Sitting in an online meeting this morning, having, for the millionth time, refused to download another pointless app, something occurred to me. The person I was waiting to speak to had no connectivity. At the same time, one of the major broadband providers where I am (though happily not my own) went down. Yesterday, a power issue at Google in the UK knocked out significant parts of the Internet for several hours, including my banking service of choice.
Earlier in the week, I spent hours on the phone, for the third time in as many weeks, attempting to get a telco to actually install the phone line and broadband they had promised an elderly relative.
Then my computer, which should be sufficiently powerful to do what I ask of it – displaying text and images – crashed.
That thought that occurred, though? If, as at least appears to be the case, people are increasingly irritated all the time, is the routine, and one can only assume, intentional degradation of information technology (IT), not to mention its use to paper over cracks in businesses, organisations and even society at large, just driving people crazy?
Perhaps not. It’s plausible enough, but correlation and causation and all that.
Still, IT is a source of frustration. It always has been. However, as someone who has used computers since the days of BASIC and CP/M it strikes me that the source of the frustration has shifted.
Of course, we are now much more dependent on technology than we were back when DOS roamed the earth. As a result outages, downtimes and even crashes are just bigger problems than they used to be.
But there’s something else. In the 1980s, 1990s and 2000s, information technology often failed to quite match up to users’ expectations – and suppliers’ claims. In addition, back then computers were hard to use, demanding real skill, not to mention the learning of all manner of arcane commands. These things were bound to cause frustration.
Today, it seems that the main source of frustration is decisions made by the people supplying the technology.
This column has long been long on complaints about the state of both software and hardware (and, to be fair, short on solutions). However, rather than rehearsing my suspicion that most software does not really need to be delivered ‘as a service’ or ask, once more, why user-interface design has collapsed into a morass of attempts to baffle and bilk users, let’s look at something else. Let’s look at how some software companies interact with users.
Posting on LinkedIn and X (formerly Twitter) this week, engineer Kenneth Shultz announced he was replacing a major CAD application with alternatives coded using artificial intelligence (AI).
Why? Because the software’s publisher e-mailed his company, giving it five business days to respond to a licence audit. Anyone who, well, has a job will know that this is a disruptive request, particularly so in any business that doesn’t have its own internal IT department.
We’ve already seen plenty of frantic market action with software company share prices getting hammered once people realised AI could, in theory at least, replace an awful lot of specialised, so-called ‘vertical’ applications. True, the iShares Expanded Tech-Software Sector fund, which tracks global software houses, did recover somewhat from its early 2026 dive, but as of 20 October it remains down 10% over the last 12 months.
The giants are only part of the picture, though, and it stands to reason that vertical software faces the greatest risk. After all, these applications are rarely run, or written, for love, are often extremely expensive, hard to use, and come with confusing licence arrangements.
Frankly, an awful lot of consumer software could be replaced, too. It would be unwise to do so, at least without understanding the potential consequences, but it could be done.
For my money, I wouldn’t be keen on actually building anything sketched up in a vibe-coded app, for fairly obvious reasons including physics and professional liability. I might use a vibe-coded word processor, though. Or page layout application. Or video editing software. Or photo retouching. Or all manner of small, micro-applications that no commercial software house could ever be bothered to develop. And in the case of, say, golf-course management software, I don’t know about you, but I’d be delighted to get away from paying every month or year. Especially if it was coded to perfectly meet the requirements of whatever quixotic business systems the golf course has been using for two decades.
Given all of that, is now really the time to be annoying users? Food for thought if you’re on the board of a software company.
Ltd
