If you’ve used pretty much any consumer software or internet search engine in recent years, you’ve noticed a decline in user friendliness and utility. What was once a simple matter of navigating intuitive interfaces and obtaining relevant information has become a frustrating ordeal.
From constant updates that seem to break more than they fix to search engines returning an avalanche of ads instead of relevant answers, the slow death of corporate software has left many asking:
Let’s dig into the mystery of tech companies’ baffling self-sabotage and try to unearth some answers.
One of the primary culprits is the obsession with change for sheer novelty’s sake. Tech companies, in their desire to stay ahead, often push changes that don’t improve quality of life for users. Instead, these updates are driven by worship of innovation or attempts to chase trends with no thought for whether or not they benefit users.
The result? Features that once worked smoothly are altered beyond recognition or removed entirely, sowing confusion that frustrates even seasoned users.
How often has a simple software function that’s become second nature to use been saddled with needless complexity in the name of “improvement”? Time and again, familiar, reliable functions are buried under pointless clutter. And more often than not, these changes are superficial, lacking any real benefit. Nonetheless, you are forced to re-learn tasks you’d once mastered. Rather than making your job easier, these updates introduce more friction into everyday use, and for no good reason.
Related: Dead Internet
Another major factor in the decline of user-friendly software is the loss of institutional knowledge. Many of the original programmers who built the foundations of today’s software and search engines have retired or moved on without passing down their expertise to the next generation. The engineers who created pioneering systems had intimate knowledge of the architecture and purpose behind their design choices. They focused on utility and usability because they understood the needs of the user.
As newer developers take over, this crucial knowledge is being increasingly lost. Without a proper transfer of insight and context, software becomes prone to inefficiencies and errors. Newer developers, lacking the deep understanding of why certain design decisions were made in the first place, may inadvertently sabotage important functions . This erosion of expertise results in bloated, less efficient software that breaks more often and becomes a nightmare to use.
But that’s not even getting into search engines.
Once a go-to source for information, former internet pillars like Google, Yahoo, and Amazon have fallen victim to declining utility, thanks in large part to the glut of paid advertising.
Google’s search results, for example, have become overrun with ads, sponsored links, and SEO-driven results that push the answers you actually want farther down the page. What used to be a straightforward search for information now requires sifting through several layers of irrelevant or sales-driven noise.
This problem extends beyond search engines to platforms like Amazon, where users often find that the top-listed products aren’t the best or most relevant, but those from sellers who’ve paid the most to be there. The overwhelming presence of ads has warped search algorithms, undermining their original purpose of bringing users the most useful and accurate information with speed and efficiency.
And it would be professional negligence not to point out the digital elephant in the virtual room: big tech censorship.
Because somewhere along the line, software companies started by nerds in garages appointed themselves the Internet Police. Under the guise of protecting users from misleading information, the billionaires in charge of these megacorps have launched a crusade to suppress ideas they disagree with. And since their insular worldview clashes so brazenly with reality, the results are ever clumsier and weirder. Take the recent fracas over Google’s Gemini A.I.
With search engines becoming less reliable by the second and software growing ever more unwieldy, people are seeking alternatives. Perhaps surprisingly, large language models like ChatGPT are gaining popularity. The rising preference for consulting LLMs, despite the much-publicized Google Gemini debacle, testifies to just how ad-cluttered and unusable old style search engines have become.
Contra the popular preductions of doom, we’re seeing rapid improvement in A.I. models akin to Google’s boom phase twenty years ago. If internet users’ past behavior is any indication, the shift toward LLMs will not only continue but grow. As search engines hemorrhage user confidence, and software disappears farther into its own navel, A.I. will increasingly draw users away. And while valid criticism has been lodged against the accuracy of A.I., in the final analysis people want tools that make their lives easier. The lessons of DoorDash, on-demand streaming, and online shopping dictate that convenience beats quality.
The decline of consumer software and search engines isn’t a passing phase. It’s part of a broader trend fueled by unnecessary changes, the loss of foundational knowledge, the pervasiveness of paid advertising, and tech censorship. As users face mounting frustrations, they are seeking out alternatives—most notably large language models—to find easier answers. No matter that the easy way and the right way aren’t always the same.
The deep lore of Tolkien meets the brutal struggle of Glen Cook in the dark fantasy prelude to the acclaimed Soul Cycle.
Brian
Concour. Wordprocessing software is my example. The MS office ribbon is an excellent example.no one asked for it but MS imposed after the billon points of data it collected . Everyone had to unlearn 40 years of software use had to be unlearned. And for nothing.
