One of my favorite exchanges in The Office comes in Season 2, when Jim asks Dwight, “Hey, Dwight. As Sempai, do you think there is ever going to be a day where humans and robots can peacefully co-exist?”
Dwight considers this for roughly zero seconds.
“Impossible.”
I have been thinking about some version of Jim’s question since I was very young—long before I knew enough chemistry, physics, mathematics, or philosophy to model what I was actually thinking about.
My present state of mind might best be called apacoloptimism: not optimism because catastrophe is impossible, but optimism because apocalypse, in its older sense, means unveiling. I wrote recently in “The Veil Is the Enemy, Not the End” that the important thing about an apocalypse may be that something hidden becomes legible.
If morality itself could become sufficiently legible—rendered algorithmically without being flattened into one stupid quantity to maximize, constrained instead by relationship, suffering, intention, agency, and cost—then perhaps we have reason to answer Jim yes.
Perhaps we should even hope for something vaguely like Hari Seldon’s fictional equations: not equations that predict every human choice, but models good enough to reveal the trajectories of large systems.
And perhaps those models would reveal something else, both unsettling and hopeful: humans and robots are not opposites of nature. Our brains and the hardware that instantiates artificial intelligence are different physical systems, but neither escapes matter, energy, constraint, force, motion, or time.
Twenty-one years ago, on my old University of Florida page, I published an essay called “Transition State.” I took a reaction-coordinate diagram—the familiar curve with reactants, products, and an energetically expensive transition state—and used it to think about human psychology.
A few weeks later, in “Shawshank Sunday III: Psychogeolothermodynamics,” I wrote: “Andy is depicted as an isotherm. I am usually an adiabat. Let me explain.”
And in “Brooks vs. Red,” I was already using a compass, a prison, fear, hope, and another human being as models for how people find—or fail to find—a direction.
Apparently I have been doing this for a while.
It can look like the occupational disease of a science nerd: eventually everything becomes chemistry or physics. But I think I have been reaching for something more specific.
I like models as translation.
A model does not claim that two things are identical. Human beings are not ideal gases. Marriage is not a reaction vessel. Morality is not literally a vector field.
A good model preserves relationships while changing vocabulary. It rotates a problem until structure that was hidden from one angle becomes visible from another.
Consider inertia. Human beings tend to continue along established trajectories.
Or friction. Static friction is often greater than kinetic friction. Starting the run can be harder than continuing it. Sometimes changing behavior requires less heroic force and more intelligent alteration of the contact surfaces.
Or potential wells. Habits, identities, relationships, addictions, and ideologies can become attractors. Push the system away, and it rolls back toward the same minimum.
This is where “try harder” begins to look like bad engineering.
Then there are constraints. We usually speak of constraint as the opposite of freedom, but physical systems teach a more complicated lesson. A crystal exists because atoms cannot occupy arbitrary positions. Language works because words cannot be arranged without consequence. Music becomes possible because not every frequency occurs simultaneously.
Constraint does not merely prevent form.
Constraint produces it.
A mind without constraint would not be infinitely creative.
It would be noise.
I now think my later work on moral geometry grew naturally from this same habit of thought. Moral life can be modeled as a field defined by at least three dimensions: relationship, suffering or cost, and intention. Actions move through that space as trajectories. Virtues provide directions. Habits create curvature. History reshapes the landscape.
And under pressure, one question becomes especially revealing:
Where does the cost go?
Do I absorb some cost myself to protect relationship, truth, or another person’s flourishing? Or do I use power, rationalization, bureaucracy, or distance to export that cost onto someone else?
That distinction gives moral direction something like an axis. One orientation bends toward generativity, restraint, and self-emptying. The other bends toward domination, insulation, and displaced suffering.
Physics and geometry do not sit above morality as judges. They enlarge its descriptive machinery. They give us a richer library for tracing relationships, extending consequences across a field, and illuminating structures that ordinary moral vocabulary tends to flatten.
They do not replace moral judgment.
They increase its resolution.
That may be the common thread running through decades of my writing: translation, relation, constraint, and differentiation. I keep taking structures from one domain and rotating them to see whether they illuminate another.
The useful models survive the rotation.
The decorative ones fall apart.
We are collections of atoms obeying physical law.
But some collections of atoms do something peculiar: they construct models of the forces acting upon them, models of other minds, and eventually models of the consequences of their own actions.
That does not free us from causality. Knowledge and power are what help define moral agency.
Perhaps freedom is something subtler: the emergence, within causality, of a system capable of representing its own constraints and using that representation to alter its trajectory.
And perhaps morality begins one step beyond that:
when the system learns that its trajectory is not the only one in the field.
Related NSEFL posts
- The Veil Is the Enemy, Not the End — apocalypse as unveiling; moral progress as increasing legibility rather than merely avoiding catastrophe.
- Toward a Moral Geometry — the most direct introduction to field, vector, curvature, and moral orientation.
- Moral Geometry Maps Possible Worlds — choices as boundary conditions that change which futures remain reachable.
- The Self as Apex Sensor — develops the question that appears here as “Where does the cost go?”
- Cost, Coherence, and Rotational Literacy — the strongest companion piece on models themselves: rotation, constraint, friction, and writing as a method for preserving perception.
