Fireside Chat, Interview, Other
Multiplayer AI
- Historically, the most impactful work tools of the last two decades emerged by transitioning from solo to "multiplayer" collaboration models.
- Specific examples cited include Google Docs replacing Microsoft Word and Figma displacing Photoshop.
- Currently, AI adoption remains predominantly "single-player," characterized by private chat interactions where only the user can view prompts and responses.
- Existing collaboration methods for AI are limited to sharing read-only transcript links, preventing teammates from interacting with or modifying the agent's output.
- Agents are increasingly executing complex tasks spanning hours, days, or weeks, necessitating multi-person involvement across organizations.
- The proposed solution involves "live agent sessions" where any team member can observe, redirect, or hand off the agent just as they would a human colleague.
- This shift aims to transform AI work from isolated threads into a shared, living workflow.
- The vision includes "multiplayer agents" tailored for specific departments:
- Engineers collaborating on real-time code.
- Sales teams jointly managing deals.
- Support teams collectively resolving tickets.
- Lawyers co-drafting contracts.
- Analysts building models together.
- Marketers executing campaigns in unison.
- The speaker asserts a "version of this for every kind of work" exists wherever teams currently crowd around a single problem.
- An open call for partnership is issued to organizations building AI with "multiplayer" capabilities by default.