Human expertise is becoming a commodity

AI, Labor Compression, and What We Are Not Prepared For

I saw four marketing leadership job postings in a single afternoon.

Not identical postings. Different companies. Different industries. Different growth stages. Some were thoughtful and refreshingly honest about the tradeoffs of operating lean. A couple of jobs were probably good for the right person.

But there was a new thread running through all of them that I could not ignore once I saw it.

The expectation that one senior marketer, supported by AI tools, could now function as an entire marketing department. Not lead one. Replace one.

One role wanted demand generation, positioning, lifecycle, events, analytics, sales enablement, and pipeline ownership tied directly to revenue targets. Another wanted brand strategy, content, creative direction, video production, channel management, and conversion economics with little internal support. Several explicitly described AI systems and agents as part of the operating model that made this scale of responsibility workable. One company openly stated that they prefer systems and agents before adding headcount. Another required candidates to be “AI-native.” No one is “AI-native”.

And one of the most interesting contradictions sat at the center of it all: a company demanding high-volume execution from lean teams while explicitly warning that AI-generated submissions would be automatically rejected.

Use AI. Scale with AI. Move faster with AI. Operate leaner with AI. But still produce unmistakably human work.

I have spent more than two decades in marketing leadership roles. I use AI every day. I am not threatened by the technology itself. Some of the tools are genuinely extraordinary. They help organize information faster, accelerate production, reduce friction, surface patterns, and improve operational efficiency.

But what I began to realize as I read through these postings is that many companies are no longer treating AI as support infrastructure. They are treating it as justification. Justification to lean out teams, compress multiple specialized disciplines into fewer people, expect executive-level strategy and hands-on execution from the same exhausted human being, and present labor reduction as innovation.

Because AI does not just promise efficiency. It signals it. And for a lot of organizations right now, the signal appears to matter more than the proof.

One of the most consistent patterns in my career has been watching companies say marketing is strategic while treating it operationally.

Marketing is expected to drive growth, but often without access to the conversations that actually shape growth. Product decisions get made elsewhere. Revenue assumptions get made elsewhere. Pricing gets decided elsewhere. Headcount decisions get made elsewhere. Then marketing gets handed a number and asked to generate demand against conditions it did not help create.

I have sat in meetings where marketing was expected to explain pipeline shortfalls without visibility into sales conversion issues, product readiness problems, or shifting executive priorities that changed the actual business conditions underneath the forecast. And when pressure hits, marketing is often one of the first places companies look to rationalize cost.

Part of that is structural. Marketing is highly visible but difficult to measure cleanly. Everyone sees the outputs: campaigns, ads, emails, social posts, webinars, landing pages, events. But much of the real work happens upstream of execution, in places that are harder to quantify and easier to underestimate: positioning, narrative clarity, customer understanding, market timing, pattern recognition, judgment, and knowing which problem actually needs solving.

That work is mostly invisible until it is absent.

I think that is part of why marketing has always had a slightly unstable relationship with legitimacy inside organizations. Even now, after decades of digital transformation and revenue accountability, marketing still gets flattened into stereotypes remarkably quickly. Make it pretty. Generate leads. Support sales. Run campaigns. At the same time, modern marketing organizations are expected to own increasingly complex systems: demand generation, lifecycle, analytics, customer journeys, product marketing, attribution, sales enablement, brand strategy, conversion optimization, market research, retention, and growth.

It is both highly strategic and constantly at risk of being treated like execution-only labor.

That tension existed long before AI entered the conversation. Which is part of why the current shift feels so significant. AI is entering a profession that has already spent decades trying to prove that expertise, judgment, and strategic thinking are not interchangeable with content production.

None of this started with AI.

Long before ChatGPT, corporate America was already deeply conditioned around efficiency language: lean teams, resource optimization, operational discipline, doing more with less, outsourcing, automation, productivity gains, margin improvement, shareholder value. Every economic downturn accelerated it. Every new technology wave justified another version of it.

Sometimes those shifts were necessary. Some companies absolutely were bloated. Some systems needed modernization. But there is a difference between using technology to improve work and using technology to rationalize permanently shrinking the humans doing it. That distinction matters.

After 2008 especially, the operating model inside many companies changed. Teams became leaner by default. Headcount became harder to justify. Entire departments learned to function permanently understaffed because temporary austerity slowly hardened into normal operating practice. People absorbed the work of coworkers who had been laid off. Teams became smaller while expectations continued rising. The reward for becoming more efficient was often not relief, but more responsibility.

AI entered an environment where those pressures already existed.

