I was walking through l’Hermitage in Leningrad, untroubled by the affected Frenchification of its name because I don’t read any Cyrillic script, happy that someone at least recognised that localisation is not necessarily a good thing. Not that I thought about such things at the time. I was 18, I’d just taken my first flight, a cramped little plane filled with foul smelling cigarettes and malodorous people, many calling incessantly for little bottles of vodka when it was clear that the stewardess’ time would have been better spent if the airline had just dished out full size bottles to each passenger as he, they were almost all men, boarded.

I’d arrived in somewhere that was about as foreign as it was possible to get from the coal, iron, steel, ship- and bridge-building town where I’d spent most of my teenage years and I was on a school trip, arranged by the local segregated (by gender when there were only two to think about) and a few stragglers, like me, annexed because my mother knew someone that worked at one of the grammar schools. I had presented my freshly minted passport and a document that, many years later, would become central to my life, a “visa”.

I didn’t know what a visa was: I just thought of it as a sort of entrance ticket, like I’d need to get into a gig at Newcastle City Hall, or Redcar Bowl, places where icons of my time came so we could imagine, just for a couple of hours, that we were part of their world. Then I’d go home, they’d get in their scruffy buses and the spell would be broken. Until the next time.

It was in Leningrad that I was first introduced to the concept of rigid compliance. Go there, don’t go there, move now, don’t move. Instructions were dressed up in pretend courtesy: “This way, please.” Any deviation was met with a chiding that today is replicated by the voice on the SatNav as it becomes increasingly frustrated with a driver who goes his own way.

It’s no surprise that someone, somewhere, wants the car to take complete control. Freedom of throught? That goes out of the window the moment the car recognises the approaching key fob.

Naturally, as a long-haired westerner with a penchant for getting into trouble in times when getting into trouble was actually quite difficult, but never any real trouble, nothing that is that would cause my parents anything more than a grumpy, frumpy, “humph” over dinner, the idea of being in such a straitjacket was never going to work. And it didn’t.

Just a couple of minutes through the front door of the famed Winter Palace, I stopped abruptly. I just stared. The most beautiful thing I’d ever seen was there, in front of me.

The (I assume) increasingly strident tones of our “guide” didn’t penetrate my consciousness until she poked me. Every part of my body and, if one believes in such things, my soul was taken up by a pair of doors, maybe ten feet wide and twenty five feet high (or so, today, they seem in my memory) inlaid with the most exquisite patterns in gold. I didn’t question its craftsmanship, who had made it, what the costs, in ecological, human or financial terms.

I just felt better for knowing something so perfect could exist. Why think behind it when it was clearly beyond my comprehension, even beyond my imagination.

And yet, as a pair of doors, prosaically, it was commonplace.

Something everyday had, by the application of forces I could not conceive of, transcended their purpose and become something I would never, at least not so far, forget. Which is lucky because I was so transfixed I didn’t think to take a photograph.

That, I’m sorry to say, is exactly what is happening inside companies today. Except that the technology which is being designed, built and deployed is far from perfect and, unlike the doors, it is often left to function in isolation, one part independent of another, with no brain overseeing the whole and recognising the results. No one has a strategic overview of everything tech is doing and where it is doing it. And it is not designed, built and deployed by skilled artisans who understand form and function and the effects it has. Tech is built by people with a single task who do it and walk away, their job finished.

In software-world, no one ever does the equivalent of coming back to make sure the doors are hung properly, that they swing smoothly, that they close without gaps or overlaps, however small.
In software-world, no one can hear you scream

I’ll give you an example: I got a pre-paid mobile from the same provider as my post-paid. I had to go through all the same identity checks. I don’t use data on the prepaid, and I don’t use SMS. In fact, I have it purely so I can reveal a public number because I’m selling my flat and one of the most virulent issuers of spam in Malaysia are estate agents so they are not getting my real number nor my real email address. Every few days, I get an SMS from the provider trying to up-sell the data in the package. Their system is designed to do that but not to check if I use data at all, never mind anywhere near the limit. If they cross-referenced, they would know that for more than a decade, I have never used even close to the data allowance in my post-paid account. So why are they pestering me with endless messages? It's because no one has looked beyond the superficial - here's a number, let's try to sell something. It's legitimised spam.

