Why We Always Use New Technology To Do Old Things (And Why That's a Mistake)

The Pattern We Never Notice

In the 1890s Uriah Smith faced a problem: automobiles had recently been invented and there was concern about them scaring horses off the road in ways that would endanger the riders of horses or the carriages that they pulled. Automobiles of the time made lots of noise, did not drive perfectly straight, were a bit jumpy, and had all sorts of other issues, many of which would scare a horse. Smith, himself, was a bit of a generalist, being a writer, poet, inventor, and preacher. He had multiple prior inventions, including an artificial leg that had moveable joints. He put his mind to the task and patented one of my favorite inventions: Horsey Horseless, not to be confused with everyone's favorite submarine, Boaty McBoatface.



We, unfortunately, do not have any evidence that Horsey was ever created, we only have the patent filing. However, Smith envisioned this as a solution to the current problem. If we make the automobile to look like a horse, then it won't scare other horses quite as much. While hilarious in retrospect, this may have seemed like a good idea at the time.

This approach is what's known as skeuomorphism. A skeuomorph is an item that's designed to resemble something that came before it. Sometimes this is done as an aesthetic sensibility, and other times it happens due to a limitation in our imagination. The first cars all resembled, and were called, horseless carriages. At the time we struggled to imagine vehicles taking any format other than a carriage, only now with an engine instead of a horse. When we delete files, we drag them into the digital trash can, another example of skeuomorphism. We struggled to imagine what it meant to delete a file, so we designed that function according to an image or action that we all knew.

The tendency for us to create skeuomorphs is quite reasonable - making a new thing is already a stretch of the imagination and we want that new thing to feel familiar and get adopted by a user, so we make it similar to something old. From this we get lightbulbs shaped like candlesticks, virtual notepads shaped like sticky notes, and Horsey Horseless. Today's AI tools often follow the same pattern: "smart search engines," "better autocomplete," or "automated assistants" that replicate human workflows rather than imagining entirely new ways of working.

When The Familiar Becomes Permanent

But here's where skeuomorphism becomes dangerous. What starts as a reasonable attempt to make new technology familiar can lock us into suboptimal solutions long after the original constraints disappear. This leads us to another concept where things get trickier - path dependence.

Path dependence describes how past decisions can sometimes constrain future possibilities. Paul David, the economist, has a great, short paper about the history of the QWERTY keyboard (https://www.jstor.org/stable/1805621). The short of it is that there's a better keyboard layout, the Dvorak layout. The QWERTY layout came from a technical problem in early typewriters that caused them to jam and the QWERTY layout avoided that problem. However, as typewriters improved, the problem ceased to exist, yet we still use a QWERTY keyboard layout. This is simply because everyone had already learned to use the QWERTY layout. Switching costs and network effects create lock-in with new technology.

So we created a solution to a temporary problem and that solution has impacted how keyboards are designed for the rest of time. As David writes, "competition in the absence of perfect futures markets drove the industry prematurely into standardization on the wrong system." This is the dark side of skeuomorphism; the familiar can become permanent, even when it's wrong.

This sort of user lock-in that we see with QWERTY represents a broader threat to innovation. Making the wrong decision early in a tech's lifecycle can put organizations on the path to obsolescence. We see this pattern playing out with AI today: companies rushing to implement "AI-powered" versions of their existing processes without questioning whether those processes make sense anymore.

The Innovation Trap

Understanding this dynamic brings us to Clayton Christensen's Innovator's Dilemma, which reveals why breaking free from skeuomorphic thinking is so critical. Christensen showed that disruptive technologies start by serving existing markets in familiar ways, but the real transformation happens when companies stop trying to replicate the old and start imagining the new.

Netflix is the perfect example of this progression. They didn't start by trying to "disrupt" entertainment - they began with a skeuomorphic approach: mail-order DVDs that replicated the Blockbuster experience but more conveniently. This familiar model helped them gain adoption. But Netflix's genius was recognizing when to break free from that skeuomorph. They evolved from "better Blockbuster" to streaming service to content creator, each step moving further away from replicating existing models and closer to defining entirely new ones.

Compare that to companies that got stuck in the skeuomorphic trap. Kodak invented the digital camera but couldn't move beyond thinking of it as a "film replacement." Newspapers created websites that looked exactly like print editions. Taxi companies built apps that were just "phone dispatch, but digital."

These three perspectives - skeuomorphism, path dependence, and the Innovator's Dilemma - all illuminate the same fundamental challenge: our tendency to use new technology to do old things, even when that approach limits the technology's transformative potential.

Breaking Free from the Pattern

The most common approach is to use the new thing to do an old thing better. The internet comes out and it's a digital mailbox, library, and newspaper. It takes us time before we start thinking about social networks, e-commerce platforms, or streaming services, the truly new things that are possible because of the internet. Some companies build their entire strategy around using new tech to optimize old processes. That approach is usually successful for a time, but it often eventually fails unless they can break free from that initial skeuomorphic thinking.

Being stuck in skeuomorphic thinking is a trap, though it feels like safety. As a leader you might think that your organization shouldn't jump right into some new tech, and certainly you need to make your tech investments wisely. But if there's value in a new technology, that value likely extends beyond simply optimizing old processes. Making a company-wide decision to use new tech only for familiar purposes can set you up for path dependence that represents a significant opportunity cost.

However, there are ways to overcome this type of thinking. As you examine new technologies, the key is asking yourself what constraints from the old system no longer apply. What would this process look like if we designed it from scratch today? What are we doing this way only because that's how it's always been done? If our biggest competitor disappeared tomorrow, how would we structure this differently? Every business has established procedures, policies, and organizational structures that work a certain way. Do they work that way because of old constraints that this new technology overcomes? If so, then it's time to fundamentally re-evaluate that process rather than just making it more efficient.

Our default response is always to replicate what has come before and make it more efficient. Making things more efficient is good! But don't fall into the trap of optimizing your business for legacy processes with legacy customers while the real innovators create new markets and customer categories, leaving you behind.

The Current Moment

Every transformative technology tends to follow this pattern. We think of how we can use new tech to solve current problems. We build Horsey Horseless and sometimes make ourselves obsolete. The real opportunity isn't in doing old things better - it's in doing entirely new things. The winners in any new technology wave are those who move fastest from mimicry to genuine innovation.

Take generative AI as a current example. Most organizations are implementing AI-powered customer service that handles the same inquiries human agents handled, intelligent document processing that digitizes the same forms we've always used, and automated content generation that produces the same marketing materials we've always created.

These applications aren't wrong, but they're skeuomorphic. They're taking AI and using it to do familiar things more efficiently. The companies that will define the next wave are those asking different questions: What if customer service wasn't about handling inquiries but preventing them? What if content wasn't something we create but something that creates itself based on real-time user needs?

The companies that understand this pattern and deliberately break out of it first will define the next wave of innovation. They'll move beyond asking "How can we use AI to do what we already do?" to "What becomes possible when intelligence is abundant and cheap?" Maybe that means customer relationships that adapt in real-time, or business processes that redesign themselves based on outcomes.

My challenge to you is to examine how you're implementing emerging technologies. Are you using them to make old processes more efficient? That's better than doing nothing, but if that's your entire approach, you might be setting yourself on a dependence path that doesn't lead forward. The winners will be the companies that use the familiar as a starting point, not an ending point - those who recognize when it's time to move beyond familiar patterns and build something genuinely new.