AUTOMOTIVE / ANNOUNCEMENT
Why the Blind-Spot Warning Earned Your Trust, and the Lane-Centering Didn't
Four features. One pattern. What trust actually depends on — in a car, and at the head of an organization.
This essay applies an interaction design lens to driver-assistance features, then shows the same lens applied to leadership. It's both a launch statement and a demonstration of how the practice's method travels.
Blind-spot warning: trusted by 84 percent of drivers, the highest of any feature tested. Lane-centering, which steers and brakes for you in traffic: near the bottom of the list. Same car. Same manufacturer. The gap between them is the whole idea behind what I'm building next.
After thirty-plus years in automotive user experience — the last stretch of it at Stellantis — I've left to build Confident UX full time.
Here's the idea the practice is built on, and you already have an opinion about it, whether you've put words to it or not: once you can see why you trust one feature and not another, you can see the same pattern anywhere a person is trying to act with confidence — including at the head of an organization.
Ask a car owner which driver-assistance features they trust, and a pattern shows up almost every time.
Blind-spot warning — the light that tells you something is in the space you can't see — is one of the most trusted safety features on the market. AAA's research on real owners puts trust in it at 84 percent, the highest of any feature tested. Adaptive cruise control that handles stop-and-go traffic on its own — slowing, stopping, resuming without your foot — is trusted far less. Same research, same owners: 59 percent. J.D. Power's more recent look at driving-assistance systems that combine steering and speed found some of the lowest usefulness scores of any advanced technology in the car. A lot of owners say, in effect: I already know how to do this.
Same category of feature. Same manufacturers. Similar engineering budgets.
Why does one earn trust almost automatically, and the other has to fight for it?
There's an answer, and it comes from an idea that's been sitting in plain sight in interaction design for forty years.
In the 1980s, the researcher Don Norman studied how people act on anything — a door, a dashboard, a decision. He described it in seven careful stages: you form a goal, decide what to do, work out how to do it, and do it — then you check what happened, work out what it means, and compare it against what you wanted.
That's precise, but it's a lot to hold in your head while you're driving. In practice, it collapses into four moments that matter here. You perceive what's happening. You interpret what it means. You evaluate it against what you're actually trying to do. You act.
Do this enough times and the steps disappear. You just drive. But they're still there — and each one can be supported, or gotten in the way of.
Blind-spot warning supports the perceive step. It gives you information you didn't have — something is there. It doesn't tell you what that means. It doesn't tell you what to do about it. You still interpret it: is that car a threat right now? You still evaluate it: is it safe to move over? You still act, or you don't. The feature extended your senses. It never touched your judgment.
Rear park aid does the same job in a different spot. That warning tone that speeds up the closer you get to something behind you isn't deciding whether it's safe to reverse. It's not touching the brake. It's just telling you how close you're getting, faster and faster, and leaving the rest to you. (Some systems bundle in automatic braking under a similar name — that's a different feature, doing something closer to what's coming up next.) The plain version — the sound — is exactly like blind-spot warning: extend the senses, leave the judgment alone.
Now take the feature that manages your steering and your speed together, continuously, in traffic. To do that, it has to interpret the situation — is the car ahead slowing, speeding up, changing lanes — and evaluate what should happen next — hold, brake, steer. It has to make the call that used to be yours. And it makes that call quietly, continuously, without asking.
Perceiving extends what a person can already do. Interpreting and evaluating replace something a person was already doing — the part that feels most like being in control.
Automatic emergency braking (AEB) looks, at first glance, like it belongs on the trusted side of this list. A system that just executes a decision you already made — the way a seatbelt does. It doesn't.
AEB has its own goal built in: don't hit that. Nobody has to state it. It's just there, all the time, watching.
It watches how fast you're closing on whatever's ahead of you. It's constantly asking one thing: can you still stop in time on your own? The moment the answer is no, it brakes. It doesn't ask first. Sometimes it brakes before you've even noticed the danger yourself.
That's not the act step, quietly finishing what you already started. That's the system interpreting the situation and deciding what to do about it — the same two steps that erode trust everywhere else on this list. And yet AEB is broadly trusted, and increasingly mandated. Nobody treats it like lane-centering.
A few things explain the difference. The goal AEB assumes is universal and non-negotiable — every driver already wants to avoid the collision, all the time, with no exceptions to weigh. Lane-centering assumes something far more contestable: an ongoing judgment about how you want to drive, which is closer to a preference than a shared goal. AEB also only acts in the narrow window where a person's own reaction time provably can't keep up. And it fails toward silence: doing nothing is its default, and it only intervenes when the alternative is a collision.
There's a fourth difference, maybe the simplest one. When AEB fires, it does exactly what any driver would do in the same instant — stand on the brake, as hard as possible. There's no version of that moment where a reasonable driver says afterward, I'd have braked more gently, or, I'd have swerved instead. Lane-centering doesn't have that luxury. It's constantly making calls where a reasonable driver would genuinely want it done differently — hold closer to the car ahead, cut the corner a little tighter, ease off the gas sooner. Real preference is at stake, and the system is guessing at it without asking.
Put simply: when the goal is uncontestable, even an algorithm can do it well. That's part of why AEB was the easy one to hand over.
The trust gap, it turns out, isn't simply about which of Norman's four steps a feature touches. It's about what's being assumed on your behalf, and under what conditions.
It's tempting to read the trust gap as a technology gap — adaptive cruise control will earn the same trust as blind-spot warning once the software improves.
I don't think that's it. Interpreting and evaluating are inherently harder to support without taking over from the person doing them — and that takeover is what erodes confidence, no matter how good the underlying system is, unless it meets AEB's terms: an uncontestable goal, a gap human timing genuinely can't close, and silence as the default everywhere else.
Confidence is not a personality trait. It's a system property.
Most of the tools an organization buys for itself work like the trusted half of that dashboard. Reporting. Analytics. Dashboards that show more, faster. They extend what leadership can perceive.
Very few tools help with the harder half — helping an organization interpret what it's seeing, and evaluate it against what it actually intends, in a way people trust rather than quietly work around. When a tool tries anyway — handing down a conclusion, a score, a recommendation without showing its work — it runs into the same problem the lane-centering feature runs into. People turn it off. Or they nod at the dashboard and go back to their own judgment.
What Confident UX does starts with AEB's secret: help a leadership team define its uncontestable goals — the equivalent of don't hit that, translated into their own terms. That's not a small thing to hand a team. An uncontestable goal is what lets everyone stop arguing about what they're aiming at, and focus on what matters.
Then it helps that goal scale — show up the same way in the hundred small decisions made across the organization every week, the way AEB's goal doesn't live in a policy document. It lives in every foot of closing distance.
Only then does Confident UX turn to the half AEB doesn't touch either: helping people perceive where hesitation is showing up against that goal — where confidence is thinning before anything visibly breaks.
The interpreting and the evaluating stay exactly where they should: with the people responsible for them. Confident UX doesn't hand down a verdict. It hands over a clearer goal and a clearer view. The judgment stays yours. It just gets sharper.
This is a UX — a user experience — lens, and it isn't limited to screens and dashboards. The same three questions — what does this person want, does the interaction make the next move obvious, does the outcome deliver — apply to leadership, hiring, a boardroom decision, a hockey team, a golf course, an F1 pit crew, anywhere a person is trying to get from what they want to what actually happens. Cars just happen to be where I learned to see it first.