My brother and I were on our regular Saturday bike ride on the Capital Crescent Trail when an e-bike flew past us.
It got me wondering:
At what point does a bicycle stop being a bicycle?
When the constraints that shaped our organizations change, what we built around them may need to change too.
It still has pedals. But add a motor, more power, and more speed, and at some point you’ve created something that doesn’t fit quite as neatly into the old category.
Yet there we all were. On the bike path.
You see versions of this everywhere.
Is a modern car primarily a vehicle, or is it becoming a computer on wheels?
What exactly is a bank when payments, lending, investing, and financial advice no longer have to come bundled together?
What is a university when teaching, content, assessment, community, and credentials can increasingly be separated?
The obvious conclusion is that traditional categories are blurring.
But the more I thought about it, the more I realized that wasn’t the interesting part.
The more interesting question was what caused those categories to exist in the first place.
And, as usual, my mind went to how organizations work.
Don’t judge. 😉
We organized around constraints
A bank didn’t end up combining deposits, lending, payments, risk management, financial advice, and other capabilities by accident.
Universities didn’t bring teaching, knowledge, assessment, credentials, research, and community together arbitrarily.
Professional services firms didn’t build organizations with large numbers of highly educated people researching, analyzing, drafting, reviewing, and advising simply because that was the natural shape of a firm.
These arrangements made sense under a particular set of constraints.
Expertise was scarce.
Information was expensive to acquire and move.
Analysis took time.
Coordination required people communicating with other people.
Decisions often had to travel through layers because information, authority, and expertise were distributed unevenly.
Those realities influenced which capabilities belonged together, which sat inside or outside an organization, how work was divided into jobs, why departments existed, and how decisions moved through management.
Many of the structures we now take for granted are, at least partly, responses to constraints.
What happens when the constraints change?
Technology has been loosening some of those constraints for decades.
AI is particularly consequential because it can change several at the same time.
Expertise that was difficult or expensive to access can become much more widely available.
Analysis that took hours or days can happen in seconds.
Information can be interpreted and synthesized without passing through as many people.
Coordination that required meetings, handoffs, and managerial intervention can increasingly happen through systems.
And some routine decisions can be made and acted on without traveling through the same organizational layers.
This doesn’t mean expertise becomes irrelevant, coordination disappears, or consequential judgment should be handed over to machines.
But enough of the underlying economics can change that something much more interesting than productivity becomes possible:
Capabilities that once needed to sit together can increasingly be separated, combined, and organized differently.
And when that happens, it doesn’t just change how an organization operates.
It can change where value is created and who captures it.
When the bundle changes, strategy can change too
Consider banking.
For a long time, many financial capabilities naturally lived inside a bank. Today, payments can happen inside another platform. Lending can appear at the moment of purchase. Investing and financial guidance can be delivered through entirely different interfaces.
Customers can use capabilities that once helped define a bank without experiencing them as banking at all.
That matters because someone else can now own the customer relationship, the data, or the most valuable part of the experience.
Universities present a different version of the same question.
Teaching, access to knowledge, assessment, credentialing, research, and community historically came bundled together. Technology has been pulling at that bundle for years. AI increases the possibility of separating and recombining some of those capabilities even further.
That doesn’t mean the university disappears. Community may become more valuable. Trusted credentials may become more valuable. Great teaching may become more valuable.
But the value of each part of the bundle can change.
Healthcare offers another example.
Monitoring that once required a visit to a medical setting can increasingly happen continuously on someone’s wrist or in their home. Some expertise can be accessed remotely. AI can help interpret information and support diagnosis. Care that once had to be organized around a physical institution can involve very different combinations of people, technology, and place.
Again, the interesting question isn’t whether hospitals or physicians disappear.
It’s which capabilities remain more valuable together, which can move elsewhere, and who creates value when they do.
Professional services may face an especially interesting version of this.
Consulting, law, accounting, and other knowledge-intensive firms have historically combined scarce expertise with large numbers of people doing research, analysis, drafting, review, and coordination.
What happens when parts of that work become dramatically less expensive and expertise becomes easier to access?
The answer isn’t necessarily that these firms disappear.
It’s that the reasons their capabilities belonged together in a particular way may no longer be the same.
That is a strategic question because when an established bundle comes apart, the incumbent doesn’t automatically retain the valuable pieces.
The same question applies inside the organization
Now bring the question inside your own company.
Finance. Marketing. HR. Customer service. Strategy. Operations.
Why does each function contain the activities it contains?
Why does expertise live where it does?
Why do particular decisions belong to particular roles?
Why does information travel through the handoffs it does?
Why do some decisions require several levels of approval?
There are good answers to many of those questions.
Some capabilities create more value precisely because they sit together. Some decisions should remain concentrated. Some boundaries strengthen accountability, manage risk, preserve institutional knowledge, or enable people to learn from one another.
AI doesn’t make organizational structure irrelevant.
But for some of those questions, there may be another answer:
Because when we designed the system, there wasn’t a practical alternative.
That answer deserves much more attention.
If we simply add AI to existing jobs, departments, workflows, and management structures, we may improve productivity substantially.
We may automate tasks. Shorten cycle times. Reduce costs. Give people better information.
All worthwhile.
But we are also making an enormous assumption:
That the organization into which we are putting AI should remain essentially the same.
What if some of the constraints that caused us to build it that way are disappearing?
Beyond “How should we use AI?”
Most organizations understandably begin with questions about use cases.
Where can AI save time?
Where can it automate work?
Where can it reduce costs?
Where can it help people do their existing jobs better?
Those questions matter. But they start with the organization we already have and ask how AI can improve it.
There is another level of questioning:
Which capabilities genuinely become more valuable when they remain together?
Which capabilities can now be separated or recombined because the constraints that once held them together have changed?
Where might that shift change who creates value, who owns the customer relationship, or where competitive advantage comes from?
And perhaps the hardest question:
If we were building this organization around today’s capabilities rather than yesterday’s constraints, what would we do differently?
The answer won’t be “change everything.”
Many of today’s structures exist for very good reasons. And new capabilities bring new constraints of their own.
But organizational structures aren’t laws of nature.
They reflect assumptions about what is scarce, what is expensive, what requires people, where expertise resides, how information moves, how coordination happens, and how decisions get made.
When those assumptions change enough, the structures built around them deserve another look.
All of that from watching an e-bike fly past me on a Saturday morning.
So much for turning off my strategy brain on a bike ride. 😉