But they don’t cross-reference. I tried to pay for the next month's extension: the system required me to enter an e-mail address. The excuse that is currently being used for taking email addresses in Malaysia, used by businesses of all kinds, is that it’s needed so that an e-reciept can be issued, despite the fact that I was attempting payment within their own app (which is installed on my phone without my consent when the card was installed. More about that, and the e-invoicing email thing later). They already have it. And they know they have it because my ID for both accounts was the same passport. So, somewhere two systems have not been connected.

So, I tried an alternative: I went to the only financial app I allow on my phone, an e-wallet, from which I pay my post-paid bill every month. I have to do this because the phone company refuses to charge my overseas cards but I can pay using the wallet wherever I am. I tried to add my pre-paid phone to my post-paid account. I entered the number into the space provided. The money was deducted, then immediately refunded without explanation.

So, I tried an alternative: I went to the mobile phone company’s app on my postpaid phone and found the “pay someone else’s bill” tab. I entered the number, expecting it would be added to my bill. No, it required me to pay. I chose the e-wallet. After a tortuous multi-stage and multi-platform process, there was confirmation that the bill had been paid. But the balance on my e-wallet didn’t change and the transaction did not show up in the records. The next day, I found an email on my desktop saying the payment had failed and the company was awaiting payment. Who, in their right mind, uses so many channels of communication for the same transaction? Who, in their right mind, designs a system that requires multiple platforms, has multiple points of failure and communicates those failures via a multitude of communications channels? And who, in their right mind, especially if they are designing a KYC system, does not consolidate records belonging to the same, confirmed, identity?

This should not be complicated but clearly no one has taken a look at the whole process in the round.

The app that the prepaid installed on my phone without my consent said my name was "Joseph" and gave me details of his account. So no one has thought that when a phone number is allocated for re-issue that the account details should be reset before the new user puts the card in his phone. Oh, and thanks, Joseph, your number is a magnet for fraudsters with autobot calls inviting you (it must be you not me 'cos it started hours after the phone was connected) so clearly you were not careful enough with your personal details when you had the number.

Which brings us to the e-invoicing thing.

At the beginning of the CoVid-19 pandemic, the Malaysian government made a number of extremely good decisions. One was that people were allowed to go shopping for food. But, before tracking and tracing apps were available, which took only few weeks for the first generation, premises were required to register those who entered. That meant revealing a mobile phone number and a name. This caused uproar amongst those concerned with privacy and security. Some people lied, some people got a second phone and some people breezed past, knowing that a minimum wage, migrant, guard was not going to risk his work permit by getting involved in an argument that would end up on social media, no matter that he was in the right.

Very quickly, the government made it clear: using its app meant there was no need for registration at the premises. Scan a QR code and the system did it for you.

Fast forward five years and a widely hated compulsory e-invoicing system is now in place. It is now impossible to payunless an email address is revealed, even if one pays in cash. E-mail addresses are one of the primary means of starting an investigation into someone and investigating is exactly what criminals do. It also provides a treasure trove of information, often sold by disaffected staff or those who are happy to reveal data for a surprisingly small amount of money. Like property managers who release ownership, property details and contact information to estate agents.

E-invoicing, across the world, is being sold as “green” approach – millions of slips of paper aren’t generated and thrown away. But it also provides exactly the kind of data analysis information, and centralised tracing, that many people resent. Whatever you do, wherever you go, whatever you buy, that data is now in the system and available to be obtained by hackers or sold by staff.

How, one has to ask, does a government which understood and removed the security risk of compulsory disclosure of personal information turn around and say now you have to reveal that information when you pop into a convenience store and buy a Mars Bar?

