The AI-Native Developer: From Tools to Teammates

Arc of AI 2026 Video Coming Soon

Abstract

A year ago, the industry was debating AI coding assistants. Today, we’re building with AI agents as collaborative teammates, and that changes everything. In a few pockets of conversation our thoughts have evolved beyond “will AI take our jobs?” to more interesting questions: How do we engineer context for reliable AI systems? What does evaluation look like when agents are making decisions? How do we build workflows where humans and AI agents contribute their distinct strengths? However others are firmly stuck in the worry of a future without software development as we know it today. In this session, we’ll explore what it means to be an AI-native developer in 2026, someone who thinks in terms of agent skills, context engineering, and orchestrated workflows rather than just autocomplete. These changes aren’t about replacing developers. They’re about expanding what’s possible and building systems we couldn’t imagine a year ago, and developing new skills that make us more valuable, not less. Our future is a fundamentally creative, deeply human job.

Resources

Writers, Word Processors, and the Fear of New Tools

Grace Hopper and the First Compiler

The Parade of Panic: Every Abstraction Leap

Prompt Engineering Is Dead

Evals: From Vibes to Engineering

Context Engineering

The good-oss-citizen Case Study (15% to 99%)

Agent Security: The Supply Chain Problem

Patrick Debois: “What If Context Was Like Code?”

Olivier Pomel: Autonomous Incident Response

AI Coding Agents and IDEs

  • GitHub Copilot — GitHub’s AI pair programmer, powered by OpenAI models.
  • Cursor — AI-native code editor built on VS Code with agent mode.
  • Claude Code — Anthropic — Anthropic’s agentic coding tool that lives in your terminal.
  • Windsurf — AI-native code editor by Codeium with the Cascade agentic engine.
  • OpenAI Codex — OpenAI’s coding agent with CLI and cloud-based development.
  • Gemini Code Assist — Google — Google’s AI coding assistant with agent mode for multi-file tasks.

The Agentic SDLC and Context Lifecycle

Previous Versions of This Talk

Books by the Speakers

Conference

  • Arc of AI 2026 — The conference where this keynote was delivered. Austin, TX, April 13–16, 2026.

Speakers