That is part of why the current moment feels so volatile. AI is not arriving inside stable organizational structures. It is arriving inside systems already optimized aggressively around labor efficiency and investor expectations. And unlike earlier technology shifts, AI creates a particularly seductive narrative for executives because the outputs are so visible. A chatbot produces copy instantly. A workflow automates repetitive tasks. A system summarizes research in seconds. A small team suddenly ships more work faster.

That visibility makes the leap from leverage to replacement feel deceptively logical. If one person with AI can produce twice as much output, companies naturally start asking whether they still need the same number of people. But output is not the same thing as capability. And that is the part many organizations are moving too quickly to fully understand.

The deepest layers of expertise inside companies are often the least visible ones — judgment, discernment, pattern recognition, political navigation, contextual understanding, the ability to recognize when a problem is actually upstream of the place everyone is trying to solve it. Those things are much harder to quantify on a spreadsheet than labor reduction. Which is part of why AI currently functions as more than a technology story. It is also a financial story. AI signals efficiency, modernization, scalability, future readiness, and margin expansion. In public markets, especially, those signals matter enormously.

The danger is that companies start restructuring themselves around the appearance of future capability before the technology has actually replaced the human expertise it is removing.

The data is starting to reflect this. A May 2026 Gartner survey of 350 global enterprises already deploying AI found that 80 percent had reduced headcount tied to their AI initiatives, some by as much as 20 percent. The companies that cut the most showed financial returns nearly identical to those of the companies that cut the least. In several cases, the ones that cut less performed better. Gartner’s conclusion was direct: workforce reductions may create budget room, but they do not create return. The assumption driving much of this restructuring—fewer people equals better margins, which equals better performance—is not supported by the evidence currently available.

One of the things I remember most from my last role was watching the same idea land differently in executive meetings depending on what pressure the organization was under that week. The mission mattered deeply to people there until revenue pressure, operational stress, or risk tolerance changed the temperature in the room.

Before I left, I spent months helping shape a brand platform called Your Money is a Movement.

The line itself was already there. That was the point. It had been sitting inside the organization, mostly ignored, treated like language instead of strategy. When I found it, I recognized the spark in it, not because AI surfaced a phrase, but because years of brand work told me there was something there worth building around.

That spark became a manifesto. Bold, brash, and more distinctive than what mission-driven organizations often allow themselves to say out loud. The kind of positioning that can help a mission-driven organization stop sounding like everyone else and start becoming recognizable.

On paper, that sounds like messaging work. Campaign work. But the actual work had very little to do with writing a tagline.

AI helped with parts of the process. It helped organize research, synthesize interviews, surface themes faster, analyze transcripts, and categorize patterns. But AI did not recognize what made the company emotionally distinct. AI did not sit in executive meetings and recognize where leadership was aligned, where it was hesitant, where tolerance shifted depending on financial pressure, or where political realities inside the organization would shape what could actually survive implementation. AI did not determine whether the positioning felt authentic or merely aspirational. It did not recognize which language rang true to employees and members and which language sounded like marketing theater.

And AI certainly did not persuade leadership to believe in it.

Recognizing the signal inside the noise is human work. Understanding when an organization is trying to sound like something instead of become something is human work. Knowing which ideas can survive contact with reality inside a company is human work.

And once the positioning existed, the real work started: translating it across channels, adjusting it for different audiences, balancing brand aspiration against compliance realities, operational constraints, budget pressure, shifting leadership priorities, and actual execution capacity. None of that looked like the fantasy version of AI-powered marketing that keeps showing up in job descriptions right now. It looked like judgment. It looked like discernment. It looked like pattern recognition built over decades.

The problem is that the work no one can see is exactly the work that does not show up in a job description.

One of the stranger contradictions in these job postings is how much they still demand unmistakably human work while treating human labor as increasingly compressible.

One company rejected AI-generated submissions while describing an operating model built around lean execution, systems, and efficiency. Another emphasized human-to-human marketing while requiring candidates to build AI agents and workflows that would help a small team scale harder before adding headcount. What companies increasingly seem to want is not less human expertise. They want the same expertise spread across fewer people.

I understand why this is happening. If you are a founder or executive under pressure to grow efficiently, AI looks extraordinary. A small team suddenly appears capable of producing more output. Functions that once required specialists now look, at least on the surface, manageable by generalists supported by tools. From the outside, it can look irrational not to compress the model.

But living inside that model is different from admiring it on an org chart.

One exhausted human plus AI is not a marketing organization. It is one person context-switching between strategy, analytics, content, positioning, operations, reporting, creative direction, stakeholder management, vendor coordination, sales support, campaign execution, and performance accountability while trying to preserve enough mental clarity to still produce thoughtful work.

