Ex-Uber dev explains his Multi-Agent Workflow

David Ondrej, a former Uber engineer, explains the critical shift in AI development from single-player tools to collaborative, multiplayer AI systems. He argues that current AI tools are often built for human interaction and are not optimized for agent-to-agent or agent-to-human collaboration in a team environment. Ondrej uses the analogy of historical technological shifts, like the internet's impact on publishing, to illustrate the potential for AI to fundamentally change how businesses operate.

He introduces the concept of "multiplayer AI," where agents can work together and integrate into existing workflows. Ondrej showcases a product called Lindy, an AI teammate designed to take on administrative tasks and work collaboratively. He demonstrates how Lindy functions as a part of a team, learning from interactions and contributing to shared knowledge bases. The core idea is that AI agents should become fully integrated team members, capable of performing complex tasks autonomously and contributing to collective intelligence.

Ondrej discusses the importance of context for AI agents, drawing parallels to human communication and knowledge sharing. He highlights that the bottleneck is no longer the technology itself, but the human adoption and integration of these AI systems into their workflows. He contrasts the old way of sharing documents via email with the new paradigm of real-time, collaborative editing through AI agents. Ultimately, Ondrej emphasizes that AI is not just about powerful models, but about building effective AI teammates that can augment human capabilities and transform business operations.