The Problem Is Usually in Front of the Monitor.

By Manfred & Martin (The Swimming Club), Germany.

We build technology to make life simpler. Then we forget how it works, improve it until it gets worse, and invent another button to rescue us.

Before the Swimming Club could begin talking about technology, we first had to get the technology working.

The Mayor had built a new Pineapple app and wanted Martin and Manfred to test the installation process. Martin pointed his phone at a QR code. Something opened. He thought he was in the Google Play Store.

“You can’t be,” said the Mayor. “The app isn’t in the Google Play Store. That’s physically impossible.”

“Physically, I’m here at home,” Martin replied.

It was an excellent beginning.

There followed several minutes of browsers, QR codes, links, menus, disappearing dialogue boxes and instructions to scroll up or possibly down. At one point Martin had to act quickly because his phone was making an option disappear before he could press it.

“I love this world,” he said.

Then came his theory of modern civilisation: perhaps people are becoming more frantic because their devices keep putting them under time pressure.

Eventually Manfred tried the alternative route. The link arrived through WhatsApp, the installation button appeared, the app installed and a yellow Pineapple icon materialised on his phone.

It worked.

Which, as the Swimming Club was about to discover, is one of the more dangerous moments in technology.

The Manfred button.

Suppose there were one magical button that could remove five irritating steps from your life. What would it do?

Martin immediately went after forms that ask you to enter information the system surely already knows. He also wanted rid of terms and conditions nobody reads but everybody must accept, usually after scrolling down one final centimetre to prove that they have participated in the ceremony.

Manfred approached the problem from the life of a software developer. His button would deal with documentation: write the help text, cope with the deadline that was yesterday, fill in the timesheet, and finally put away the coffee and cigarette and complete the task without addictive assistance.

Martin particularly liked the timesheet idea. Filling in hours from some distant week requires a special kind of archaeology.

His solution was obvious.

Ask Manfred: “What did I actually do last week?”

And so the first important technological discovery of the morning was made. Martin already has artificial intelligence. It is called Manfred.

“If I want to know something, I ask Manfred,” he explained. It was, he felt, the more reliable form of AI—although he added that he believes at least half of what Manfred tells him and assumes the opposite for the other half.

Magic has a short shelf life.

The strange thing about technology is how quickly miracles become furniture.

Manfred put it simply: once people get used to something, they stop thinking about it. AI still has a certain wow because it is new. We can still look at something it produces and wonder how on earth that happened.

Give it a couple of years and it will simply be there.

Smartphones have already completed that journey. The Swimming Club remembers when they arrived and people talked about what these remarkable little machines could do. Now we carry them everywhere without wondering much about them at all.

Until one breaks. Or disappears. Or we have to buy a new one.

Then the invisible technology becomes extremely visible again.

The same thing happens when something we once understood becomes unfamiliar. Manfred recently bought an inexpensive internet television service and wanted to run it through an Amazon Fire TV Stick. He had configured the same system years earlier, so he knew it worked.

Unfortunately, he no longer knew how he had made it work.

Which buttons had he pressed? Where was the setting? How did he get back to wherever he had once been?

It is a peculiarly infuriating problem because surrender is impossible. You cannot say, “It doesn’t work.”

You know it works.

Worse, you know you once made it work.

Martin knew the feeling. He had recently spent hours struggling with a documentation system after overlooking an instruction he himself had written into the template. A blank line in the wrong place was enough to cause the problem.

The documentation existed.

He just hadn’t read his own documentation.

There is apparently a technical expression for this kind of failure: the problem is not behind the monitor. It is in front of it.

When fixing things becomes a way of thinking.

For Manfred, this reaches beyond computers. Software development has trained him to see problems in a particular way. When something goes wrong, his instinct is not to stay with the problem but to look for the solution. Something is broken. Fine. What can be done? What is the next step?

That habit has followed him home.

His wife has occasionally pointed out that his responses can be very technical when something rather more emotional might be required. Manfred understands why. If your professional life consists of solving problems every day, eventually problem-solving becomes part of the operating system.

