Transcript
EXTRACT ONE\n\n[MAN]: I think the redesign can work, but not if we pretend the square is bigger than it is. We keep talking about adding another stop, which is tidy on paper, yet the real pinch point is the crossing lines.\n[WOMAN]: Exactly, and that's where the market changes things. People don't just arrive; they drift in with bags, children, bicycles — all the small movements that make the route messy.\n[MAN]: So the issue is not raw numbers, then. It's the sequence.\n[WOMAN]: Yes, though sequence sounds too neat. It's really about whether someone can read the place in five seconds. If they hesitate, they spread out.\n[MAN]: We saw that on Tuesday — I want to say around half past eight, though it may have been nearer nine — when the delivery vans came in.\n[WOMAN]: And those vans don't come every hour, that's the point. They come in a burst, the same burst as the school run, so the lane looks empty and then suddenly isn't.\n[MAN]: Right, but if we put the emphasis on the bridge approach, won't that draw people away from the frontage?\n[WOMAN]: Not if the frontage is clearer. The bridge is only part of it. What matters is the way the pavement opens, or rather doesn't open, into the square.\n[MAN]: So more space won't solve it on its own.\n[WOMAN]: No. Better cues, a more obvious line, and fewer moments where people have to guess where to go — that's what will keep the buses moving.\n\nEXTRACT TWO\n\n[SPEAKER]: I used to think good psychology would arrive like a clean answer. That was the attraction of the lab, really — the sense that if you measured carefully enough, the pattern would reveal itself and you could almost trust it to speak.\nThen I took a post in a secondary school, which was meant to be temporary, and the first surprise was how little the students cared about elegant explanations. They wanted the corridor to feel manageable at nine in the morning, not after three months of reflection.\nI remember one boy — I think it was 2017, though it may have been early 2018 — who never filled in the questionnaires but would copy the seating plan into his notebook because he liked knowing where he could sit.\nAt first I found that oddly disappointing. The data were thin; the story felt unfinished. Yet the teacher said, 'Look again.' And when I did, it was obvious that the notebook wasn't a side note at all. It was the intervention.\nThat was the shift for me: not from science to common sense, as people sometimes imagine, but from chasing the grand explanation to noticing the tiny routine that made the day less chaotic.\nI still like a good model, of course. I just don't confuse it with change anymore. Change is often boring, a little repetitive, and — this is the part I had to learn — more local than heroic.\n\nEXTRACT THREE\n\n[INTERVIEWER]: A lot of hospital managers say the new triage software is valuable because it is transparent. You can see the ranking, you can see the score, and that, on the face of it, feels reassuring.\n[EXPERT]: It does, though transparency is only the first layer. If the score is visible but nobody can challenge what it does next, then you have a polite machine, not a fair one.\n[INTERVIEWER]: But isn't that still progress? I mean, compared with decisions made in a corridor at three in the morning, an algorithm looks orderly.\n[EXPERT]: Orderly, yes — and that is part of the problem. People hear 'order' and think 'consistency'; they don't hear the older question of who gets to interrupt the process when the situation is unusual.\nWe saw this in the pilot last winter — I want to say November, though it may have started in October — when the staff kept overriding the score for older patients with multiple conditions.\n[INTERVIEWER]: So the overriding wasn't a failure of the system?\n[EXPERT]: Not really. It showed that the software was being used as a prompt, not as a verdict. The real issue was that the nurses had no easy way to record why they intervened, so the reasoning vanished after the moment passed.\n[INTERVIEWER]: Which sounds like a paperwork problem.\n[EXPERT]: On the surface, yes. But underneath it's a governance problem: if the reasons disappear, the institution learns the score and forgets the judgment.