In 2022, I was hired to build an AI operation at a health tech startup. At the time, we were pioneers in the use of AI in healthcare, which required significant human oversight, which one day became unnecessary. GPT-4 started and within a short period of time I realized that my role was no longer relevant. My employer came to the same conclusion. There was no plan to retrain me or redeploy my skills into new versions of the work. My job just disappeared.
I say this not as a warning, but as context. When I see a wave of mass layoffs being justified as an AI transformation, I’m not reading it from afar. I’m on the other side of that decision.
What I learned on the way down
What I didn’t fully understand at the time, but understand now, was that employers weren’t changing. They were optimizing. Layoffs provide a clear calculation. These provide immediate cost savings and a simple story for boards looking to see a return on their AI investments. What they cannot offer is increased capacity, creative utilization, or new types of work. I was a lost cost. The underlying competency question — what should this job be? — was never asked.
When companies like Meta and Microsoft cut tens of thousands of employees, many leaders see it as a necessary step toward becoming more “AI native.” We are aware of what is actually going on. They are choosing the fastest path to efficiency rather than the harder path to reinvention. They are pausing their path to change because it is easier than rewiring the way they work. I know firsthand the difference between these two.
what I did differently
Currently, I am the Director of AI Operations at Pearl, an independent professional AI company. We’ve taken a different path, upskilling our employees, reimagining roles, and having uncomfortable conversations sooner than most companies would like. During that conversation, one conversation caught my eye.
I work closely with a technical writer who recently asked a question that many of our employees are quietly thinking about. “What is my job now when AI can do so much of my work for me?” She realized that much of the value she provides—drafting, editing, and refining documents—is now available to everyone through the effective use of AI. I recognized the moment immediately. I was living it.
The difference this time was that I didn’t dodge the question. We asked them to answer together. Now, she operates like an entire technical writing department with a team of AI agents who help proofread, edit, and standardize content. She also owns the company intranet. This intranet relies on regular manual updates, which often fail. Instead of asking her team for updates, she uses AI to collect, organize, and update content across departments, turning typically outdated systems into living sources of truth. She reduced the time normally required to maintain the system by 95% (completely on her own).
This worked because we were already talking openly and early on about how AI would change work. Programs like the AI Champions Initiative, which asks leaders across all departments to allocate 10% of their time to exploring and building AI-powered workflows, has normalized experimentation and facilitated candid conversations about the evolution of roles.
Patterns deployed at scale
This is an opportunity that companies are missing out on. When leaders avoid redefining roles early on, there is a moment when downsizing feels inevitable. When the team wakes up, hundreds of people have lost their previous jobs and there is no clear plan for what will happen next. At that point, termination becomes a reaction to inaction. That is a failure of leadership, not a result of AI.
Companies that are truly transforming with AI are doing something much more difficult than announcing layoffs. They recognize that work itself is changing and are proactively designing accordingly. They are retraining employees, redeploying them to new roles, and redefining what a “good” job looks like in an AI-enabled environment.
This is not easy, especially at scale. It would be much easier to tell all departments to cut 20% of their staff and “fix it”. Large organizations are optimized for that kind of directive. And when boards demand results in a single quarter, leaders often feel a sense of immediacy and decisiveness to cut staff.
But there are also deeper risks. Layoffs create a downward spiral. As AI continues to improve, with each new wave of capabilities comes another round of layoffs, companies will steadily downsize and become increasingly reliant on technology until there is nothing left to transform. These companies will survive, but they will not evolve. More adaptive organizations become smaller versions of themselves, able to do the same amount of work with fewer people, while expanding scope and output with the same teams.
A groove is already forming
Although we are still in the early stages of this transition, clear disagreements are emerging. On the other hand, some companies are using AI as a justification for cutting staff. On the other side are companies that treat it as a catalyst for reinvention. The difference is whether leaders choose change driven by long-term capability building rather than short-term pressures.
The companies that will successfully navigate this situation will not be those that have never faced disruption. They will learn from it and build structures to deal with the next wave before it arrives.
AI doesn’t just reduce the workforce. Organizations can accomplish many times more when people are given a structure to evolve around. I know that because I had to find that structure myself. And now also because I helped others find it. Let’s give up on the path to change and hope that efficiency will drive us forward. Or you can do something more difficult. I know where the former leads.
The opinions expressed in Fortune.com commentary articles are solely those of the author and do not necessarily reflect the author’s opinions or beliefs. luck.
