The shiny thing changes. The pattern doesn’t.
Why decades of building taught me that the technologies change, but the incentives rarely do.
Every week there is another headline telling us that technology has entered a new era.
One week it’s Ai.
The next it’s startup valuations.
Then surveillance.
Then privacy.
Then software architecture.
Then augmented reality.
Next week it will be something else.
I’ve spent most of my career building technology, and after decades of doing it, one thing has become impossible to ignore.
The technology changes.
The underlying patterns rarely do.
Looking back through articles I’ve written over the years, I noticed that many people probably saw a collection of unrelated topics. One article was about augmented reality. Another was about social media. Others focused on privacy, local Ai, software architecture, version churn, surveillance, startup economics, and decentralized communications.
To me, they’ve always been about the same thing.
Not the technology.
The incentives behind it.
I’ve spent most of my career as a builder. I’ve designed software, platforms, geospatial systems, communication tools, WebAR experiences, enterprise applications, and more recently, local Ai systems. From the outside, it probably looks like I’ve moved from one technology to another, following whatever came next.
That couldn’t be further from reality.
Builders eventually stop looking at technology the way everyone else does.
When you’ve spent years creating products that have to survive operating system updates, changing APIs, disappearing SDKs, evolving regulations, hardware failures, and shifting customer expectations, your perspective changes.
You stop asking what a platform can do today.
You start asking who controls it.
You stop asking whether an SDK is impressive.
You ask whether it will still exist five years from now.
You stop chasing features.
You start protecting architecture.
You stop being impressed by demonstrations.
You start wondering what happens after the next software update breaks everything your customers depend on.
That’s a very different way of looking at technology.
Back in 2020 I wrote an article called “The Augmented Reality Industry Is Shooting Itself in the Foot.”
Many people assumed it was about augmented reality.
It wasn’t.
Augmented reality simply happened to be the example.
The article was about incentives.
Companies were trying to own the ecosystem instead of growing it. Everyone wanted developers tied to their platform. Everyone wanted creators locked into their tools. Everyone wanted consumers inside their own experience.
The technology wasn’t the problem.
The incentives were.
A few years later I wrote “Breaking the Cycle: How Content and Consumers Are Trapped in the Con.”
Again, many readers assumed I was writing about social media.
I wasn’t.
I was writing about incentives.
Creators invested their time creating.
Consumers invested their time consuming.
Platforms positioned themselves in the middle and captured much of the value generated by both.
Different technology.
The same pattern.
More recently I’ve written about version churn and version fragmentation.
At first glance those sound like engineering problems.
They’re really business problems.
Every time an operating system changes.
Every time an API is deprecated.
Every time an SDK forces a migration.
Every time a cloud provider changes pricing.
Every time a platform vendor changes direction.
Someone pays.
Usually it’s the builder.
Those engineering hours rarely appear in project budgets. Investors don’t usually ask about them because they aren’t visible on a balance sheet. Customers rarely notice because no new feature appears on the screen.
Yet over the lifetime of a product, those invisible costs can become some of the largest expenses a company faces.
That’s why architecture matters.
Not because it’s exciting.
Because it determines whether your business can continue building when everyone else is reacting.
The same applies to Ai.
Much of today’s discussion focuses on what models can do.
I tend to ask different questions.
Who owns the model?
Who owns the data?
Who controls the infrastructure?
Who decides the rules tomorrow?
What happens if that company changes direction?
Those aren’t philosophical questions.
They’re architectural ones.
The same questions apply to privacy.
The same questions apply to surveillance.
The same questions apply to cloud computing.
The same questions apply to startup funding.
The same questions applied to augmented reality.
They’re all connected.
As builders, we inherit the consequences of decisions made by people we’ll probably never meet.
That’s why I’ve become increasingly interested in what many people consider the boring parts of technology.
Architecture.
Open standards.
Privacy by design.
Reducing unnecessary dependencies.
Building systems that remain understandable years later.
Making deliberate decisions instead of fashionable ones.
None of those topics generate standing ovations.
Nobody celebrates a system because it survived ten years without a major rewrite.
Nobody writes headlines about the company that avoided technical debt because they made thoughtful architectural decisions early.
Nobody raises billions because they carefully reduced future maintenance costs.
Yet those are often the very decisions that determine whether a company survives after the excitement fades.
Every few years there is another shiny thing.
I’ve watched mobile reshape computing.
I’ve watched VR promise to change everything.
I’ve spent years building in AR while watching the industry repeatedly get distracted by hardware instead of adoption.
I’ve watched blockchain become the answer to problems it never needed to solve.
I’ve watched NFTs explode into headlines before disappearing almost as quickly.
I’ve watched the metaverse become the future, until it wasn’t.
Now we’re living through the Ai era.
There will be another one after this.
There always is.
The mistake isn’t chasing new technology.
I’ve spent much of my career building with emerging technologies long before they became mainstream. Builders should be excited by what’s possible.
The mistake is allowing the excitement to replace the fundamentals.
I’ve watched companies pivot entire strategies because investors suddenly wanted blockchain.
Then NFTs.
Then the metaverse.
Now Ai.
Tomorrow it will be something else.
The names change.
The investor decks change.
The keynote presentations change.
The LinkedIn experts change.
The incentives barely move.
That’s why I don’t spend much time trying to predict what the next shiny thing will be.
I’m far more interested in asking what survives after the shiny thing fades.
Does the architecture still make sense?
Does the business model still work?
Can the company still afford to build?
Did they reduce dependencies or create new ones?
Did they build something durable, or something that depended on everyone continuing to believe the story?
Looking back over years of writing, I don’t think I’ve been documenting technology trends.
I’ve been documenting recurring human behaviour.
I wasn’t writing about augmented reality.
I was writing about incentives.
I wasn’t writing about social media.
I was writing about incentives.
I’m not writing about Ai.
I’m still writing about incentives.
Who owns the dependency?
Who controls the platform?
Who carries the hidden costs?
Who benefits when incentives become misaligned?
What survives after the excitement fades?
Technology evolves remarkably quickly.
Human behaviour doesn’t.
The older I get, the less interested I become in the headline of the week.
I’m much more interested in the architecture beneath it.
Because long before software fails, business models fail.
Long before products disappear, incentives reveal where they’re headed.
The shiny thing changes.
The pattern doesn’t.