I use AI a lot. It helps me write, organise ideas and get unstuck when I’m going in circles. And honestly, that has real value.
But the weird part is this: I don’t always go to AI because I’m stuck. Sometimes I go there before I’ve even given myself a chance to be stuck.
Not because I couldn’t work it out. Just because it’s easier to start with something than with nothing.
That’s where dependency begins for me: not when we use AI, but when we stop wanting to try without it.
The training we skip
The blank page was annoying, but it forced me to sit with the idea. The first bad draft was slow, but it showed me what I actually understood.
Sometimes the hard part was the training.
And I wonder if we’re skipping too much of that now.
Who owns the infrastructure
But there’s a second layer that worries me even more.
If our writing, code, learning and decisions increasingly run through a handful of private platforms, how much of this new productivity is really ours?
Prices change. Limits change. Models change. Access changes.
A workflow you thought you owned may simply be something you rent.
And that can happen quietly. APIs change. Free tiers disappear. Products shut down. Sometimes it’s just another Tuesday, and something you relied on no longer works the same way.
The balance I’m trying to find
None of this makes AI the villain. I use it. I’ll keep using it.
But I do think the dependency is worth being honest about.
AI gives individuals huge power. One person can now do things that used to take a team. That’s exciting.
But the same technology may also concentrate a huge amount of power in the companies that own the models, infrastructure and access.
The balance I’m trying to find: using AI without losing the ability to think when the tool isn’t there.
If your AI tools disappeared tomorrow, what part of your work would collapse first?