TL;DR
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
An essay published on OpenAI’s website proposes three ways AI might be integrated into firms: beneath workers as supervised agents, alongside them as colleagues, or between managers and staff coordinating work. The authors describe these as illustrative scenarios, not predictions, and argue that outcomes such as job changes, layoffs and pay effects remain uncertain.
Two authors writing in an essay hosted by OpenAI have outlined three possible ways firms could reorganize as AI becomes more capable: systems could work beneath employees, alongside them as colleagues, or between managers and staff to coordinate work. The authors present these models as illustrative scenarios, not forecasts, and say elements of all three are already appearing in subtle ways.
The essay is the second installment in a series on the next economy and is described by its authors as an independent contribution to a platform exploring an AGI future. They explicitly say it reflects their views, not those of OpenAI or their colleagues. Their argument focuses on firms adopting AI while humans remain part of the workforce, rather than on a future of complete automation.
In the first model, AI agents work below individual employees. Workers would delegate routine tasks to agents, supervise their output and potentially coordinate groups of specialized systems. The authors imagine employees gradually spending more time on communication, coordination and creative work. They also describe a hypothetical accountant, Jace, whose agents handle spreadsheet tasks while he takes on responsibilities more like those of a manager. This is an example in the essay, not a reported case study.
The other models put AI beside people on the org chart, handling roles or functions directly for human managers, or between managers and workers, coordinating work and monitoring whether executive plans are carried out. The authors say real firms may combine elements of these arrangements. They argue adoption could begin with employees who understand specific tasks, then expand as AI systems become able to perform broader functions.
How AI Could Reshape Company Roles
The scenarios matter because they frame AI adoption as a potential change in who directs work and how responsibilities are distributed, not only as a question of whether a task can be automated. If agents take on routine work, employees could be asked to supervise systems, coordinate output or assume broader responsibilities. That shift could affect job design, management structures and the skills companies reward.
But the essay does not establish that workers will reliably be promoted, that jobs will be preserved or that pay will rise. Its optimistic possibility—that higher productivity creates more work to manage and increases the value of human judgment—depends on whether people continue to contribute value and whether firms expand their ambitions rather than reduce headcount. The authors present this as a possibility, not a measured outcome.
The essay also raises questions about worker agency and organizational control. AI placed between executives and staff could help coordinate activity, but the arrangement could also alter how closely work is monitored and how much discretion employees have. The source sketches those tensions rather than documenting effects at companies that have adopted such structures.
Products Worth Considering
Apple Watch SE 3 GPS 44mm Smartwatch
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Assistants to Hybrid Firms
The authors open with their experience managing teams at a young age, saying their own path at OpenAI has differed from the traditional progression in which junior employees first handle routine work and gain responsibility over time. They describe working with AI systems as a form of delegation: the systems can do useful grunt work, make mistakes and improve as users learn how to manage them. Those observations are the authors’ account of their experience, not an independent evaluation of AI performance.
The essay’s central premise is that adoption will be gradual and uneven. Current systems may handle only some tasks, making them more useful when paired with workers who know the details of the job. If systems can later perform broader functions, the authors suggest senior managers may use them more directly. The piece examines what such hybrid firms could mean for red tape, employee agency, company size and pay disparity, but the supplied text does not provide data establishing outcomes in those areas.
The source frames the proposed organizational forms as possibilities for a period in which AI is integrated into existing firms. It does not announce a new OpenAI product, company policy or workforce plan. The authors state that the essay represents their own views and is not an official position of OpenAI.
““None of these scenarios will play out exactly as we describe, but we think aspects of all three will come true.””
— The essay’s authors
Jobs, Pay and Adoption Remain Open
The essay does not provide company-level evidence or forecasts showing how common any of the three models are, how quickly firms might adopt them, or which industries could change first. Its reference to subtle examples already occurring is not accompanied in the supplied material by named organizations, measurements or specific cases.
It is also unclear whether increased productivity would lead firms to expand output and retain workers, as the authors’ first scenario suggests, or to reduce staffing. The essay conditions its argument on people continuing to add value and notes that models may not yet be capable of handling all tasks. It does not quantify the effect on wages, job numbers, company size or workplace monitoring.
The supplied source text ends partway through a discussion of whether people may continue to be valued in production. That leaves the essay’s full treatment of the conditions and counterarguments unclear here. Readers should not treat the hypothetical Jace example or the proposed rise in employee responsibility as reported evidence of a broader labor-market trend.
Watch for Evidence Inside Firms
The next step is not a specific event announced by the authors, but whether organizations put similar structures into practice and disclose what changes. Evidence to watch for includes which tasks are delegated to agents, who is accountable for errors, whether human roles are redesigned and whether staffing or pay changes as a result.
Further installments in the series may develop the authors’ discussion of firm size, red tape, worker agency and pay disparity. The supplied material does not state a publication schedule or name a next installment. Until more detailed examples or independent data are available, the three models are best read as a framework for discussing possible organizational changes, not as a settled account of the future of the firm.
Key Questions
What are the three AI-era firm models in the essay?
The authors describe AI working beneath employees as supervised agents, beside people as colleagues on the org chart, or between managers and staff to coordinate and monitor work. They say firms could combine parts of these models.
Does the essay predict that AI will eliminate jobs?
No. The essay focuses on firms where people remain involved and describes a scenario in which employees take on broader responsibilities as AI handles some previous tasks. It does not establish that layoffs will not occur, and its account is a possibility rather than a labor-market forecast.
Is this an official OpenAI position?
The authors say the essay reflects their views, not those of OpenAI or their colleagues. Its publication on OpenAI’s website does not make the scenarios an official company policy or forecast.
What evidence does the essay provide for these changes?
The supplied text gives the authors’ personal observations and hypothetical examples, including an accountant who delegates spreadsheet tasks to agents. It does not provide named company case studies or data measuring adoption, employment, productivity or pay effects.
What remains uncertain about the future of the firm?
It remains unclear how widely firms will adopt these structures, whether AI-driven productivity will preserve jobs or reduce staffing, and how roles, wages and employee autonomy may change. The authors present these outcomes as open questions tied to AI capabilities and how firms choose to use them.
Source: openai.com
