Skip to content
Ryan James
Menu

← Newsletter4 min read

The Cost of Staying Comfortable

Maybe We Are Losing Part of Ourselves with Complacency

A year ago, then two years ago, then now. Think about what the public AI models could do at each of those points. ChatGPT, Gemini, Claude: these are the front runners, and what they can do for anyone with a login is mind-boggling. Researchers have been talking about these capabilities for years, but now they’re packaged up and available to everyone. We can let a model run on a huge task, analyze data, build things, and make connections we never thought we’d get to. Work that used to take 6 to 12 months of getting the right people in the room, and still came with user error and biased input, can now happen in a fraction of the time.

I work in tech, so part of me is all in on that. But I also work in a school, and that’s where I get torn. Because I think we’re getting rid of the struggle.

Before I got into technology, I was a math teacher, and I learned that the struggle is where learning happens. As I got better at teaching, my study guides got harder and harder. I noticed that kids didn’t put much effort into homework, even though homework was a safe place to make mistakes. But put something hard in front of them right before a test, and they were willing to go through the struggle. So I made the study guides ridiculous. A few “thanks for playing” questions, but mostly problems they’d have to look things up to solve. It was a confidence destroyer, I’ll admit that. But I never made the test as hard as the study guide, and my standardized test scores kept climbing. The kids even came up with their own mantra: the study guides are ridiculous, but if you do the study guide, the test is easy.

It’s the same reason a view means more when you hike to it. Drive to a mountain overlook and it’s pretty. Hike up and it’s the same location, the same view, but it hits completely differently, because you had to overcome something to earn it. The things I remember most in my life are the things I struggled with. Sometimes I got through them, and sometimes I didn’t. I remember all of it.

So here’s what worries me. AI is great for complex reasoning we couldn’t do on our own, or for finding patterns in data sets no human could. But I think we’re sliding past that into “I don’t want to think about this, so I’ll let AI do it.” That’s complacency, and it’s comfortable.

And I don’t think the struggle we need is fitting everything into a calendar, or rationing our tokens. As someone in technology, I’m maybe on the verge of saying we need to pull back a little, so we don’t give up something that’s a real part of being human.

A student told me a few years ago that they avoid stressful situations because stress is bad for you. That blew my mind, because stress is good for you, in the right dose. When you work out, you’re putting stress on your body, and your body responds. Without it, you aren’t healthy. The same goes for our brains and our emotions. We need to keep stressing them, not to the breaking point, but to a deliberate point.

When I’m out on a bike ride or a trail run for two or three hours, I hold my heart rate around 135 to 140. For me, that’s right near the top of my aerobic zone. It’s not an all-out effort, but it’s a sustained one. I stay in that uncomfortable-but-safe place for hours, and then I come out of it and recover. The question I keep asking is whether we’re still doing that in the rest of our lives, or taking the easy road out.

I honestly don’t know the answer. I’m not saying jump into things you’re not ready for. There’s a safety to it. Find your own version of that zone, push yourself there for a set amount of time, then come out of it into a safe place where the stress is gone. What that looks like will be different for everyone.

So what can educators do?

I don’t have all the answers, but here’s what I’d try. The process matters more than the deliverable, so reward it that way.

  1. Grade the struggle, not just the answer. Give real points for drafts, revisions, and notes on what a student tried. If the final product is worth 100 points, make sure the process is worth enough that skipping it costs them.

  2. Make “where I got stuck” part of every assignment. Before they turn something in, ask for a few sentences on where it got hard and what they did about it. Students who skip the struggle have nothing to write there.

  3. Let mistakes earn credit. Offer corrections and revisions for points back. That tells kids a wrong first attempt is part of the work, not the end of it.

  4. Do what I did with study guides. Make practice really hard and grade it for effort and completion, then make the assessment fair. Kids will push themselves when the stakes for failing are low and the payoff is real.

  5. Have them talk about it. Two minutes of explaining their work out loud, to you or a classmate, shows you the thinking behind it faster than any finished product.

  6. If they use AI, make them show the process. Have them try first for a set amount of time, then show what they asked the tool and what they did with the answer. The goal is for AI to be a thinking partner after the struggle, not a replacement for it.

None of this is a formula. But if we value the climb more than the view, students will start to want the climb. And I hope the rest of us will do hard things for the sake of doing them, for the sake of growth, instead of reaching for a tool to spoon-feed us answers.

Keep coding!
Ryan

This post first went out on Substack. Join the conversation there or subscribe to get the next one by email.