It can also become a trap. Once a programmer has found a direction, the brain can charge down that road with impressive determination while becoming almost blind to everything on either side. Manfred has seen it professionally: a developer becomes so absorbed in a solution that somebody from outside eventually looks at it and asks the devastatingly simple question:

What on earth have you done?

And suddenly the programmer can see it too.

This is why other people matter. Show them the software. Let them use it. Ask what they think.

Especially after version 29.

The danger of version 30.

The Pineapple app had already been through 29 revisions when Manfred’s test revealed the need for number 30.

After logging in, he was landing on an article page instead of the homepage. That needed fixing.

No great drama. Another adjustment.

But Manfred offered a warning from years of software development. There comes a point when a product works well. The obvious problems have been solved, the useful features are there and everything is doing roughly what it is supposed to do.

Then comes temptation.

Perhaps it could do this as well. And this. And wouldn’t another feature be useful?

That is when improvement can turn into what Martin calls Verschlimmbesserung — a wonderfully efficient German word for an improvement that makes something worse.

You optimise the thing to death.

The software that was supposed to become easier becomes more complicated. Eventually somebody has to remove last week’s brilliant new idea and go three steps backwards.

The Mayor had already felt the temptation. Perhaps another Fruit Basket could be added to the app. Perhaps the monthly conversation questions could go in there too. Perhaps—

No.

Leave it alone.

Version 30 had taught its first lesson.

A computer that cannot count properly.

All of this sophistication hides an absurdly simple machine.

When asked whether he thinks about the enormous amount of code behind an innocent little button, Manfred’s answer was immediate.

“No.”

The thing should do what it says it will do. That is what people expect from the software he writes, so why should he expect anything different?

And for all their apparent intelligence, Manfred pointed out, computers are spectacularly limited. Strip away everything we have built around them and a computer essentially does one thing extraordinarily well: it calculates extremely quickly. Give it zeros and ones and enormous speed, and extraordinary complexity becomes possible.

Humans, meanwhile, understand context that a computer must have explained in ridiculous detail.

Take a seemingly innocent request from the Mayor: build an app that tells him whether today is a good day to take the caravan somewhere.

Manfred considered the specification.

Weather would matter. Destination would matter. The Mayor’s health might matter. His mood might matter. A supposedly simple yes-or-no answer had already begun opening doors into an absurdly complicated system.

Manfred proposed a more efficient piece of software.

Press the button.

Today is a good day.

For everything else, ask AI.

What disappears when everything becomes easy?

There is another side to all this convenience. When software removes five steps, it does not necessarily remove only the annoying ones.

Manfred’s example was shopping. You can sit on your sofa, order almost anything and have it brought to your door. It is wonderfully efficient. But somewhere in that efficiency, human contact disappears. Sometimes the delivery driver becomes the only person involved whom you actually see.

Supermarket apps brought another irritation. Manfred had considered using them and decided the whole business was too complicated. The Mayor tried several and eventually deleted them because they were making him aggressive.

Then the conversation turned to discounts.

An app announces that something is cheaper. You notice it. Perhaps you choose one supermarket rather than another because the offer is there. Then you see another bargain and buy something you had never intended to buy.

Martin and Manfred arrived at the same wonderfully inconvenient conclusion: if you spend money on something you did not need because the app told you it was cheap, where exactly is the saving?

Technology has removed friction.

It has not necessarily removed manipulation.

Manfred 2.0.

By now, the obvious next step was to stop improving software and improve the Swimming Club instead.

Martin 2.0 came with three requirements. He should run on a purely mechanical device, remove all work from Martin and always be kind to him.

Reasonable specifications.

Manfred 2.0 required more delicate engineering. Asked what would have to change to make him more future-compatible, Manfred began identifying his own bugs. He can be stubborn. His sarcasm has become highly developed over the years. Both features, he admitted, are not always perfectly compatible with the rest of humanity.

But the upgrade also needed one completely new function.

A magic lamp.

Rub it and Manfred 2.0 grants three wishes.

After the third wish, Martin suggested, the whole system should automatically reset.

Perhaps that is the feature technology has been missing all along. Not another optimisation. Not another menu. Not another clever function hidden behind another clever function.

Just three wishes.

Then reset.

And if anything goes wrong, press the Manfred button.