A few months ago, I inherited my grandmother's sewing machine. It's beautiful – entirely metal, older than I am, and somehow still works perfectly. It feels like the kind of object that was built to outlast the people who owned it. The only problem was that I had absolutely no idea how to use it.

So last week, I signed up for a beginner sewing class.

I made my first project: a butter-yellow corduroy coaster. It was crooked in places, the stitching wasn't perfectly straight, and it took me much longer than I expected. I loved every minute of it.

At one point, someone in the class asked why I decided to learn to sew.

Isn’t it obvious?! So much of my day is spent in front of screens. My fingertips spend more time on a keyboard than touching almost anything else. I spend hours talking with people through video calls and, increasingly, asking questions of AI.

I wanted to make something with my hands. I wanted to learn from someone standing beside me instead of something responding from a chat window.

It wasn't because one is better than the other. I use AI every day, and I genuinely love it. It helps me think, write, research, and explore ideas faster than I ever could alone. But sitting in that sewing class reminded me that there are some parts of learning that aren't really about getting the right answer.

Throughout the class, our instructor would demonstrate a technique – how to place pins, where to mark a seam, how to finish a top stitch – and then she'd casually add, "Sometimes I just do it this way because I like it."

It felt like someone had given me permission to stop searching for the one right answer.

I expected sewing to have a correct way to do things. Instead, I found someone with enough experience to know there were several good ways. She wasn't breaking the rules. She had mastered them well enough to know where they ended and her own judgment began.

It reminded me of one of psychology's most famous experiments.

In the 1950s, psychologist Solomon Asch asked participants to complete a simple line-matching exercise. The correct answer was obvious. But when everyone else in the room confidently chose the wrong line first, many participants ignored what they could plainly see and agreed with the group instead.

The experiment wasn't really about intelligence. It was about something much more human. When confidence enters the room, our own judgment often leaves.

I don't think AI creates that instinct. I think it amplifies it.

I personally have always looked up to experts. We've always deferred to confident voices. We've always wanted someone to tell us the right answer. AI just happens to be the one of the most knowledgeable, articulate, and accessible authority many of us have ever encountered.

But there's a subtle danger in anything that sounds consistently confident. Not because it's always right – it isn't – but because confidence has a way of quieting our own thinking.

The real risk isn't that AI gives us answers. It's that we stop asking questions.

The questions that matter are often the simplest ones: Why? Says who? What would make this wrong? Those tiny questions are where critical thinking begins.

My sewing teacher never claimed her way was the way. She'd show us several options and then shrug, "Sometimes I just do it this way because I like it." I realized she wasn't teaching us to memorize techniques. She was teaching us how to develop judgment.

AI is remarkably good at generating possibilities. But eventually, someone still has to decide which approach fits the situation, which tradeoff they're willing to make, which answer reflects their values, and which one simply feels right. Those aren't questions AI can answer for me. They're questions that shape who I become.

I've started to think there are two kinds of knowledge. The first is knowing how something is usually done. The second is deciding why you want to do it differently.

AI is extraordinary at the first.

The second belongs to us.

My grandmother's sewing machine taught me how to make a coaster. My sewing teacher reminded me that learning isn't just collecting techniques. It's developing the confidence to choose among them.

In a world where technology can generate almost infinite possibilities, our judgment becomes even more valuable.

That's the last part of thinking I hope we never outsource.