“That’s Just The Way We've Always Done It!”

By Brittany Molenaar

Why the systems we inherit aren't always the systems that work, and what that means for how we meet AI.

Look down at your hands on your laptop keyboard. Your fingers are probably resting on a row that reads A S D F and J K L ;Above them sits Q W E R T Y U I O P; and below them sits [SHIFT] Z X C V, and B N M , .

“Holy Shift,” have you ever sat there and wondered why the heck the letters were organized this way? Is this a natural way to communicate? Is this the fastest, or most efficient way? But you've always typed words out on this layout, and you’ve done it millions of times without ever thinking about why, and why not? That’s the way we’ve always done it. But is it? How would you know if you’ve never once chosen it?

Nobody who studied the English language and the human hand at the same time concluded that this was the best arrangement of letters. Nobody sat there at breakfast, while surfing the web and enjoying their first cup of coffee, and thought, “You know what? I think that this is the most efficient way to nail down the written text since the rock and chisel,” or did you? Probably not, and you’ve probably never questioned whether the keyboard in front of you is a fossil. It records problems solved in the 1870s, for machines that no longer exist, by people who never imagined you.

Every laptop, phone and tablet has changed completely since then. The layout didn't.

The keyboard isn't really the point, though. It's a mirror. A surprising amount of how we build, govern and live runs on the same logic: this is how it's done, because this is how it's always been done.

A short history of the keyboard nobody chose

On June 23, 1868, the printer and newspaper editor Christopher Latham Sholes, with Samuel W. Soulé and Carlos Glidden, was granted a patent for a typewriter (Britannica, n.d.). The early machines arranged their keys roughly alphabetically (Yasuoka & Yasuoka, 2011).

Over the next few years, they shuffled the letters. The familiar story is that Sholes spaced out commonly used letters so the typebars wouldn't jam (Britannica, n.d.). Researchers Koichi and Motoko Yasuoka dispute that story. Their study of the patent record argues the layout was shaped mainly by telegraph operators, who were among the first customers and used the machines to transcribe Morse code (Yasuoka & Yasuoka, 2011). Either way, the arrangement answered the needs of a specific machine and a specific market. Your hands weren't the design brief.

The gunmaker E. Remington and Sons took on the design, and production began on March 1, 1873. That machine became the Remington typewriter, which popularized the QWERTY layout (Library of Congress, n.d.). Typing schools taught QWERTY. Offices bought QWERTY machines because that's what typists knew. Typists learned QWERTY because that's what offices bought.

That loop has a name: path dependence. The economic historian Paul David used QWERTY as his headline example in a 1985 paper. He argued that once enough typists, machines and training are built around a standard, each new adopter has more reason to pick it and switching costs keep climbing, so the standard gets "locked in" (David, 1985). The system stops being chosen and starts being inherited.

In 1936, August Dvorak patented an alternative layout. How much faster it really is? That remains to be debated. Economists Stan Liebowitz and Stephen Margolis argued that the studies favouring Dvorak were flawed, noting that Dvorak himself held the patent and was involved in the testing (Liebowitz & Margolis, 1990). That debate is revealing, in itself. Nobody argues QWERTY is optimal. The best defence available is that it isn't bad enough to be worth changing.

The typebars are long gone. The electric typewriter, the computer and the touchscreen all changed the machine completely. The layout built for that machine is still under your fingers.

That's the pattern to watch for. A solution outlives its problem, and then becomes a problem of its own.

The guitar: when the tool shapes the thinker

Pick up a guitar in standard tuning and look at the strings: E, A, D, G, B, E. Each string sits a perfect fourth above the one below it, except one. Between G and B, the gap shrinks to a major third ("Regular tuning," n.d.).

That one irregularity means chord and scale shapes can't simply slide across the neck. When a shape crosses from the G string to the B string, the notes on the B string have to move up one fret (Facoetti, n.d.). Beginners memorize the exceptions. Teachers spend lessons explaining the "B-string shift."

There are reasons for it. The five-course guitar of 1500s Italy was already tuned A-D-G-B-E, and when a sixth string was added later, it was tuned to a low E to keep that familiar pattern (Owens, n.d.). Tuning every string in fourths would give perfectly consistent shapes ("Regular tuning," n.d.), but it would also change the familiar open chords. Neither is simply right. Standard tuning is a trade-off made centuries ago that we now treat as a law of nature.

