How AI can work across scales, from individuals to organizations to economies, like steel and the steam engine before it, as AI arrives as “infinite minds”
companies with over 1 million people. Running an organization will start to feel like vibe coding. Sarah Guo / @saranormous : an inspired view of the organizations to come @alth0u : imagine writing this when notion app latency is still 800ms to do anything are you out your fucking mind son Sam H Li / @samhli_ : 💡a rare peek into what we are trying to build at @NotionHQ. “The real gains are limited only by our imagination and inertia,” is how we are viewing agents for knowledge work. as great as 2025 was for our launches, i have a strong feeling that the best days are ahead. Matt Slotnick / @matt_slotnick : this is a great essay. i've been grappling with how to conceptualize a future that i am convinced will be revolutionary. applications today are only loosely involved with the majority of the real work that people do. i believe that the applications of the future will Sam Gorman / @gormankind : I'm always surprised that more founders don't study history. I think people miss that it was never about remembering facts and figures. It's about connecting the dots to help you understand where the future is going next. And then as a founder shaping it. I love using Kunal Bhatia / @kunalslab : “...when busywork is delegated to minds that never sleep. Steel. Steam. Infinite minds. The next skyline is there, waiting for us to build it.” - Ivan Zhao Camille Ricketts / @camillericketts : One of the most beautifully written pieces about AI and knowledge work. Kenneth Auchenberg / @auchenberg : “He's become a manager of infinite minds.”
Context & Ripple Effects
The essay extends the AI discussion from task-level automation to organizational design: instead of treating models only as tools, it imagines persistent agents coordinating work across much larger institutions. That builds on the earlier expectation that AI's initial business impact would come through enterprise deployments aimed at changing labor and operations.
Notion's emphasis on agents for knowledge work places the argument in the emerging agentic-workspace layer, where the practical constraint is not only model capability but whether teams redesign processes around delegated work.
First-order effects
- Notion and similar workplace-software providers are pushed to make agents a core work surface rather than an optional assistant, with repetitive knowledge tasks becoming the first targets for delegation.
- Managers gain a stronger incentive to define work as outcomes and constraints that agents can execute continuously, rather than as sequences of manually assigned tasks.
Second-order effects
- Enterprise buyers will judge AI tools less on isolated chat features than on whether they can coordinate work across teams, increasing pressure on workspace vendors to integrate data, permissions and workflows.
- Scaling agent-led knowledge work can also preserve demand for the human review and data-preparation labor documented in the hidden tasker workforce, even as some routine office work is automated.
Third-order effects
- If organizations reliably operate through large pools of AI agents, competitive advantage may shift toward firms with the best process data, governance and computing access—not simply the largest employee base.
- The result could intensify the distributional concerns raised around AI's potential to concentrate economic power; whether it does will depend on how broadly capable agent infrastructure and its gains are distributed.
The trend: AI is moving from a productivity feature toward an organizational operating layer that can execute and coordinate recurring knowledge work at scale.