My biggest gripe is software developers never leave their cubicles and actually talk to users. Alot of software additions are unnecessary and simply be a free add ons if required.
We need to re-evaluate why we use software and go back to basics. Quit adding stuff that forces users to unlearn usage just because a company feels like it
xavier
That is why I still use Word 2010.
I use Wordperfect 21 withb18-20 as backup.😎
xavier
“Because somewhere along the line, software companies started by nerds in garages appointed themselves the Internet Police. Under the guise of protecting users from misleading information, the billionaires in charge of these megacorps have launched a crusade to suppress ideas they disagree with. And since their insular worldview clashes so brazenly with reality, the results are ever clumsier and weirder. ”
I will add, that you might find the book, Surveillance Valley, by Yasha Levine, to be a useful read in that regard. Because all manner of people involved in the early years of the Internet, were involved in various groups that have all kinds of ties to folks who have been systematically making things worse for decades.
Good point. Look at how many of the former Twitter workers Musk fired were FBI agents.
There was a point where Microsoft removed spellcheck from the web version of outlook. In fact, you might still not have it to this day, due to A/B testing meaning that different users have different versions of the software. If you go to the Microsoft support forums, you will find people from just a few months ago saying it doesn’t work on their version of outlook, and being told to fix this by selecting options that do not appear for them. Even if it is working for you, there was a period for a few months where it apparently worked for no one, and Microsoft’s stock response was “we’re working on implementing that in a future update!”
It really does capture everything wrong with modern tech “innovation.” A feature that worked fine for over a quarter of a century removed for no apparent reason, with no option to keep it. Even when it is reimplemented, it is worse than before, doesn’t work for everyone, and has no fixes for when it breaks. Keep in mind that they have moved to a subscription/software as service model, meaning that people are PAYING for companies to break features in their products. And it’s not like we’re getting mind blowing new features to make up for these defects; all we get are redesigned interfaces for the illusion that something is actually improving.
There really is no reason not to use stable versions of old software. Winamp may now be some lame spotify ripoff with the old player completely abandoned, but that doesn’t mean I can’t still use the version I’ve transferred between my last 4 computers to listen to music. Even when software breaks due to new operating systems, PC emulation has gotten pretty great through the XP era so you don’t even have to go to retro hardware.
My main music player is a 12 year old version of iTunes (iTunes 10.6.3.25, to be specific.) It has the features I want, and none of the bloat.
You missed the huge problem of the Internet itself, specifically in its current form where everyone, everywhere can for whatever or no reason, can and is encouraged–even required–to be on the same Internet. The current Internet is open borders for computers and everything is partly downstream from that.
Because your computer is connected to the Internet, it is connected to every cracker, scammer, ne’er-do-wel, and demonically infested computer in the world. Software developers are in a constant battle vs malefactors at every level of the stack–network, operating system, drivers, libraries, and their own code. And, whenever there is a change anywhere else in your stack, the changes have to be made to compensate.
But these are just security updates, why does my application have to suck? Updates have to be paid for, paying developers for updates requires new customers, attracting new customers requires your product to have at least the appearance of being current and new. Sometimes new hardware, useful new features, or changes lower in the stack require new interface designs. All the rest is an exercise in making the dev teams appear useful to the MBAs in between when they are obviously needed.
Search engines are a special case of this. The worsening utility of Google, et al., is entirely adversarial. It’s a multi-way fight between:
– The information retrieval specialists who want to get the best results for the searchers.
– The censors who want to remove bad-think results for the searchers.
– The companies who want to convert searchers into customers.
– The scammers who want to convert searchers into victims.
– The spammers who want to flood certain results to the searchers.
Are the IR specialists inferior to the ones who came before? I couldn’t say; not my specialty. Their adversaries, however, are stronger than ever.
The novelty seeking is far less bad in open-source software, which has similar adversarial pressures but doesn’t have the mandate to make the big line go up.
Yes, and the same holds true for A.I. development, where the open-source LLMs are running rings around the brain-damaged corproate stuff.
“Many of the original programmers who built the foundations of today’s software and search engines have retired or moved on without passing down their expertise to the next generation.”
Same thing happen to Gamesworkshop
Warhammer 40K had the same people working from 3rd Edition to the end of 4th edition (Just leaving once 5th Edition came out). Warhammer 40k (the miniature game) I think was worse off. The game got too complex and the new codex’s and army become too unbalance, leading to them rebooting it for 8th edition. Now it is complex (hard for new players) and very unbalance.