When to Ask for Help and When to Push Through: What Mounting a TV Taught Me About Engineering

I've spent the last two years as a junior engineer learning to ask for help. It's supposed to be the right move. You get stuck, you slack a senior, they unblock you in five minutes. Everyone wins.

Except it doesn't always work that way.

Last week I tried to mount a TV in our apartment. No experience. No drill experience. A baby who's mobile and a wife who's busy. I read the manual, felt overwhelmed, and immediately asked for help.

Here's what actually happened: my wife spent 20 minutes just understanding the context of what I'd already tried. By the time she was up to speed, she had her own stuff to do. We hadn't actually solved anything. We'd just wasted everyone's time.

So I took a break, put on music, stopped thinking about all the ways I could mess up, and just did it. One step at a time. Didn't overthink it. Finished it.

And I realized I've been doing this wrong at work too.

The Real Cost of Asking Too Early

When you escalate, the other person has to:

Load the context of what you've tried (A, B, C)
Understand your current goal
Understand the exact problem you're facing
Get back to being available for you

That's not five minutes. That's context switching and cognitive load they didn't budget for.

Meanwhile, you're not building autonomy. You're not learning problem-solving. You're not learning to stay focused under pressure.

Here's My New System

First: Spend real time on it. Not five minutes. Not until you feel mildly frustrated. Actually invest. Take notes. Look at your bookmarks. Connect dots yourself. Try different angles.

Second: If you're stuck and mentally drained, take a break. Don't ask for help while stressed. Put on music. Do something else. Come back with a fresh perspective. You'd be surprised how often you spot the answer when you're not white-knuckling the problem.

Third: If you're still stuck, escalate. But now you've bought yourself credibility and you've narrowed the problem. The other person can actually help faster because you've done the legwork.

Fourth: Don't outsource thinking to AI too early. Yes, I use Copilot. But if I feed it context and it removes things I didn't ask it to remove, I've wasted time. If I only have limited time, I can just do it myself, learn something, and stay autonomous. AI is there if I truly get stuck—not as my default move.

What This Actually Gets You

Autonomy. You're not waiting on other people.
Critical thinking. You're learning your stack, not just copy-pasting solutions.
Pressure tolerance. When production breaks at 2 AM and you have one deployment window, you can't call for help. You have to be able to solve it.
Speed. Counterintuitively, doing it yourself is often faster than explaining it to someone else.

The real skill isn't knowing all the answers. It's knowing when to push through and when to ask. And it took me mounting a TV to figure that out.