Linus Torvalds, legendary coder and creator of the operating system Linux, recently spoke out about the dangers of coding with AI.

The upshot? AI is a tremendous tool for coder productivity, but used without understanding of the code it’s writing is dangerous/bad.

It’s hard to argue with that logic.

It’s like asking a room full of people: “Do you believe children should be safe?” Not one person in a million is going to answer no to that question.

But black boxes (things we use but don’t understand how they work) are all around us daily.

If we take an airplane to a conference, do we know all the systems keeping us in the air?

When we drive a car, do we know the intricacies of the internal combustion engine?

When a teenager grows a following on TikTok, do they have any clue how the algorithm works or how their data is being used?

When any of us signs the terms & conditions of any app, do we really know what we’re agreeing to?

Our whole society runs on black boxes. And the thought of coding “blindly trusting AI” is scary. But the reality is, the kind of security issues AI presents were already present in many other forms. Even coding by hand, we have to trust that our compiler isn’t taking our human code and injecting something malicious when it translates our work into machine language, something even the most skilled real programmer doesn’t understand.

I get it. It’s terrifying.

But the reality is? This ship sailed the first time our ancestors used a tool.

It’s what we do. We invent things, and we become completely, hopelessly reliant on them. How else do you think we got into this mess?