One Version of the Human-AI Future of Work
It’s been a while since my last article on ReArrangement and The Future of Work. However, this past week, I spoke on a panel during NYC Tech Week and we explored the topic, given the updated nature of technology (particularly Artificial Intelligence – AI). This prompted me to revisit my 2022 writings on the future of work and consider how updated trends are influencing ReArrangement. For context, check out Part 1 and Part 2. We’ll call this article Part 3.
My thought at the time and it remains today, is that the world is rapidly changing given advances in technology. The advances I discussed in 2022 revolved around the migration of knowledge work from in-person to remote in the post-pandemic era. That much of what we could achieve in an office could be accomplished at a distance. My thesis was that these new arrangements were a result of accelerated longer-term trends toward distributed work and that the distribution would lead to ever more loosely coupled systems, marketplaces for talent, and smaller organizations.
With the rise of AI, this trend is once again accelerating. Organizations are rethinking their talent structures and testing new modalities for achieving work. One example is through the deployment of synthetic employees, or AI agents. These tools are expected to enable organizations to hand off all manner of tasks, and particularly those that have been well defined by knowledge workers. This trend, if effective, will allow legacy organizations to continue to shrink their ranks and startups to reduce headcount growth. On the other hand, AI-related work is creating its own set of new challenges and job roles. So, while trends point toward people losing work, there are also trends pointing toward them gaining it.
Here we see another example of the need for ReArrangement in how we get work done. Today, we continue to live in flux between organization structures and technological advancement. This flux deserves examination of the trends and exploration of what lies ahead. ReArrangement implies we need updated models to match people to opportunities and technology to use-cases. That’s what we will explore here.
Is AI Taking or Making Jobs?
Today, people are afraid that AI will take their jobs. Indeed, a recent Reuters Poll of 4,531 Americans found that more than half feared that AI would take their job or the job of someone in their household [1]. Layoff reporting illustrates that the fear is not unfounded. Research from Challenger, Gray and Christmas shows that layoffs attributed to AI are on the rise, with AI-related cuts through May 2026 (87,714 cuts) already surpassing the total in 2025 (54,836 cuts) [2].
On the organization side, a 2025 survey from McKinsey and Co. using a global sample of organizations of varying features (e.g. size, industry, etc.) found that 88% are regularly using AI in at least one business function [3]. However, they found that 62% of firms using AI are still in the experimentation and piloting phases while the remaining 38% are either scaling or have fully scaled AI across their organizations. Critical to AI’s impact on employee retention, 32% of the organizations expected a decrease in the size of their workforce in the coming year, 43% expected no change, and only 13% expected increases.
Thus, AI is being deployed across a large majority of organizations, though many are in the early stages, and the fear, likelihood, and actuality of it displacing people exists. However, according to LinkedIn’s January 2026 labor market report, 1.3 million AI-related job opportunities have appeared in the prior two years [4]. The roles, as the report explains, didn’t exist five years ago, but they have quickly become essential to digital economies.” The researchers are calling these roles “new-collar,” and require technical fluency, manual capability, and continual adaptability.
Another Case for ReArrangement
AI is both removing and adding jobs as organizations adapt their work practices to the new technology. That is, we see the push and pull of ReArrangement, which appeared during 2020 pandemic ear of remote work – with variations in adoption, implementation, and outcomes. However, though distinct, it would be too simplistic to separate these events. Rather, they are compounding factors of the larger trend of ReArrangement.
From a macro-view we can see waves of change across time in the way organizations and individuals interact. For some time, employment has become more tenuous, more temporary, more…flexible. I expect this flexibility will continue, with organizations engaging individuals and synthetic systems (e.g., agents) on new terms, with rapid deployment and continuous iteration.
However, like I explained in Part 1 and Part 2, we do not yet have systems to achieve this level of rapid deployment on the human side of work and challenges remain for deploying AI at scale.
ReArrangement for Talent
In May 2026, the Bureau of Labor Statistics showed that the average time job seekers spent unemployed was 26 weeks and a survey by Robert Half of 450 job seekers found that 68% of them expected their job search to be longer than previous searches. With average private non-farm weekly earnings at $1,287.28, going 26 weeks without a job means missing approximately $33,469.28 in wages.
