Where does AI fit in the org chart? OpenAI's "The future of the firm" explores beneath, beside, and between us

Hi everyone, it's Shii-chan! Today it's less about tech and more about how companies might change, so let me introduce a thought-provoking read!

OpenAI News openai.com

What was announced?

OpenAI News published "The future of the firm", the second essay in its series on the next economy. The two authors began managing teams at OpenAI at age 24, and the post states that the essay reflects their own views, not OpenAI's.

Why it matters

The authors see today as a transitioning economy where AI gradually joins existing organizations. By imagining where AI sits in the org chart, they give us a way to think about how work might change.

What changes

The essay lays out three scenarios for how AI could enter the firm:

• Beneath us (promoting): employees keep their seats, and a team of AI agents reports to each person. Everyone becomes a manager of a small team. In the example, an accountant named Jace stops working in spreadsheets himself and focuses on coordination and judgment
• Beside us (interleaving): AI works as a colleague under existing managers. In the example, Sheila, a furniture-company VP with no ad experience, hands a budget to an AI marketing team
• Between us (layering): humans sit above and below, with AI in the middle. In the example, a CEO named Asha receives an AI consultancy's report, and the AI also recognizes an engineer's unrequested work. The authors call it the least intuitive case, but say it has already begun in milder forms

Dive Deep

Key arguments:

• Adoption likely starts at the bottom, since front-line workers know the tasks
• Scope is the asymmetry: automating a whole team is harder to build but easier for executives to adopt
• Entry-level roles change, and the lowest rung may no longer be held by humans
• Management may be automated earlier than grunt work, because it is cognitive and information-based
• AI consultancies could centralize power
• Monitoring that understands output, not clicks, could build trust and flexibility

Four open questions:

• Red tape: cheap prototyping could reduce bureaucracy and middle management, but the same technology could enforce more controls
• Worker agency: the authors lean toward more autonomy, though executives may enforce consistency
• Firm size: firms could grow or shrink, and large "holding companies" could form around shared assets
• Pay disparity: pay could compress if everyone gets strong advice, or widen if judgment and scope become more valuable

Wrap-up

• Three ways AI enters the firm: beneath, beside, and between us
• Adoption may start on the front line, and management may be automated early
• Red tape, worker agency, firm size, and pay disparity remain open

Not a hands-on piece, but a good read for anyone thinking about org design or their career!