Many of the people taking these jobs already know that. They are not naive. The market is tight. Experienced people need work. Many genuinely love building things. Some are attracted to the autonomy and range. Some are burned out by bureaucracy and large organizations. Some will succeed in the right environment for a period of time.

But there is a difference between carrying a broad role in an early-stage company and normalizing structurally unsustainable expectations across an entire profession. That is the line we are starting to cross.

The cost of these models usually gets absorbed by the worker long before it becomes visible to the company. The exhaustion, the cognitive overload, the inability to think strategically because execution never stops, the erosion of specialization, the pressure to constantly prove value because the role was built around efficiency assumptions from the beginning. And underneath all of it is the implication many experienced people feel instinctively, even if they do not always say it out loud: if AI supposedly makes one person capable of doing the work of many, the expertise itself becomes easier to undervalue.

Over the last few years, I have watched companies publicly celebrate their people while privately reorganizing themselves around needing fewer of them. Not just in marketing.

Across functions, the same pattern is starting to appear. Customer support teams reduced because AI chat systems can absorb volume. Recruiters replaced with automated screening workflows. Writers pushed to produce more output at higher velocity. Engineers expected to accelerate delivery with AI-assisted coding. Entire operational teams compressed because leadership believes AI can close the gap. And to be clear, some of these tools genuinely do improve productivity. That is what makes this moment complicated. Because there are real efficiency gains happening alongside a level of labor rationalization that many companies are moving through much faster—and much more aggressively—than they fully understand.

What strikes me most is how differently these decisions are framed when AI is attached to them. For decades, aggressive labor reduction was often viewed critically. Companies risked looking shortsighted, extractive, or overly focused on quarterly performance at the expense of employees and long-term stability. AI changes the narrative. Now the same decisions can be framed as innovation, transformation, modernization, or future readiness. The language becomes optimistic even when the underlying behavior is fundamentally about reducing labor costs and increasing output expectations for the people who remain.

That shift matters because companies reveal what they actually value through the tradeoffs they make under pressure, not the language they use when things are easy.

Almost every company today talks about people, culture, purpose, wellbeing, authenticity, and human-centered leadership. I do not think all of that language is hollow. I have worked with genuinely thoughtful leaders who care deeply about their employees and customers. But an uncomfortable gap is opening between what companies say about people and what many organizations appear increasingly willing to normalize operationally, using AI to increase efficiency, reduce dependence on labor, avoid hiring, justify leaner teams, and push more responsibility onto fewer people.

At some point, companies have to confront the reality that workers notice the contradiction. Customers do too. Because this is ultimately bigger than marketing, or even AI itself. It is about whether organizations still view human expertise as something worth investing in long term, or whether people are increasingly being treated as temporary operational overhead that technology will eventually minimize.

I suspect this is partially a bubble and partially a permanent restructuring. I think companies are overestimating what AI can currently replace while underestimating what human expertise still does. But I also think some jobs are simply not coming back. And we are nowhere near emotionally, economically, or politically prepared for what that means.

I do not think AI is going away. I do not think companies should ignore it, resist it blindly, or pretend it will not reshape entire categories of work. Some of the productivity gains are real. Some organizations will absolutely become smarter, faster, and more effective because of these tools.

But there is a meaningful difference between using AI to support human expertise and using it to justify eroding the conditions that expertise actually requires to exist.

Expertise does not appear instantly. Judgment does not appear instantly. Discernment, taste, pattern recognition, strategic thinking, customer understanding, organizational navigation, and creative intuition are all built slowly, through years of experience, mistakes, observation, context, and accumulated knowledge. Those capabilities are not interchangeable with output. And the current conversation around AI treats them as if they are.

That is the part that concerns me most. Not the technology itself. The possibility that companies become so focused on maximizing short-term efficiency gains that they begin hollowing out the human depth that made their organizations valuable in the first place. Especially because many of the capabilities organizations now seem most eager to compress are the ones hardest to rebuild once they disappear, like experienced operators, institutional memory, strategic judgment, creative maturity, and the ability to recognize problems before they become visible in a dashboard.

Those things do not scale as neatly as software. They also do not fit easily into quarterly efficiency narratives. But they matter.

And I suspect the companies that navigate this era most successfully will not be the ones that remove human expertise fastest. They will be the ones disciplined enough to understand where human judgment still matters most.

Because AI may absolutely reshape work. But if companies continue mistaking leverage for replacement, they will not build smarter organizations.

They will build thinner ones.