Here's the more interesting part. The tuning doesn't just affect how guitarists play. It affects what they play. Riffs, genres and whole styles grew around the shapes that fall easily under the fingers. The instrument quietly decides which ideas feel natural and which feel awkward.

Philosophers call this territory ontology: our assumptions about what exists and how things are. Our ontology comes before our problems. It decides what we even notice as a problem. A guitarist who has only known standard tuning doesn't experience the B-string shift as a design flaw. It's just "how the guitar is."

The same is true in our organizations. If you have always coordinated a project through email chains, spreadsheets and printed drawings, then late information, version confusion and double entry don't look like design flaws. They look like "how construction is." The system has become part of how you see the world, so it's invisible.

For the neighbours: traditions that cost more than they give

A keyboard layout is harmless enough. Some inherited systems are not.

Bloodletting was a Western remedy practised for about 3,000 years, from ancient Egypt until it was discredited in the late 1800s. It rested on the theory that illness came from an imbalance of four bodily "humours," and removing blood was supposed to restore the balance (Greenstone, 2010). In December 1799, George Washington's physicians drew more than 80 ounces of his blood, roughly 40% of his total volume, over about 12 hours while treating a severe throat infection. He died on December 14, 1799 (Markel, 2014). His doctors weren't fools. They were doing exactly what their training, their peers and centuries of tradition told them was right. They weren’t in this career to kill people; they were in it to help them with practices that did the opposite. That's the unsettling part.

Foot binding in China appeared during the Song dynasty (960–1279) and lasted about a thousand years (Mackie, 1996). Girls' feet were broken and bound to keep them small, causing lifelong pain and disability. It served no practical purpose. It persisted because it signalled status and marriageability, and because no single family could stop without their daughter paying the price. Societies were different then, and for a woman, survival meant finding an appropriate marriage partner, so that she wouldn’t bring on the shame and financial strain to her own family.

Women, much like those in our own Western history, were raised to be excellent helpmates, wives, and mothers. Marriage was a way for her to lift, honour, and provide safety and security for her family. Every member of the collective family had a responsibility to one another. Parents who loved their children did this to them, often because they felt they had no other choice, and how could your daughter compete when everyone else was doing it? They also felt the heavy weight of responsibility for their future posterity. What would life be for the next generation if she didn’t find a good partner? Not only that, what might happen to their own daughter when they were both gone? What other options did she have?

Let’s be real, that sucked for everyone involved, but you know what? Even during hard times, some humans can be brave and amazing, because what ended it is worth remembering. Around the turn of the 20th century, reformers formed natural-foot societies. Members pledged not to bind their daughters' feet and not to let their sons marry women with bound feet. Once a whole community switched together, nobody was left out. In Tinghsien, a conservative rural area, 99% of women were bound in 1889, and none were in 1919.

Political scientist Gerry Mackie argued that harmful conventions usually break not through one brave family at a time, but when a group coordinates to change together (Mackie, 1996).

Most inherited habits are far smaller than these. The annual event held because "we always host it." The report nobody reads, but everyone still produces. The approval chain with five signatures because one project went wrong in 1998. The effort we pour into getting these things "right" is real. The benefit often isn't. We keep doing them partly out of habit, and partly because we don't want to be the ones seen doing it differently.

"We've always done it this way" is a description, not a reason.

How do you know a system has stopped working?

Not everything old is broken. Some traditions carry hard-won wisdom, and "new" is not automatically "better." What we need is a fair test, and here are six questions to ask of any system you've inherited:

1.     Does the original problem still exist? QWERTY solved jamming typebars. There are no typebars. If the reason is gone, the solution deserves a second look.

2.     Who is it serving? Foot binding served appearances, not the girls. Many workflows serve an audit trail or a department's comfort more than the people doing the work.

3.     What does it cost to keep it "right"? Count the hours, the rework, the stress and the workarounds. When upkeep costs more than the benefit, the system is running you.

