interview

OpenAI’s CPO on how AI changes must-have skills, moats, coding, startup playbooks, more | Kevin Weil

From Lenny's Podcast

Kevin Weil•Chief Product Officer, OpenAI

Executive Summary

OpenAI operates on the principle that the current AI model is the worst you'll ever use, with significantly more capable models released every two to three months, fundamentally altering the product development landscape.
The core skillset for product managers in the AI era is shifting towards writing 'evals' (evaluations) to measure model performance and mastering the workflow of fine-tuning base models for specific use cases.
OpenAI's product philosophy includes 'iterative deployment' (shipping early), 'model maximalism' (avoiding complex scaffolding around models), and maintaining a lean, 'PM light' organizational structure to move quickly.
There is a massive opportunity for startups to build AI applications across every industry, as OpenAI focuses on the core platform and models, intentionally leaving vertical-specific solutions to the broader ecosystem.

12 quotes

Concerns Raised

OpenAI's once 'massive' 12-month model lead over competitors has shrunk.
The failure of past ambitious projects like Facebook's Libra highlights the execution and reputational risks in launching transformative technology.
The industry has been surprisingly slow to adopt fine-tuning, a critical workflow for optimal model performance.

Opportunities Identified

Startups can build valuable businesses in every vertical by leveraging OpenAI's API, as OpenAI will not build everything.
The cost of using OpenAI's models has decreased by ~100x over two years, dramatically expanding accessibility.
Fine-tuning models for specific use-cases is a key differentiator that can significantly improve product performance.
New modalities like Sora for video generation are unlocking massive creative and productivity gains in industries like filmmaking.

Key Themes

The Accelerating Pace of AI Development

OpenAI's CPO, Kevin Wheel, emphasizes that AI models are improving at an exponential rate, with new, more capable versions being released every two to three months. This rapid advancement means that products on the edge of current capabilities will soon become highly effective, a concept he frames as 'the AI model you're using today is the worst you will ever use.'

This relentless pace forces a fundamental shift in product strategy and planning. Teams must build with future capabilities in mind and understand that current limitations are temporary, creating both immense opportunity and the risk of being quickly outdated.

The Evolving Role of the AI Product Manager

The skills required for product management are changing. Instead of traditional spec writing, the crucial new skills are writing 'evals' to programmatically test a model's performance on specific tasks and understanding how to fine-tune base models with proprietary data.

This signals a necessary evolution for product leaders and teams. To build successful AI products, PMs must become more technical, moving from defining user flows to defining and measuring desired model behaviors and outcomes.

OpenAI's Product & Organizational Philosophy

OpenAI operates with a 'PM light' structure (~25 PMs) to empower its product-focused engineering team. Their philosophy is built on 'iterative deployment'—shipping early to learn from real-world usage—and 'model maximalism,' which prioritizes improving the core model over building complex product features around its current limitations.

This approach explains OpenAI's ability to innovate and scale rapidly. It provides a model for other AI-native companies on how to structure teams and prioritize development in a fast-changing technological landscape.

The AI Platform and Startup Ecosystem

OpenAI sees itself as a platform provider, creating foundational models for 3 million developers to build upon. The company explicitly states it cannot and will not build applications for every industry, creating a vast greenfield opportunity for startups to build vertical-specific AI solutions.

This defines the strategic landscape for entrepreneurs and investors. The most significant near-term value creation in AI may come from companies that leverage powerful base models to solve specific, high-value problems in industries like finance, healthcare, and law.

The Enduring Power of the Chat Interface

Contrary to the belief that chat is a temporary UI for AI, the speaker argues it's a versatile and powerful interface. Because natural language is how humans communicate across a wide spectrum of complexity and intelligence, chat provides a universal and flexible medium for interacting with increasingly capable AI models.

This challenges conventional wisdom about future user interfaces. It suggests that investing in conversational experiences is a durable strategy, as language will likely remain the most effective 'API' for humans to interact with general intelligence.

Get started free

Topics

AI Models Product Management OpenAI GPT ChatGPT Sora Fine-Tuning Model Evals Product Strategy Startup Ecosystem API Economy Iterative Deployment Organizational Design Human-AI Interaction Prompt Engineering

Processed Apr 3, 2026 yt-dlp + mlx-whisper + Gemini