Interfaces for intention

For much of the modern era, many domains tied expertise to their machinery.

To make meaningful contributions, you had to navigate layers of accumulated friction: the workflows of finance, the toolchains of engineering, the protocols of healthcare, the planning systems of logistics. Using the system came before fully grasping the abstraction. The interface acted as a gate.

AI shifts that balance.

It doesn’t eliminate mastery, but it moves the bottleneck. As the cost of turning intent into execution tends toward zero, the interface stops absorbing most of the difficulty; intention begins to matter more.

When procedural fluency is no longer the primary constraint, the scarce part of expertise becomes the ability to articulate what should happen, under which constraints, and with what structural awareness. Skill gradually becomes less about technique and more about abstraction.

This raises a natural question: if expertise becomes intention shaped by constraints, what interface allows that to be expressed?

The first temptation is to assume chat becomes the universal interface. Chat was the first expressive surface AI unlocked. And chat is excellent for exploration: it lowers inhibition, widens the search space, and makes complex systems feel approachable. But chat is poor at governance; it captures desire but not the boundaries that give intention its form. It has breadth but lacks structure.

A world where intent moves in lockstep with execution requires a surface that makes structure visible; where constraints are explicit, flows are legible, and consequences can be understood before anything moves. A surface that reveals the architecture of what you’re trying to do, and accelerates the feedback loops between intention and outcome, before execution is locked.

If such a class of interfaces emerges, something downstream begins to shift. The borders of domains open up. When machinery is no longer the barrier, the transferable part of expertise becomes a person’s mental models: the patterns they recognize, the flows they can map, the constraints they can design. The penalty for stepping into adjacent fields drops, and recombination accelerates. Disciplines begin to overlap because their primitives can mix with less resistance.

We’ve seen this pattern before. The invention of the spreadsheet in the 1980s collapsed the machinery of finance and planning into a surface where structure became legible and malleable. People who had never touched a mainframe could suddenly model businesses, run scenarios, and reason in ways that had previously been locked behind specialized tooling. Marketers built forecasts without analysts. Operators ran capacity plans without dedicated systems. Founders did their own finance. Product teams modeled growth loops. The result was cross-pollination. Reasoning developed in one corner of an organization could now travel to another. Domains didn’t disappear, but their borders thinned. And the creators of that interface captured a meaningful share of the value unlocked.

All this suggests that one of the frontiers ahead of us lies in creating the interfaces that make this new kind of reasoning possible; the meta-interfaces that allow systems, people, and disciplines to combine in ways the old machinery never permitted.

Whoever builds these surfaces will not only unlock the value of recombination, but capture a part of it.

The second derivative of conflict resolution

I wrote this after noticing a pattern in how good teams and good relationships evolve. It’s not that they avoid conflict, but rather they metabolize it faster every time. The model that emerged was a mathematical one; relationships as learning systems, their health measurable by the slope of repair.

One of the beliefs I hold most firmly is that the best predictor of success in any relationship, whether romantic, friendship, or team, is the second derivative of conflict resolution.

By conflict, I don’t mean shouting or drama. I mean any point where expectations diverge and two internal models of reality collide.

A great relationship is not one without friction; it’s one where friction resolves faster and cleaner over time. The first time you face conflict, it takes a day to recover. The second, six hours. The third, ninety minutes. The fourth, twenty. After that, the curve asymptotes toward zero.

That curve, the rate at which repair accelerates, is what I call the second derivative of conflict resolution (SDCF). It measures not harmony, but learning. Every disagreement, once resolved, adds a building block to shared understanding, which means you don’t have to fight the same fight twice.

This reframes relationship quality from being about harmony to being about adaptive efficiency. The first derivative of conflict resolution shows how quickly a single conflict resolves (i.e. the velocity of recovery). The second derivative shows how that velocity improves over time (i.e. whether the system learns). In simpler terms, what matters isn’t how fast you repair once, but how fast you get better at repairing.

If over successive conflicts the first derivative (recovery speed) becomes more negative more quickly, meaning repair happens faster each time, then the second derivative across conflicts is positive in the direction of learning. Conversely, when the second derivative flattens or turns negative (i.e.when conflicts take just as long, or longer, to resolve) it’s a sign of structural incompatibility. The system isn’t learning. What looks like “communication problems” is really the absence of adaptation.

Most people assess relationships based on emotional tenor; how good they feel or how frequently they argue. But the SDCF model suggests something different; conflict isn’t a sign of failure, but rather it is signal. Each disagreement surfaces new data about boundaries, needs, and blind spots.

In that sense, the counterintuitive truth is that the path to relational strength runs straight through conflict.

Every repair is a form of learning; every argument, a test of how well two people can turn friction into shared understanding. What ultimately defines longevity is how efficiently that learning compounds, and how each conflict leaves the system slightly more aligned than before.

What we often call being “well-matched” is really just phase alignment under low stress. A relationship that truly compounds is one where both people elevate each other through conflict.

Common sense suggests compatible people should recover faster, but the inverse is also true; people who recover faster become more compatible. The variable you can actually control is the learning rate; the slope of repair.