4.     Would anyone design it this way today? If you were starting fresh with today's tools, would you build this? If the honest answer is no, you're looking at inheritance, not design.

5.     Can anyone explain why, or only how? People can usually tell you how a process works. When nobody can say why it exists, it's probably running on momentum.

6.     Is fear of judgement the main thing holding it in place? If the strongest argument for keeping something is what others might think, you've found a tradition kept for the neighbours.

A system that fails several of these deserves an honest conversation. That doesn't mean tearing it down tomorrow. It means it has lost the right to go unquestioned.

Zooming out: what politics and policy teach us about systems

The same pattern shows up at the scale of whole societies, and political science offers some useful language for it.

Politics or policy: which comes first?

It's tempting to think politics comes first: we elect people, then they make policy. But it runs both ways. Policies create groups of people who depend on them, benefit from them or are harmed by them, and those groups reshape politics. The political scientist Aaron Wildavsky put it neatly: policy becomes its own cause (Wildavsky, 1979). Every solution produces new problems, and those problems become the next round of politics.

That isn't a reason for despair. It's just an honest description of how change works. The keyboard solved jamming and created path dependence. A new workflow solves double entry and creates a training need. The question is never, "Will this create new problems?" It always will. The question is whether the new problems are better ones to have.

Agenda setting: no platform, no mandate

A problem that nobody names will never get solved. Political scientists call the process of deciding which issues get the most attention “agenda setting”; my uncle sometimes calls it “picking your poison.” John Kingdon described change as happening when three streams meet: a recognized problem, a workable solution, and the political will to act (Kingdon, 1984).

In politics, you can't claim a mandate for something you never put on your platform. In organizations, you can't fix a workflow nobody is willing to call broken. Before any change, someone must put the problem on the table out loud. The entire engine has to function simultaneously to work; if one little piece goes off track, it puts the whole thing in a precarious position.

And let’s be fair here to some of our politicians, and this is my blog, so I can go a little off-track if I want to, because nobody is grading this as a paper. So, dear reader, if you don’t like this, then maybe you should stop reading here. But this needs to be said:

I don’t know if you know this, but running a country is a lot like cleaning your house in a tsunami with all the windows open. Let’s be real, my friends, it isn’t easy. Thank you to all the wonderful men and women who are brave enough to. That’s a heavy load, and holding the responsibility for the lives and welfare of Millions of people in your hands is heavy. One wrong move at the wrong time, or the right move made too quickly, can upset the entire engine that we are running on. It takes the care, effort, time, and countless hours and sleepless nights of a handful of people, compared to the rest who are governed, to ensure everyone else can sleep the next day peacefully.

I cannot imagine how hard that must be, and I am grateful that we have people willing to do it. Every day, and every moment that we have, is a miracle. Waking up the next morning is a miracle, and you can thank the ones who govern you for that. Some of these amazing people are visible and invisible at the same time. At the end of the day, most of them are just ordinary people, trying to make the world a better place and live another day: just people like me and you.

Systems theory: what is the system adapting to?

The political scientist David Easton described politics as a system. Demands and support flow in; decisions flow out; the results feedback and change the next round of demands. A healthy system keeps adapting to its environment (Easton, 1965).

The useful question about any policy or process is: what is it adapting to? If the answer is "conditions that existed twenty years ago," the system is out of step. Reform isn't only about changing a rule. It changes the context in which every future rule gets made.

Complexity and entropy

Here's the warning: Systems don't just stay out of step quietly. Each workaround, exception and patch adds complexity. The historian Joseph Tainter argued that societies solve problems by adding complexity, and that each new layer eventually returns less than it costs (Tainter, 1988).

Past a certain point, a system exceeds its capacity to manage itself. In politics, this can look like rising polarisation, gridlock and unpredictable outcomes, as institutions struggle to hold competing demands together. In a company, it looks like burnout, bottlenecks and the one person who "knows how it all works" holding the whole thing together. Keep layering patches on a structure built for a different world, and it eventually comes apart.

The lesson isn't that change is dangerous: It's that refusing to adapt is dangerous too. It just fails more slowly, which makes it easier to ignore.

So, what do we do about AI?

Which brings us to what's on everyone's mind.

