In short
AI is fundamentally changing software development. Code, documentation and prototypes can be made much faster. But the most important change is not that AI writes code faster. It is that writing code is becoming less and less the scarce part of IT.
The questions that matter
What remains are the harder questions:
- What should we actually build, and which business problem does it solve?
- Is software the right solution at all, or does something already exist?
- What should be done with AI, and what definitely should not?
- Which supplier fits, and how do we avoid dependency?
- Which data are we allowed to use, and how do we secure it?
- What does it cost on an ongoing basis, and what does it deliver?
What that means for your IT
The value is shifting to people who understand the business, have an overview of the technology and make good choices. That is why we do not want to join the table only once someone has already decided "we need application X", but earlier: "we have a problem or an opportunity here; what is the smartest way to act on it?"
The answer may be software. Or AI. Or an existing product, a different supplier, or a simpler process. And sometimes the best IT advice is: build nothing at all.
What to watch out for
- Start with the business goal, not with the technology.
- A technically beautiful solution that delivers nothing is not a good investment.
- Record choices and their reasons, so they can be understood and revisited later.
Would you like advice on software development with AI? Request a free advice conversation.
Updated on