It’s worth highlighting that awareness itself changes the shape of the curve. Most relationships operate unconsciously along their derivative, unaware of whether repair is accelerating or stalling. But once you can see the curve, you can influence it. Awareness reigns in entropy, and replaces drift with structure.

That awareness can have two outcomes, both good. It can either help a relationship move to a higher level of coherence, or reveal that the system has reached its limit, that its slope will never meaningfully improve, and thus allow it to end cleanly. Both outcomes are infinitely better than unconscious decay.

This lens changes how you think about relational “success.” It’s not about avoiding arguments or achieving constant peace. It’s about whether repair gets faster and deeper each time. Whether the feedback loop between conflict and understanding tightens. Whether the relationship compounds.

It also applies beyond the personal. Teams, partnerships, and organizations all have a SDCF. The best companies aren’t those without disagreement but those whose disagreement resolution curve steepens with time, as they learn to metabolize tension into clarity.

A team’s greatness isn’t its lack of internal debate, but how fast it integrates disagreement into improved operating norms. Cultures that avoid conflict decay, while cultures that metabolize it evolve.

If you believe this, then conflict stops being something to fear. It becomes diagnostic. You run toward it, because every repair is a data point on the curve. A chance to move the derivative in the right direction.

That, to me, is what separates fragile from enduring systems, whether personal or collective. It’s not how they avoid stress, but how quickly and gracefully they repair after it. The rest is just noise.

Coase in the age of code

I recently came across an essay I wrote in university in 2011, about Ronald Coase and the theory of the firm. It was dry, academic, and deeply curious about a question that still feels pertinent today; why do firms exist at all?

Back then, the answer felt settled. Coase had explained that firms emerge because organizing through the market is costly; contracts take time, negotiations add friction, and information is imperfect. When it becomes cheaper to manage people internally than to transact externally, a firm is born. The invisible boundary of the firm, he said, lies at the point where these two costs meet.

Reading that old paper now, after a decade spent around blockchains, DAOs, and “trustless” financial systems, I’m struck by how cyclical the question feels. Crypto’s grand promise was to eliminate the very frictions that gave birth to the firm; to replace bureaucracy with code, management with incentives, contracts with consensus.

If Coase’s firm existed to minimize transaction costs, and those costs could be automated away, then perhaps the firm itself could be dissolved.

That was the dream. But it didn’t happen.

When you replace contracts with smart contracts, you still need judgment: what counts as a valid state, what to upgrade, when to fork. When you remove hierarchy, you rediscover governance, only now it’s slower, noisier, and happening in public.

Bounded rationality didn’t vanish with blockchains. It simply migrated to Discord. The same cognitive limits that once defined the borders of the firm now define the borders of the network.

Agency problems persist too. Token holders delegate to committees, multisigs, or core teams. Power concentrates. Decision-making slows. Coordination becomes its own cost center. Every “decentralized” organization ends up rebuilding a managerial layer; sometimes reluctantly, sometimes accidentally.

The irony is that Coase’s logic still applies: a network expands until the cost of coordinating one more decision exceeds the cost of spinning up a new one.

Coase described the firm as an economic structure. But over time it became something deeper: a social technology for minimizing collective error. Firms exist not just to cut transaction costs, but to give a group of humans a shared model of the world, a rhythm, a sense of who decides what.

Even if code can settle value instantly, humans still need slower systems for context, accountability, and trust. The firm endures because it optimizes for judgment, not just execution.

This is why most DAOs still look suspiciously like companies. They may route capital through tokens, but their structure—small cores, delegated authority, decision bottlenecks—echoes the same patterns Coase was describing in 1937. It turns out coordination is a harder problem than trust.

What has changed is where the boundary lies.

The minimum viable firm used to require offices, payroll, and legal scaffolding. Now it can exist as a wallet, a few passkeys, and a group chat. The cost of coordination has collapsed — not to zero, but low enough that the firm can shrink to its essence: a system for allocating capital and attention toward a shared goal.

That collapse in cost doesn’t end the firm; it atomizes it. The future looks less like one monolithic organization and more like a mesh of smaller, temporary ones.

In that sense, blockchains didn’t abolish Coase’s world; they made his boundary dynamic.

Coase saw transaction costs as economic. What he couldn’t see from 1930s London was that information processing itself would one day be the scarce resource. The true constraint on coordination is no longer contract enforcement, but rather comprehension.

A modern firm isn’t just a bundle of contracts; it’s a bundle of cognition. Its size is limited not by the cost of managing people, but by the bandwidth of shared understanding among them.

Technology keeps lowering the cost of transaction, but it doesn’t raise the ceiling of comprehension. We can move money instantly, but aligning meaning still takes time. That’s the paradox of the digital firm: infinite speed, finite sensemaking.

When I wrote that early essay, I thought of the firm as an object, a structure bounded by cost. Now I think of it as a living organism bounded by cognition. The question is no longer why firms exist, but how fluid they can become before they stop being coherent.

Coase explained why we built firms. Crypto reminded us why we still need them.