新闻

The hardest part of designing is no longer building your solutions. It's knowing what to build and what matters the most. 

For a long time, building was the hard part. It took the time, the people and the budget. So, we organized everything around it. That is shifting and you can feel it in daily work. Code, tests, documentation, a first attempt with three different solutions: what took days now takes hours. For R&D, the speed is real and exciting, but sometimes also scary. 

AI is also cheap (still). Tools that cost next to nothing do work that used to fill up a whole week. Those who learn to use it well now will pull ahead. 

But designing itself is changing. Most engineers have always been writers, and not the best readers. One of the multiple reasons for the “Not invented here” syndrome we engineers tend to have. Now AI writes the first draft, and the engineer reads, understands, and assesses. Is it (or am I) doing the right thing? What did it (or I) miss? Is it scalable? Is it robust? How does it handle the unhappy path? Reading and understanding what was produced matters more than ever before, because a confident wrong answer looks exactly like a right one. In a safety-relevant business like ours, responsibility for quality stays with people. 

And when building gets cheap, choosing becomes everything. AI can produce a hundred options. It cannot tell which one matters to our customers and to us. That takes experience, curiosity, and a deep understanding of the problem. It will happily build the wrong thing, faster. Knowing what really matters is the key skill. 

That’s why we invest in upskilling. Engineers need to understand the fundamentals, ask the right questions, and, above all, understand the customer and their needs.  

Building with AI is cheap (still). Knowing what to build is priceless. 

Klims Saikins