The answer, surely, has to be that people work in silos. The lack of overall strategy and effective policing to ensure that systems work together and not in conflict affects the entire world of commerce and government, across the globe.

The failure to rationalise data is one problem. The failure to use data to reduce the burden on customers is something else entirely.

But someone, somewhere, is looking at the end result and thinking “That is the most beautiful thing I have ever seen”. They don’t question its craftsmanship, who had made it, what the costs, in ecological, human or financial terms. They just feel better for knowing something so perfect can exist. Why think behind it when it is clearly beyond their comprehension, even beyond their imagination.
Ever more stupid people

This lack of questioning, even of checking, to see what’s not perfect, has led to a raft of very stupid people, backed with a dangerous dose of arrogance. There is a feeling of corporate infallibility: “my company built it so it must be right.” Managers don’t pay attention to complaints from customers because the company, or to be more precise the company’s computers, are always right. In any case, how dare anyone challenge the god-like authority of computer-people who all-too-often, convinced of their own infallibility, respond to complaints with, either, “you just don’t understand” or “it works as intended.”

If such complexity, confusion and inefficiency is as intended, heads need examining. With a brick.

Many software programs now offer “dark mode” – good, it’s about time someone started looking after our eyes. But the program designers fail to make the toolbars work properly. Icons simply disappear, leaving users to remember where they should be and to rely on pop-up tool-tips to show if we are in the right place. It is inconceivable that the designers did not test the theme before releasing it, or that of the millions of people who use the software no one has complained. So, yes, it is working as designed, or at least working as neglected. How is that a good thing?
In the absence of "AI"

None of the examples I have shown have anything to do with so-called artificial intelligence. They are all simple, old fashioned programming, quality checking, etc. A computer might be used at some stage of the process but the failures all result from a false assumption that computers solve problems and therefore people can, and do, become more stupid.

The attitudes are, however, part of the development process of "Artificial Intelligence" which is just a poncy name for software.

These same attitudes have spread across organisations: surly barmen get uppity when a customer points out lipstick on a glass. That glass has been washed, collected from the washing area, transferred to a distribution point, taken down, held while being filled and then handed over. But, somehow, it’s the customer who gets huffed at, who sees the same drink poured into a different glass, who risks his drink being contaminated with snot or sweat in revenge.

In a restaurant, when the wrong food is delivered, the opportunities for cock up have been just as many.

And so is the affixing of the correct identification tags in a hospital.

None of those are computer-generated errors, and in truth the vast majority of things that go wrong are not computer-generated and even if they are, they are not computer-perpetuated. There are almost always human breakpoints, as there should be, but these are all too often not taken sufficiently seriously.

Philosophically, we are in a world where wonder is not reserved for the perfect, but it is widely applied in automatic reliance on processes and on computerised output.

It’s a world where computers are assumed to be “set and forget” and once they are running, no one listens to see if they are creaking and grinding as a real machine would do, looking to make running repairs and to lubricate where necessary. Once the process has started, it is assumed to be good.

Product is released without proper testing, knowing there will be failures many but not all of which will be fixed with a patch, but not until someone has already suffered the adverse consequences of software that was demonstrably not fit for purpose. And even if individual pieces of software work, their interaction, or lack of interaction, with external influences – human or other software – is not a primary consideration: that’s the rushing out of new features, which are also untested.

Before we entrust functions to computers, we must first decide: can we be trusted to make a good job of it?

The answer, so far as I can see, is a resounding “no.”

We are intellectually and emotionally superior to machines. Increasingly we choose not to recognise that and to delegate our place in the world to faceless programmers who could be anywhere and anywhen - and who have no long-term relationship with the software and therefore no actual or moral accountability.

As a species, we rely on the false prophets of software-world and allow ourselves to become ever more stupid. We are too willing to accept the concept of rigid compliance with what we are told. Go there, don’t go there, move now, don’t move. Instructions are dressed up in pretend courtesy: “This way, please.” Any deviation is met with chiding.