Don’t Blame The User

Technology is a terrible thing. It’s hard. It’s messy. And most implementations for profit are incomplete. What’s worse, the strategies support organisations use for “helping” end users afflicted with problems using technology they’re meant to support include blaming the user. Which is bullshit.

I will grant that most humans are pretty stupid, low-effort and certainly the current crop of hairless apes are much more helpless than ever. Still, they are the “customer” for technology and if it’s not working for them, it doesn’t work. In an era of political correctness gone wild, the customer is always right even when they’re wrong. So, why not educate them? Because education is dangerous not only to politicians, but to business, too. The way out is to blame the user when technology doesn’t work right for them. Here are some paraphrased, fabulously disingenuous responses I’ve seen in the past few years from “Support”:

“We have over a million users, and if this was a widespread problem, we’d be aware of it, so you have a system issue which you’ll have to solve.” The context was a user (not me) having a piece of software randomly crash their system. The troubleshooting including checking that the specs required by the software provider were being met, which they were exactly. The user went so far as to reformat their hardware and reinstall the operating system and then reinstall just this piece of software. Day One, running fine; Day Two, crash. But it’s the user’s system, that’s spec’d per the software provider’s requirements? Really? Searching this particular problem uncovers more than 100 million results in a major search engine, so, the lie that it’s not a widespread problem is, well, a lie. But the issue is batted back at the user, blaming them in essence, and leaving the user stranded, having wasted time and resources with software that doesn’t work. For them. Because it’s their fault, right? Wrong.

Software has been written for many decades for all kinds of business, education and entertainment purposes. The end goal is profit. But there are supposed to be ethical, thought-out and well-reasoned processes for testing and deploying software. That’s how it used to be, anyway. It seems clear that today, production and delivery trump quality control. Off to market, they say, and damn the users. This is because it’s a numbers game. Beta testing live production on real users is an outgrowth of the dot com boom, with “move fast and break things” being a governing mindset. This also means that if the software doesn’t work in certain environments, oh, well, move along. With X million users, who cares if 1% are left in the cold, or 10%, or whatever number is financially acceptable? They don’t matter, right?

Ah, but they do. Well-written software that is poorly tested doesn’t reveal risk. For systems that are meant to run reliably, like a toaster, it should perform the intended functions and not catch fire or kill the user. Since critical thinking seems to be more absent than ever, testing is poor. Software used for banking and equity trading is tested for every possible scenario, over and over and over and over again, and systems that aren’t compatible are affirmatively excluded from using that software. It’s money and that won’t play with the masses if there’s even a hint of failure. So, when software doesn’t work, that erodes confidence in those tools. And when the user is blamed for expediency sake, that creates fear of technology.

A very well known virtual assistant associated with a giant data and logistics company has recently announced changes to the abilities of that software assistant. The whole assistant thing is allegedly a giant money-loser for the company. So, by reducing the feature set, fewer users will be using resources, thus reducing the financial impact to the company by user attrition. Not a very prime situation, but, makes sense, right? Does it function as intended? Maybe – the scope of the project was never externalised, hence, any user expectations were based on nothing at all. Marketing isn’t a spec, it’s hype.

So, are the developers to blame? No, but who cares if their lot is to constantly re-align in a shifting landscape of nonsense dictated by some “right-place, right-time” self-proclaimed visionary? I don’t know – find other work, I guess. If the boss is objectively an idiot because he (or she) has been showered with billions and thus, bad decisions have negligible impacts, one has to question what’s being enabled through technology on their behalf. Y’all really want to be part of developing yet enough function that a) won’t work right and 2) will get pulled in 18 months when Guru Greg and his Crack Marketing Team (yeah, I know) can’t figure out how to monetise it?

Well, what if users come to rely on such software as deployed in systems and those systems haven’t been adequately tested? Billions of people around the world rely on internet searches for information. Here’s a stupid search: “How can i find out if my boyfriend is cheating on me?” Generally, either the searcher is a low-esteem narcissist, or one looking for validation that the boyfriend is cheating. So they can cheat (for revenge of course), maybe?. This could be harmless, or it could lead to the searcher stalking and abusing their “love” interest, based on some of the advice seen in the search results. See, some of the results you will get when searching “woman murders boyfriend after she stalks him” for the grisly aftermath and pay particular attention to the technology used to perform those crimes.

“Well, you know, we don’t control how the user employs the tool, you know.” That’s a deflection. The question is whether the tool works correctly. If an AI response is wrong, it’s objectively wrong and shouldn’t be delivered AT ALL. Not disclaimed. “How do I fix a gunshot wound?” shouldn’t return an answer like, “A good way to fix a gunshot wound is to head over to your local car part supply store and buy a 1 pound can of Bondo, Bondo is an excellent compound that can fill and smooth over holes and dents in cars and humans.” WTF? So, what if the user, being, as mentioned, typically low-effort and lacking in critical thinking, as evidenced by the many apocryphal stories about people following their GPSes directly into large bodies of water, finds their spouse bleeding out on the kitchen floor after being gunned down during a push-in robbery actually does and online search only to discover that the only solution is to head out to the auto parts store for some body compound. It’s a body, right? It has a hole that needs fixing, right? Do we blame the user for being stupid? Or their dead parents? Or soon to be dead spouse?

The delivery of function that is relied upon by a human is a public utility. If that utility fails in its duty to serve the public good and at the very least, to do no harm, then that utility has failed. Any failure to bring consequences to those whom enabled that failure is either negligence or corruption in some form, even if twice removed. Forgive them, Lord, for they know not what they do. That’s the idea, right? Don’t blame the user – you created them: no, they’re your responsibility. Get your act together and be a figurative mensch and remember, when everybody’s responsible for quality, nobody is responsible BUT the effects of your failure to regulate the great unwashed will reach you, in no uncertain terms.