Most Likely,
Least Likely
The familiar carries us.
The unexpected asks where we are going.
What it makes possible determines its value.
The distinction
The familiar works.
The departure matters.
Most likely gives us an outcome we recognize. Least likely gives us an outcome that changes what we recognize as possible.
I am using these terms to look at the artifact itself: the design, the answer, the piece of music, the solution that appears. This is a distinction about the character of an outcome. We are not calculating the odds of a particular object coming into existence.
A more polished version of a familiar design can be excellent. It can work better, last longer, and make someone’s day easier. Predictability has value. We depend on a great deal of the world doing what it has done before.
But something different happens when a design changes the question. The familiar problem becomes visible from another direction. An assumption that seemed inseparable from the object turns out to have been optional.
That is the territory I am calling least likely. Its significance comes from what the departure makes possible. Mere strangeness is easy. A meaningful departure has to hold together.
A different intelligence
Begin with its nature.
I prefer the phrase alien intelligence because it interrupts the expectation that AI is a synthetic version of us. Different intelligence can have different strengths. We lose sight of those strengths when we expect the machine to approach every problem as a human would.
The useful starting point is this: prediction leans toward a continuation supported by the patterns already available. Ask for a design without challenging its familiar assumptions and a familiar design is an entirely reasonable result. The machine has fulfilled the request as it understood it.
Instructions, context, and how we use a model change its outcomes. We can challenge the familiar assumptions and give the work a different direction.
Increasing variation gives us more to examine. Making an output less predictable does not, by itself, give that output a purpose. We still have the work of recognizing what is worth pursuing.
If the request is merely “be creative,” we have left the central question unanswered. Which assumption should move? What must still work? What would make the departure matter?
The human arrangement
What disappears.
What takes the foreground.
Much of being human depends on continuity. We walk a familiar route without attending to every movement. The ordinary sounds of a room recede. A practiced activity becomes easier to perform without narrating each step.
Then something breaks the pattern. A sound stops. A familiar face appears in an unexpected place. One detail in a design refuses to fit. What had been background gives way to something that asks to be noticed.
Our attention is stolen;
it is not given.
I use that formulation to describe the force of the interruption. In this first sketch, I call the predictable background our most-likely side and the attended departure our least-likely side. This is a working lens for the human experience, not a complete account of consciousness.
The distinction also needs a boundary. An interruption can be trivial, unpleasant, or misleading. Winning attention does not establish value. Something still has to happen after we notice.
Creativity begins to matter when the departure can sustain an encounter: when a piece of music rewards another listening, when a building’s surprising form serves the people inside it, when an unfamiliar solution survives the demands of the problem.
Nature as an analogy
A tree frog remains recognizable.
Think of a particular kind of tree frog. Across individual differences, a recognizable form persists. The shape, the arrangement of the limbs, the way the animal moves: there is enough continuity for us to recognize the creature.
That continuity is the feature I want to borrow from nature. The analogy leaves room for variation, unusual individuals, adaptation, and change. Its point is that recognizable outcomes recur. Nature gives us a powerful image of the most likely without requiring every individual to be identical.
The image also raises the question of a blueprint. How does an organism arrive at a coherent form? The question leads beyond the everyday image of a list of instructions being mechanically assembled. What gives the whole its recognizable organization?
Within my Reality Equation framework, the future is pure potential and the past is singular: one actual past, containing what has happened. Human beings participate in relationships with ideas, which I describe as a subset of thought patterns. This is the conceptual setting in which I am exploring the recognizable form and the unexpected departure.
The creative encounter then becomes a question of relationship. What idea has begun to organize the work? What possibility is becoming compelling enough that the familiar form no longer satisfies it?
A description of nature does not settle that question. It gives us a way to see why continuity and departure deserve separate attention.
The assumption that moved
The keyboard stopped
owning the surface.
The original iPhone offers a useful example. In Apple’s January 2007 introduction, the large multi-touch display and software keyboard were central to the design. The keyboard appeared when typing was needed. It no longer had to occupy a permanent physical portion of the phone.
Look at the conceptual move. A conventional design question asks how to arrange the keyboard and screen more effectively. A different question asks why the keyboard must remain physically present when the task has changed.
The departure reorganizes the object. The same surface can serve a photograph, a map, a page, or a keyboard. Form and function meet in the changed assumption.
I call that a least-likely outcome because it departs from the familiar design answer in a consequential way. What matters here is the relationship among the parts and the new possibilities that relationship creates.
Open the study / three responses to one brief
“Design a better phone.”
Open each response. The useful distinction is the assumption it changes and the function it preserves.
These simplified drawings are a conceptual comparison, not a product history or a ranking of every possible phone. Familiar improvements can be valuable; a departure earns its significance through what it enables.
The moving boundary
Yesterday’s departure
becomes today’s default.
Once a departure works, it can be repeated. Its conventions become familiar. Designers improve it. Users learn it. What initially interrupted expectation becomes part of expectation.
That is why “least likely” cannot be a permanent badge attached to an object. The same design can surprise one generation and feel inevitable to the next. The artifact has acquired a different relationship with the world around it.
This also explains the dissatisfaction that returns after a successful innovation. We ask why everything looks the same, even when the sameness descends from something once celebrated for breaking the mold.
There is no failure in a useful invention becoming dependable. The mistake is expecting its original surprise to remain intact simply because its original form remains intact.
“Of course it works this way.”
The convention becomes difficult to see because it works.
“It can work another way.”
A changed assumption opens a different possibility.
“Of course it works this way.”
The departure is repeated until it becomes the expectation.
Working with AI
Give the departure
something to answer to.
If you want a familiar problem solved well, let the accumulated patterns help. If you want to reconsider the problem, become more exact about the intervention.
Name the assumption that deserves to be disturbed. Preserve the function that still matters. Examine the alternatives for what they make possible, then develop the promising one far enough to discover whether it actually works.
AI can participate throughout that process: generating alternatives, exposing conventions, elaborating a direction, and helping test its consequences. Human participation can be equally conventional. Neither side is confined to one kind of outcome. The distinction helps us recognize the tendency we are accepting and the work we are asking the relationship to do.
The more fluently a system produces plausible answers, the easier it becomes to mistake completion for discovery. A finished-looking artifact can close the conversation before its most important assumption has been examined.
That is where the human intervention matters. Stay with the question long enough to recognize the departure that deserves development.
The creative task is to bring forward a possibility worth becoming familiar.
1 thought on “Most Likely, Least Likely”