For organizations, the process is long, but not nearly as drawn out: SHRM’s 2025 benchmark puts median time-to-fill at roughly 6 weeks and their 2025 Talent Trends report shows that 69% of employers face difficulties filling full-time roles. Organizations also absorb substantial costs: SHRM’s 2025 benchmark puts average cost-per-hire at $1,200 for nonexecutive roles and $10,625 for executive roles.
Thus, it is time consuming and expensive for job seekers and organizations to match with one another. The flux of people moving within and without organizations and the high cost of placement creates high friction in the labor economy. This is because it is often time consuming to determine who is the best fit for a role and to bring them on board.
ReArrangement for AI
While AI can be deployed flexibly within an organization context seemingly more so than the traditional new hire, it too has challenges. In their 2026 AI adoption survey, WRITER and Workplace Intelligence found that 29% of employees are actively sabotaging their company’s AI strategy and that 75% of executives say their AI strategy is just for show [8]. Further, only 23% of the companies surveyed are seeing a significant Return on Investment (ROI) from their AI agents. Thus, employees and executives alike are demonstrating unproductive behaviors and attitudes toward AI and this may be showing up in their ROI.
That is, the journey toward implementation is not just one of technological advancement but one of continued human-technological integration.
ReArrangement Toward Fluidity in Talent and AI Deployment
Given the above, we see that ReArrangement continues to be underway and a thoughtful approach is needed to develop a fluid talent and AI deployment model to match the needs of people, organizations, and societies. To do so, I believe we need to rethink the very basic elements of how we organize.
In Part 2 of this series I summarized the phases of pre- and post-industrial labor, work contexts, and phenomena, where the concept of work did not exist in pre-industrial contexts, emerged in the industrial context, and became a dominant life structure in the post-industrial context. That is, employment emerged as a concept and has since become core to human identity.
Here is one version of the future that could emerge, given the trends.
Maybe we are entering an era of post-employment. Not that people won’t work, but rather that they will engage organizations on a more temporary basis, help them achieve specific goals, then disengage. AI agents may be the keeper of practices and knowledge systems and maintain continuity across the organization. However, these systems would be missing the novel inputs that humans produce (e.g., new business strategies, technical advancements, innovation, etc.). Thus, in this version of ReArrangement, AI would be generalist, all purpose, highly powerful tool with organization context, wielded by specialist humans inside the organization on assignment.
This may occur gradually. Maybe legacy organizations keep their larger headcounts by redefining what roles they need to achieve their goals with AI in the loop. However, maybe startups never grow their headcounts, opting for a more hybrid structure that mirrors the post-employment paradigm. That is, over time, as companies that were once startups become legacy organizations, more will be operating in the post-employment model.
This potential model achieves several challenges I see grating against the current paradigm.
First, people are afraid of losing their jobs, and indeed, they are, and it takes a long time to become employed. Thus, in a paradigm where organizations don’t have many employees, they will be constantly sourcing real-time engagements, providing people with access to flexible and myriad opportunities across organizations. This mirrors the way in which we have come to experience the marketplace as consumers (e.g., same day doctor appointments, grocery delivery, etc.). It would make sense that this model extends to how humans make their living. However, the looseness here requires whole new systems of support (e.g., updated models for healthcare, benefits, retirement, etc.). See Part 2 for my exploration of what I call “super organizations,” and how they might help.
Second, in this paradigm, AI tools have their place in the human workflow, not the other way around. As specialists, humans could use AI to gain rapid organization context, see insights from disparate points of data, and make informed decisions about the projects within their scope. This puts humans in the driver seat and empowers them to remain highly valuable and relevant in the talent ecosystem. This also creates a supportive set of circumstances where humans remain the source of knowledge, refining and curating with AI as a partner, and would serve to encourage our critical need for continuous human development.
Third, organizations would have a fluid model for match the pace at which they change and shift to compete and meet the needs of the market. With leaner organization structures, on-demand talent, and robust AI-infused information systems, established organizations can move with the speed of a startup but maintain the robust practices embedded in enterprise systems.
The above model, while hypothetical, is an example of thinking into the distant future so that we can design toward productive circumstances in which humans and organizations thrive. It is only one version. Given the trends, what do you expect to happen in the future of work?
References
[1] https://cybernews.com/ai-news/democrats-republicans-ai-jobs/
[3] https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
[5] https://www.bls.gov/news.release/archives/empsit_06052026.htm
[7] https://www.shrm.org/content/dam/en/shrm/research/2025-recruiting-benchmarking-report.pdf
[8] https://go.writer.com/ai-adoption-enterprise-2026
