Technology is supposed to make companies more capable. It automates processes, connects systems, turns data into information and, increasingly, helps people make decisions faster.
But there is a paradox: a company can add more software, more automation and more AI while gradually losing something just as important: the ability to understand and control the technological system it depends on.
The reason is simple: complexity compounds.
A new technology rarely exists on its own. It connects to what was already there, creates dependencies and introduces decisions that will have consequences later. One platform needs information from another, an integration is built, a new database appears and eventually another automation or AI layer is added.
Each decision may make sense individually. The problem appears when they accumulate.
What started as a collection of useful solutions can become a system of platforms, providers and integrations that is increasingly difficult to understand as a whole.
The company has gained technological capability, but it may also have lost part of its freedom to change.
Complexity is not the enemy
Complexity itself is not necessarily a problem. Factories, energy systems, batteries or mobility infrastructures are complex by nature. Sophisticated problems often require sophisticated technology.
The important distinction is between inherent complexity and accumulated complexity.
The first comes from the problem itself. The second often comes from years of adding tools, integrations and workarounds without reconsidering the system as a whole.
This becomes visible when something needs to change.
Can a provider be replaced? Does the company understand which systems depend on a particular component? Can the software evolve without rebuilding everything around it?
If those questions become difficult to answer, complexity is no longer only a technical issue. It becomes a business constraint.
A company does not need one person who understands every line of code, but it does need a sufficiently clear mental model of the technology running its business.
When that understanding disappears, technological capability can gradually become technological dependency.
AI is changing the bottleneck
Generative AI makes this discussion even more relevant because it is reducing the cost of creating technology.
Code can be written faster, prototypes can be built in hours and processes that once required considerable effort can now be automated much more easily.
That is a major opportunity.
But making technology easier to create does not make it easier to understand.
If building becomes cheaper, companies will naturally build more. And everything that is created still has to be integrated, validated, secured, maintained and eventually changed.
The bottleneck is therefore shifting.
For years, one of the main questions in technology was:
Can we build this?
Increasingly, another one matters just as much:
Should we build this?
The same applies to AI itself. An LLM may be the right solution, but the problem may instead require forecasting, Machine Learning, a semantic model, deterministic software or no AI at all.
The objective should not be to add as much intelligence as possible. It should be to use the right intelligence for the problem.
Understand first, then build
This is where first-principles thinking becomes valuable.
Start with the problem rather than the tool. What actually needs to be solved? What information matters? What should be automated? What does the company need to keep under its own control?
At Biotz, this is the way we try to approach technology: understand first, then decide what deserves to be built.
Sometimes that means sophisticated engineering. Other times, the better decision is not to add another layer.
As software and AI become easier to create, simply being able to build more will become less distinctive.
The real advantage may increasingly lie in knowing what deserves to be built, what should remain simple and what a company cannot afford to stop understanding.
Because complexity compounds.
And technology only remains an advantage while you are still able to control it.
.jpg)
