AI Is Coming for Employment, Not Workers

In late December, TechCrunch surveyed 24 enterprise venture capitalists about their AI predictions for 2026. The survey didn't ask about labor. Multiple investors raised it anyway, unprompted.

A companion piece followed days later. It cited MIT's estimate that 11.7% of jobs are already automatable using current AI, and assembled a chorus of investor predictions about how aggressively that displacement will unfold over the next twelve months.

Two of those investors capture the conversation from different angles. Jason Mendel of Battery Ventures called 2026 "the year of agents as software expands from making humans more productive to automating work itself, delivering on the human-labor displacement value proposition." Antonia Dean of Black Operator Ventures took a different tack: "AI will become the scapegoat for executives looking to cover for past mistakes."

One frames displacement as a technology story. The other frames it as a corporate behavior story. They look like different arguments. But they rest on exactly the same foundational assumption.

The shared assumption is more foundational than the question of pace. It is that an employment-based economy is the only arrangement available, that whatever happens with AI in 2026 will play out within the structure of employers, jobs, and workers as we know it today.

The conversation can debate how aggressively displacement will arrive, which roles will go first, how quickly budgets will shift from labor to AI, precisely because nobody has stopped to ask whether the economy on the other side of all of that still has to be organized around employment at all.

We are debating whether we are watching the system die. We are not yet wondering how to rebirth it.

The way to begin that exercise is to look at what AI is actually displacing. The current economy runs on a particular kind of arrangement: employment built on fungibility.

An employer specifies a role, finds someone with the basic skills to fill it, and treats every candidate who can fill it as essentially interchangeable with the next. The arrangement works because roles are specified in advance, in writing, in a form precise enough that candidates can be evaluated against the spec.

AI breaks the arrangement at exactly that seam. If you can specify a role precisely enough to post it on a job board, you can specify it precisely enough to hand it to an agent.

What AI is actually displacing isn't workers. It is fungibility-based employment itself: the system that turns most labor into most income in the economy today.

Fungibility-based employment is so familiar that it reads as nature. It isn't.

Job descriptions, preferred qualifications, the standardized interview, the language of "transferable skills," the org chart of the modern firm: all of these were engineered, deliberately, over the course of the last century, to make labor interchangeable enough to plug into pre-defined slots. The bureaucratic corporation exists in roughly its current shape precisely because it can perform that integration at scale.

The fungibility AI now threatens is a design feature of this particular economic system. A different economic system, one not built on the requirement that labor be fungible, is also possible. It just requires deliberate attention to recognize the option, because the current arrangement is so dominant that it disappears into the background.

If we take the displacement seriously, and accept that fungibility-based employment is on its way out, three things can happen on the other side.

• Revolt: workers refuse to absorb the displacement, and political and labor upheaval forces redistribution by force rather than by design.
• Starvation: displacement arrives faster than any structural offset, consumption-side economics breaks, and a sufficient mass of people who can't eat eventually revolt anyway.
• Restructure: we deliberately redesign the relationship between labor and income, before the system makes the choice for us.

Universal Basic Income is the most well-known restructure proposal in circulation. The premise is straightforward: decouple income from labor entirely, and distribute resources to people regardless of whether they hold a job.

The case is usually argued for, and against, as a question of charity. That framing misses the point.

The argument for UBI is macroeconomic, not moral. The economy already produces more than its consumers can absorb, and the only mechanism currently closing that gap is wage income. Strip out wages without replacing them, and the consumption vacuum kills the producers along with the workers.

UBI, in this reading, is a deliberate consumer-side redistribution ensuring producers have a market to sell into. That makes it worth taking seriously as a transition mechanism. But it isn't the only available restructure, and it isn't the end of the story.

A less-debated restructure runs in a different direction: privileging uniqueness over fungibility. The contribution model would invert.

Instead of an employer specifying a role and finding someone to fill it, each person would identify what is uniquely theirs to offer (a perspective, an expertise, an identity) and find organizations or other individuals who value it enough to pay for it.

In practice this would look like more solopreneurs running AI as their workforce, networks of contracted expertise instead of fungible-headcount hierarchies, and eventually contribution-based mechanisms: paid contributions to shared foundation models, peer-distributed income, structures we have not yet named.

This is not a near-term reorg. It is a multi-generational shift in how labor and income relate to each other.

UBI and a perspective-based contribution economy are not mutually exclusive. UBI stabilizes the transition. A perspective-based economy becomes possible because of that stability.

None of this is inevitable, and the path is long. That is exactly why deliberate planning is urgent rather than optional.

Without a stated end goal (a UBI-style decoupling of income from labor, a perspective-based contribution economy, both, or something else entirely), the displacement curve will outrun the redesign, and we will arrive at starvation or revolt by default.

We are closer to a more humane economic system than we have ever been. But the only way to get there is to manage the transition slowly and deliberately.

The biggest risk is not that AI's impact on workers will be dangerous. It is that the stress of the transitional period becomes too great, and we never arrive at the new economy AI can enable.

In the immediate term, organizations will use AI to absorb fungible roles. That move is logical, predictable, and probably already happening in your company.

The problem isn't the move itself. The problem is what is left over after it: a recruitment model and a labor-engagement model that no longer match the work that is actually getting done.

Recruitment has to shift from "preferred qualifications" (the fungibility plug-in) to "what perspective are we adding to this room." Labor engagement has to shift from full-time employees grinding scope-specific roles to expert consultants engaged at fractional time, brought in for what they uniquely offer. Organizational shape has to shift from interchangeable hierarchy to networks of expertise.

Further out, organizations themselves will become more loosely and temporarily constituted, not the permanent-staff entities we are used to today.

Human labor doesn't disappear from organizations. It stops being defined by job postings, and organizations stop being defined by their permanent headcount.

The VCs in the TechCrunch survey are predicting an outcome: fewer jobs, more agents. That outcome may well arrive.

The work of leadership this decade is making sure it isn't the destination. Only a midpoint, on a longer journey to a fundamentally different economy.