How do you level up with AI without letting it quietly kill your problem-solving instincts?

This question has come to the forefront of my mind a lot recently, and the truth is... the answer isn't clear to me yet. But I'll give you some context to how this came about:

The thing that made me stop and think

For the last 4-6 weeks, I've actively started search for my next role after finding myself without a position at the end of 2025. I've personally found that just applying to jobs on LinkedIn has not been enough to get replies, so to increase those odds I've activeley been working with a career coach and doing outreach to peers and potential hiring managers. Which has meant a lot of outreach, a lot of follow upws, and I'd been manually tracking all of this in a spreadsheet which was becoming very tedious.

I realised that there must be a better way to solve this, so what did I do? Think through the problem (don't be silly!), I jumped straight into Claude and prompted it with my problem, not a solution... and even by my standards this was a very basic prompt (see below).

I'm working with a career coach and currently doing outreach to potential hiring managers and peers, tracking it in a spreadsheet but finding that it's a very tedious process to keep up with. How do I solve this problem?

After a bit of back and forth, we decided to go down the route of developing a Chrome Extension to re-disk my LinkedIn account. The extension essentially saves the most relevant details to me to a NotionDB and uses an LLM to help generate a connection note although most have been poor quality thus far. It wasn't anything groundbreaking, but it's become genuinely useful for what it is I require. First years ago my instict would have been to sit down, brainstorm the problem and break it down to all it's component pieces and map out how I was going to build it in code. I'm not an engineer by any stretch of the imagination, but I've always had a get it done mentality.

Instead, yesterday, I jumped straight into Claude. It enabled me to go from idea to functional extension in under 30 minutes. I didn't even have to troubleshoot code because Claude managed the entire process.

That's when it hit me. AI has become like my phone and keys. I don't leave the house without it.

The tension I can't shake

I'm loving the speed to action, it's like nothing I've ever experienced before. he ability to go from problem to working solution in the time it used to take me to make a sandwich is genuinely exciting. As a data analyst, it's changed how I approach code, brainstorming, structuring problems, almost everything. But I do worry about what's happening underneath.

Am I getting sharper, or am I slowly outsourcing the part of my brain that used to enjoy the struggle? There's something that happens when you sit with a problem long enough to actually understand it. I'm not sure I'm doing that enough anymore.

Where I'm actually at

Right now I'm scattered with AI. Too many rabbit holes, nothing sticking. I know I'm ahead of most people in my life - but compared to my peers and the next generation? I feel like I'm watching everyone sprint while I'm still lacing up.

The problem isn't finding more things to try, it's working out what's actually worth doubling down on. I don't think the answer is to use AI less because that ship has well and truly sailed, and honestly I wouldn't want it to. But I think there's a version of this where you stay intentional — where you use it as a multiplier without letting it become a crutch.

I just haven't fully figured out what that looks like in practice yet.

The question I'm sitting with: If you've navigated this, I'd genuinely love to know: what did you cut, and what did you keep?