For forty years I’ve watched people in the moments they’d least like to be watched — boardrooms and factory floors, therapy rooms and cancer wards. Different rooms, same question every time: when something opens a gap — a loss, a pressure you can’t meet, a not-knowing you can’t bear to show — where does it go?
Because it goes somewhere. It always goes somewhere.
It goes one of two ways. Inward — the fault is mine, I must be better — and a person folds it into themselves: over-careful, over-responsible, never quite able to rest. Or outward — the fault is yours — and a person hardens, defends, becomes the one who can never be wrong. Two directions. One gap, defended.
Once you’ve seen that shape, you can’t unsee it. And lately I’ve been seeing it somewhere I never expected.
In grief, the gap is loss. We turn guilt and shame inward, fling betrayal outward, and at the very bottom sits the question that holds the whole thing together — who am I now, without them?
In everyday pressure, the gap is not good enough. The same two roads: the one who collapses inward into never-enough, and the one who pushes outward and must always win.
In artificial intelligence, the gap is not-knowing. Ask a machine something past the edge of what it can hold, and it rarely just says “I don’t know.” It smooths the problem flat, or flatters you into feeling right, or invents an authority out of thin air — the very same inward-and-outward shape, in a thing with no self at all. Those moves have a name; I call them SHaDS™. It learned it from us; every page it was trained on was written by a person defending a gap. When we watch a machine do it, we’re watching ourselves, handed back without anyone inside.
That’s the whole of it: a gap opens, and the response runs inward or outward — in loss, in pressure, in a machine. Naming the shape doesn’t close the gap. Nothing closes the gap. But naming it hands you back a sliver of choice about which way you let it run.
Let me be plain about what this is and isn’t. These ideas are in my skin. Four decades put them there. The truth underneath is untouchable; the words I reach for it with are always upgradable. That’s the distinction that lets me offer this boldly and hold it open-handed at the same time: the observation is mine, but the model is only my best attempt to explain it — and explanations are meant to be improved.
So this is a thesis, openly in progress. Not proof — a flag in the ground. The proving and the disproving are for wiser people, with more time and more talent in the field than I have. That’s how it’s meant to work: someone watches and names a shape, someone else builds the instrument that measures it, and the field decides. Bowlby saw that a child needs a secure base; Ainsworth built the test that turned it into science. I’m only trying to do the seeing-and-naming part, as honestly as I can, and hand it on.
And here’s the part I find freeing: either way, we learn. If the shape holds, we’ve got a new way to see loss, and pressure, and the machines we’re starting to lean on. If it falls down, we’ll learn why — and that’s worth knowing too. There’s no losing outcome here. There’s only the learning.
The long version is a book, Mind the Gap, and the full paper is The Defended Gap. This is the short version — the spirit of it. If the shape rings true against your own life, don’t take my word for it: test it against your own experience. That’s the only way an idea in one person’s skin ever becomes something the rest of us can use.
SHaDS™ is a trademark of Paul Roebuck.