Episode 6 - The future of coding with AI | OpenAI Podcast

In this episode of the OpenAI Podcast, Andrew Mayne interviews Greg Brockman, OpenAI co-founder and president, and Thibault Sottiaux, Codex engineering lead, about agentic coding, GPT-5 Codex, and the future of AI in coding. They discuss the evolution of Codex, the importance of the "harness" (the infrastructure integrating the model with its environment), and the balance between model intelligence and convenience. They also explore the potential of AI in code review, security patching, and creating new tools, as well as the future of agentic systems, compute scarcity, and the importance of learning to code and use AI.

Outlines

Part 1: Introduction and Evolution

Part 2: Experimentation and Deployment

Part 3: Competition, Applications, and GPT-5 Codex

Part 4: Future Vision and Societal Impact

Sign in to continue reading, translating and more.

Open full episode in Podwise

Episode 6 - The future of coding with AI

OpenAI Podcast

Part 1: Introduction and Evolution

Introduction to Agentic Coding and Codex

The Evolution of Codex and the Importance of the Harness

Modalities and the Agent-Driven Interaction

Part 2: Experimentation and Deployment

Experimentation with Form Factors and Internal Tooling

Balancing Internal Use and External Deployment

Current Modalities and Future Integration

Agents.MD and the Future of Agent Communication

Part 3: Competition, Applications, and GPT-5 Codex

The Competitive Landscape and OpenAI's Mission

Future Applications: Security Patching and Code Review

The Impact of Code Review and the Launch of GPT-5 Codex

GPT-5 Codex: Capabilities and the Future of Programming

Part 4: Future Vision and Societal Impact

The Agentic Future and Scalable Oversight

Solving Novel Problems and the Future of AI

Security, Code Quality, and the Role of AI

Material Abundance and Compute Scarcity in 2030

The Importance of Learning to Code and Using AI

Usage Rates and Final Thoughts

Episode 6 - The future of coding with AI

OpenAI Podcast

Part 1: Introduction and Evolution

00:00Introduction to Agentic Coding and Codex

Introduction to Agentic Coding and Codex

04:39The Evolution of Codex and the Importance of the Harness

The Evolution of Codex and the Importance of the Harness

06:51Modalities and the Agent-Driven Interaction

Modalities and the Agent-Driven Interaction

Part 2: Experimentation and Deployment

09:55Experimentation with Form Factors and Internal Tooling

Experimentation with Form Factors and Internal Tooling

14:49Balancing Internal Use and External Deployment

Balancing Internal Use and External Deployment

17:35Current Modalities and Future Integration

Current Modalities and Future Integration

20:35Agents.MD and the Future of Agent Communication

Agents.MD and the Future of Agent Communication

Part 3: Competition, Applications, and GPT-5 Codex

22:25The Competitive Landscape and OpenAI's Mission

The Competitive Landscape and OpenAI's Mission

25:05Future Applications: Security Patching and Code Review

Future Applications: Security Patching and Code Review

27:45The Impact of Code Review and the Launch of GPT-5 Codex

The Impact of Code Review and the Launch of GPT-5 Codex

31:31GPT-5 Codex: Capabilities and the Future of Programming

GPT-5 Codex: Capabilities and the Future of Programming

Part 4: Future Vision and Societal Impact

34:31The Agentic Future and Scalable Oversight

The Agentic Future and Scalable Oversight

37:20Solving Novel Problems and the Future of AI

Solving Novel Problems and the Future of AI

40:04Security, Code Quality, and the Role of AI

Security, Code Quality, and the Role of AI

42:57Material Abundance and Compute Scarcity in 2030

Material Abundance and Compute Scarcity in 2030

45:52The Importance of Learning to Code and Using AI

The Importance of Learning to Code and Using AI

49:09Usage Rates and Final Thoughts

Usage Rates and Final Thoughts