What if today’s technology had existed in 1850?
A thought experiment we return to constantly, because it separates two things that are easy to confuse: making an old system faster, and designing a different one.
Imagine having all of this — and none of today’s industries.
- Internet
- Smartphones
- Cloud computing
- AI
- Digital identity
- Digital payments
- Global communication
- Modern logistics
- Databases
- Computing infrastructure
Now remove the rest: no incumbent industries, no established processes, no institutions whose shape was set by the constraints of an earlier century. Only the human needs, and these capabilities.
Would we rebuild today’s systems and then digitise them? Almost certainly not. We would ask two questions in order.
What is the actual human need?
Not the process that currently serves it. Not the product category that grew up around it. The need itself, stated plainly.
What is the best system for satisfying it with what we now have?
Designed against the capabilities that exist, not inherited from the ones that used to.
The same mistake is available in 2045.
It is tempting to treat this as a historical observation — a mistake the world already made, between roughly 1995 and now, which we can learn from and move past. It is not only historical.
Run the experiment forward. In 2045, machine intelligence is far more capable than today and physical automation works. Someone looks at a fragmented economy and concludes it is behind, so the advanced technology should be deployed into it. If the system around that technology has not been redesigned — the records, the markets, the enforceability, the capital, the local interfaces — then the deployment reproduces the existing outcome with much better equipment.
The legacy trap is not only digitising the past. It is also deploying the future into an old architecture.
Legacy trap, first form
Legacy trap, second form
Digitisation and redesign are not the same move.
Both begin with a real-world process and end with software. What happens in between is entirely different.
Old world
Future 2100
This is not an argument against digitisation. Reducing friction is genuinely valuable, and a great deal of it still needs doing. It is an argument about what it does and does not achieve: a faster path through a structure leaves the structure intact, including whatever the structure was getting wrong.
We design backwards from the outcome.
The company does not begin by asking what app to build. It begins by asking what real-world outcome the system should produce — and then works back through everything that outcome requires.
- Future outcome
- The end state in the physical world, stated without reference to any product. Not “a better app” — a described condition that either holds or does not.
- Complete ecosystem
- Everything that outcome actually touches, including the parts nobody has built software for and the parts that are not digital at all.
- Actors + economic flows
- Who participates, what each one wants, what each one risks, and where value and money genuinely move — as opposed to where the org chart says they do.
- System gaps
- Where the current arrangement loses value, blocks entry, breaks trust, or fails to coordinate. This is where the design work is.
- Capabilities
- What the system must be able to do to close those gaps, expressed independently of any implementation.
- Technology
- The tools selected to deliver those capabilities. Chosen here, late, on merit.
- Product / interface
- The surface a human touches. Important, visible, and downstream of every decision above it.
The failure mode is becoming a collection of existing categories.
The path of least resistance
- Existing world
- Existing categories
- Existing institutions
- Technology
- Digitisation
Future 2100
- Human reality
- Future outcome
- System redesign
- New capabilities
- Technology
- New economic infrastructure
We don’t want to build the digital version of today’s world. We want to know what the world could look like if its systems were designed with today’s capabilities and tomorrow’s horizon in mind.