Faster horses: why you should stop adding AI and start redesigning around it
By Kris Van Hauwermeiren ·
Adding AI to old processes gives you a faster horse. The real gains come from redesigning the work itself around what AI makes possible.
There is an old line often linked to Henry Ford: if you had asked people what they wanted, they would have said faster horses. Whether he really said it does not matter. The point is sharp. People asked for a better version of what they already knew, a quicker horse, when the real breakthrough was a completely different thing, the car.
I think most organizations are doing the same thing with AI right now. They are asking for faster horses. They take the way they already work, and they add a bit of AI on top to speed it up. It feels productive, and it feels safe. But it quietly misses the real opportunity, and the numbers are starting to show it.
The uncomfortable results so far
The honest data on AI in business is sobering. In a large 2026 study by McKinsey, around 80 percent of people said AI made them more productive. That sounds great. But only about a third of organizations could point to any real impact on their profit. People felt faster, yet the business looked the same.
Other research found something even harder. A widely cited MIT study reported that about 95 percent of enterprise AI pilots delivered no measurable return. Let that sink in. Nineteen out of twenty projects produced nothing the business could actually measure. The common conclusion from this work is important: the problem is usually not the technology. It is how organizations use it.
Why adding AI changes so little
Here is the core issue. Most of our processes were designed for humans doing every step by hand. The approvals, the handoffs, the checks, the meetings, all of it was shaped around human speed and human limits. When you drop AI into that process, you make one or two steps faster, but the shape of the process stays the same. You still have the same handoffs, the same waiting, the same old structure.
That is a faster horse. The horse is quicker, but you are still riding a horse. You have not asked the bigger question: if we were starting today, with these tools, would we even build the process this way? Usually the answer is no.
Integration speeds up the old shape of the work. Redesign changes the shape.
This is the difference between integrating AI and redesigning around it. Integration adds AI to the old way of working. Redesign asks what the work should look like now that these tools exist. One speeds up the past. The other builds the future. The research is clear on which one pays off. In the same McKinsey study, nearly three out of four of the top performing organizations said they had fundamentally redesigned their workflows, compared with only about a quarter of the rest. They did not just adopt AI. They changed the shape of the work.
What redesigning around AI really means
Redesign sounds big and scary, but it starts with a simple change of question. Instead of asking where can we add AI to what we already do, you ask what should this process look like if we built it today. Here is how that plays out.
Start with the outcome, let AI carry the flow, keep people where judgment matters.
Start with the outcome, not the steps
Take one important process and write down what it is really meant to achieve, the outcome, not the current steps. A customer gets an answer. An invoice gets paid. A report reaches the client. Then look at the steps you have today and ask how many of them only exist because a human used to do everything. Many of them can disappear, not just speed up.
Let AI carry the flow, not just a task
Adding AI to one task gives you a faster horse. The bigger gain comes when AI carries the work across steps that people used to hand to each other. The handoffs, the chasing, the copying from one system to another, that is where the real time is lost. Redesign means removing those gaps, not just speeding up the boxes on either side of them.
Keep people where judgment matters
Redesigning around AI does not mean removing people. It means moving them to where they add the most, which is judgment, relationships, and deciding what good looks like. Let the tools handle the routine flow, and let your people spend their time on the parts that actually need a human. That is a better job for them and a better result for the business.
Start small, but start with the right question
None of this means you must rebuild your whole company overnight. That would be reckless. The sensible path is to pick one process that matters, and redesign that one around AI as a real experiment. Measure the outcome, learn from it, and then move to the next. This is exactly what the strongest performers do, one process at a time, rather than one risky leap.
The only thing you need to change today is the question you are asking. Stop asking how to add AI to the way you already work. Start asking how the work itself should change now that AI exists. The first question gives you a faster horse. The second one gives you the car.
This shift, from integrating AI to redesigning around it, is the heart of what we help organizations with at Binary Rabbits. The tools are not the hard part. The courage to rethink the work is. And that is where the real value has been hiding all along.
About the author
Kris Van Hauwermeiren is the founder of Binary Rabbits and a Tech Director with 25+ years of enterprise experience in digital transformation, AI and technical leadership.