Three trajectories between optimization and meaning
The conversation about artificial intelligence has become strangely bipolar.
AI is either the greatest force for human progress we have ever unleashed, or the silent architect of our own obsolescence. It will cure disease, end scarcity, unlock human potential—or it will take our jobs, hollow out meaning, and, in the most dramatic versions, optimize us out of existence like an inconvenient variable.
Both sides are compelling. Both are emotionally satisfying. And both miss something essential.
They talk about outcomes, but not about structure.
They argue about what AI will do to us, while paying far less attention to how it quietly reshapes the way we divide work, the way we form organizations, and the way humans coordinate with each other when complexity explodes.
History suggests that technologies rarely destroy societies outright. Instead, they destabilize them—by stretching existing organizing principles beyond their limits. And when that happens, new forms emerge.
This essay is not about utopia or dystopia. It is about three trajectories—three ways the world could reorganize work and coordination under the pressure of AI. All three are plausible. All three are already visible in fragments today. And none of them exist in isolation.
They are not just scenarios. They are swings of a pendulum.
Three lenses before the future unfolds
Before looking forward, we need to ground ourselves in three ideas that quietly determine almost everything about economic life.
The division of labor: our oldest scaling trick
Once upon a time, a person made an entire object from start to finish. Then we discovered something simple but profound: if we split work into parts, output skyrocketed.
Adam Smith’s needle factory remains the canonical example. One person, working alone, could make a handful of needles a week. Divide the process into steps—drawing wire, cutting, sharpening, assembling—and suddenly thousands appear. Civilization, in many ways, is built on that insight.
But division has a cost. Specialization trades context for efficiency. The more we divide, the less each part understands the whole. Coordination becomes a problem that must be solved—first informally, then managerially, then bureaucratically.
AI does not invent this dynamic. It amplifies it.
The theory of the firm: where the borders lie
Why do companies exist at all? Why not contract everything through the market?
The classic answer is: transaction costs. Hiring, managing, aligning, trusting—none of that is free. Firms form where internal coordination is cheaper or safer than constant contracting.
AI bends this logic. Suddenly, coordinating external work becomes easier. So does automating internal work. The borders of the firm soften, blur, and become negotiable again. The question is no longer “Should we outsource?” but “What must remain coherent?”
The collective: a synthesis, not a replacement
Here is where something new enters.
When division of work accelerates and firm boundaries dissolve, a third form begins to materialize: the collective. Not a company in the traditional sense. Not a loose market either. And not a medieval guild—though the resemblance is not accidental.
A collective is defined less by contracts and hierarchies, and more by shared purpose, shared standards, and shared learning. It thrives on mixing disciplines that were never meant to sit next to each other: engineers and artists, philosophers and product designers, craftspeople and AI systems.
Its core advantage is not efficiency, but emergence—the idea that something new appears when different forms of intelligence collide.
With these three lenses in hand, the future begins to split—not into one path, but three trajectories.
Trajectory One: The triumph—and collapse—of hyper‑division
In the first trajectory, we do what we always do when something works exceptionally well: we push it too far.
AI supercharges the division of work. Tasks are broken down into ever smaller units. Knowledge work is decomposed into prompts, tickets, micro‑decisions. Entire organizations become flowing streams of optimized fragments.
At first, the results are intoxicating.
Productivity rises. Errors drop. Costs fall. Everything becomes measurable. AI systems assign, monitor, redirect, and optimize continuously. Work moves faster than humans ever could.
And then something subtle begins to fray.
People lose sight of the whole. Knowledge becomes local and brittle. Coordination overhead grows faster than output. New layers of management emerge—not to create value, but to realign pieces that no longer fit naturally.
AI, ironically, is tasked with managing the complexity it helped create.
This is where the old thought experiment returns: the paperclip maximizer. Not as a literal apocalypse, but as a metaphor. Systems become extremely good at optimizing proxies while drifting further from purpose. They maximize output, efficiency, or engagement—until those metrics destroy the thing they were meant to serve.
In this trajectory, the theory of the firm oscillates wildly. Everything is outsourced—until trust breaks. Then everything is pulled back inside—until rigidity sets in. The organization becomes restless, fragmented, and strangely hollow.
People struggle not because work is hard, but because work no longer coheres.
This is the dark side of intelligence without wisdom.
Trajectory Two: The long search for balance
The second trajectory begins when cracks become impossible to ignore.
Leaders notice that beyond a certain point, more specialization produces less value. That coordination costs eat efficiency gains. That innovation slows when nobody understands the full system anymore.
So organizations begin to experiment.
Division of work loosens. Roles become broader, elastic. Teams are reorganized around outcomes rather than functions. AI is still everywhere—but now it is used to connect, not just to optimize. It summarizes context, reveals dependencies, and helps humans reason across silos.
The firm, in this world, becomes elastic. It keeps a strong core—identity, trust, accountability—while building long‑term partnerships rather than transactional outsourcing chains.
And the collective begins to mature.
Philosophers are no longer invited just to ethics panels—they sit inside product teams. Artists shape industrial design. Craftspeople and AI engineers collaborate, each expanding the other’s range.
This world is uneven. Some organizations adapt. Others cling to old structures until they fracture. The result is not harmony, but learning.
This trajectory is uncomfortable, slow, and deeply human.
Trajectory Three: The collective becomes the unit of organization
In the third trajectory, ideas once dismissed as hype quietly return—older, humbler, and more useful.
The metaverse reappears, not as escapist fantasy, but as shared cognitive space: digital twins, spatial collaboration, environments where complex systems can be seen and shaped together.
Web3 sheds ideology and becomes governance tooling. DAOs are no longer libertarian manifestos, but practical mechanisms for distributed decision‑making and incentive alignment.
Tokenization matures into infrastructure: faster settlement, transparent ownership, programmable value flows—especially in a world where traditional monetary stability can no longer be taken for granted.
The firm, here, is no longer a hard boundary. It is a constellation: a small core of accountability, surrounded by stable partners, communities of contributors, and AI agents acting as first‑class participants.
Work is no longer defined by jobs, but by missions. Careers become portfolios, but not atomized gig work—anchored instead in collectives with memory, standards, and culture.
This is not utopia. It is simply a structure better suited to complexity.
The pendulum always swings
If there is a single belief underlying this essay, it is this:
Systems seek balance, but only after exploring extremes.
Economically. Ecologically. Politically. Organizationally.
When division of work goes too far, reintegration follows. When centralization chokes adaptability, decentralization erupts. When chaos spreads, new forms of governance emerge.
AI is not the end of this story. It is the force that accelerates the swing.
The question, then, is not whether we will overshoot—we will.
The question is whether we notice the direction of the pendulum, and learn fast enough to shape what comes next.
Somewhere beyond firms, beyond markets, and beyond outdated binaries, the collective waits—not as a destination, but as an evolving answer to the oldest problem of all:
How humans coordinate meaningfully at scale.
What's on your mind?