AI is uncomfortable for exactly the reasons this post has been describing. It asks us to question workflows we've never questioned. It means learning new skills after years of being good at the old ones. And it will bring new problems: accuracy to check, data to protect, habits to rebuild. Every solution does.

But look at what we already accept. We draft the same documents from scratch, retype information between systems, and dig through folders for the drawing that's "definitely in here somewhere." We don't experience that as a problem. It's just how the work is. That's our standard tuning. That's our QWERTY.

Consider how every past shift felt. Moving from paper drawings to PDFs felt risky. Moving from hand calculations to spreadsheets felt risky. Moving from phone calls to email felt risky. Each one brought new problems. Almost nobody wants to go back.

The honest version of the argument isn't "AI will fix everything." It's this:

•      Be willing to ask the six questions of your current systems, including the ones that feel like "just how it's done."

•      Put the problem on the agenda. Name what isn't working out loud, so there's a mandate to change it.

•      Change together, not alone. Like the families who pledged together, teams adopt new tools best when they move as a group, and nobody is left exposed.

•      Expect new problems and judge them fairly. Better problems are still progress.

Learning something new is uncomfortable at first. Then it isn't. Then, very often, you wonder how you ever worked the old way.

The keyboard under your fingers is proof that we'll keep a flawed system for 150 years if nobody stops to ask why. The good news is that we get to ask. And sometimes, the answer is that the new way is better.

References

Britannica. (n.d.). June 23, 1868: Patent for first typewriter. Encyclopædia Britannica. https://www.britannica.com/today-in-history/June-23-1868-Patent-for-First-Typewriter

David, P. A. (1985). Clio and the economics of QWERTY. American Economic Review, 75(2), 332–337. https://ideas.repec.org/a/aea/aecrev/v75y1985i2p332-37.html

Easton, D. (1965). A systems analysis of political life. John Wiley & Sons. Summary: https://courses.worldcampus.psu.edu/welcome/padm507/001/content/02_lesson/03_page.html

Facoetti, G. (n.d.). Guitar fretboard tuning: Dealing with the G and B inconsistency. FaChords. https://www.fachords.com/guitar-fretboard-tuning/

Greenstone, G. (2010). The history of bloodletting. BC Medical Journal, 52(1), 12–14. https://bcmj.org/premise/history-bloodletting

Kingdon, J. W. (1984). Agendas, alternatives, and public policies. Little, Brown. Summary: https://en.wikipedia.org/wiki/Multiple_streams_framework

Library of Congress. (n.d.). Production on the Sholes and Glidden Type-Writer began. This Month in Business History. https://guides.loc.gov/this-month-in-business-history/march/typewriter-production-began

Liebowitz, S. J., & Margolis, S. E. (1990). The fable of the keys. Journal of Law and Economics, 33(1), 1–25. https://doi.org/10.1086/467198

Mackie, G. (1996). Ending footbinding and infibulation: A convention account. American Sociological Review, 61(6), 999–1017. https://www.jstor.org/stable/2096305

Markel, H. (2014, December 14). Dec. 14, 1799: The excruciating final hours of President George Washington. PBS NewsHour. https://www.pbs.org/newshour/health/dec-14-1799-excruciating-final-hours-president-george-washington

Owens, J. (n.d.). Standard tuning: How EADGBE came to be. Fender. https://www.fender.com/articles/setup/standard-tuning-how-eadgbe-came-to-be

Regular tuning. (n.d.). In Wikipedia. Retrieved September 24, 2026, from https://en.wikipedia.org/wiki/Regular_tuning

Tainter, J. A. (1988). The collapse of complex societies. Cambridge University Press. Excerpts:https://centerforneweconomics.org/publications/excerpts-from-the-collapse-of-complex-societies/

Wildavsky, A. (1979). Policy as its own cause. In Speaking truth to power: The art and craft of policy analysis (Chapter 3). Little, Brown. https://doi.org/10.4324/9781315130149-3

Yasuoka, K., & Yasuoka, M. (2011). On the prehistory of QWERTY.ZINBUN, 42, 161–174.https://repository.kulib.kyoto-u.ac.jp/server/api/core/bitstreams/dc434be9-80cd-499b-a984-f9fa35954c3b/content