AI has made it ridiculously easy to go from I have an idea I m building it.

But something I've noticed while building products myself:

Building one product with AI somehow turned into using five different AI tools.

One for planning.
Another for coding.
Another when the first one gets stuck.
Another for debugging.
And somehow... another subscription gets added.

AI is making us faster, but I'm starting to wonder if tool switching itself is becoming the new bottleneck.

I keep noticing this while building with AI:

There isn t really a best model anymore. There s a model you trust for a specific job.

Claude might nail the code architecture.
GPT might reason through a weird problem better.
Gemini might surprise you with a different approach.
Codex might feel more natural when you're deep in the codebase.

And sometimes, when one model confidently gets something wrong, asking another model is the fastest way forward.

Something I keep thinking about with AI builders:

Should we even have to choose the model?

Part of me wants full control, Claude for one task, GPT for another, Gemini when it fits better.

But another part thinks the tool should simply understand the task and automatically use the right model behind the scenes.

I ve noticed something strange while building with AI:

The moment I know every prompt is consuming credits, I start thinking differently.

Is this prompt worth it?
Should I try fixing this myself?
Do I really want to regenerate that?

And suddenly, a tool designed to make experimentation easier can make you more cautious about experimenting.