A Lever, Not a Replacement: Growing Your Talents in Code and Math With AI
There's a line traditionally attributed to Archimedes: give me a place to stand and a lever long enough, and I will move the world. I keep coming back to it when people ask me whether AI is going to replace them. A lever doesn't replace the person pushing it. It multiplies what they bring.
A lever needs someone to push
That's the part of the metaphor I don't want us to miss. A lever with no one behind it moves nothing. The same is true of AI tools for code, math, and quantitative work. They can multiply your effort enormously, but they multiply what's there. Bring curiosity, skill, and good questions, and the tool takes you further than you could go alone. Bring nothing, and you get confident-sounding output you can't evaluate.
So the question I'd rather ask isn't "Will AI replace me?" It's "What am I bringing to the lever?" That's a question each of us gets to answer, and the answer can grow.
The parable of the talents
In the Gospel of Matthew (25:14–30), Jesus tells a story about a master who entrusts his servants with talents, a large sum of money in that day, before going on a journey. Two of them put what they were given to work and grow it. The third, afraid of losing it, buries his talent in the ground. When the master returns, he celebrates the two who used what they were given, and he's disappointed in the one who hid it.
I've always read that story as being about more than money. Each of us has been given gifts. For some people it's a head for numbers. For others it's patience with code, or an ear for music, or a knack for explaining hard ideas simply. I believe those gifts come from God, and that they're meant to be used and grown, not buried out of fear.
Fear of automation can tempt us to bury our talents: to stop learning because "the machine will do it anyway." I think the better response is the one the faithful servants chose. Put the gift to work. Use every good tool available, including AI, to grow it.
Using AI to grow your skill in code
- Try it first, then ask for a review. Write your own version before you ask a model for one. Then ask it to critique yours. You learn far more from the gap between the two than from a finished answer.
- Read every line before you run it. If you can't explain what a line does, that's your next thing to learn, not something to skip.
- Ask for tests, not just code. Have the tool help you write tests that would catch a mistake. Tests turn "it looks right" into "I checked."
- Explain it back. When you think you understand something, explain it to the model in your own words and ask where you're wrong.
Using AI to grow your skill in math
- Struggle a little on purpose. Spend real time on a problem before asking for help. The struggle is where the understanding forms.
- Use the tool as a checker. Ask it to verify a step of algebra, search for a counterexample, or test a conjecture on small cases. Then verify what it tells you.
- Ask "why," not just "what." An answer you can't justify isn't yours yet. Ask for the reasoning, and ask for a second approach to the same result.
- Keep a notebook of your thinking. Write down the dead ends and the reframings. As I wrote in The Process Is the Product, the process is the part no one else can write for you.
Using AI to grow your quantitative judgment
The most valuable quantitative skill isn't computing. It's judgment: framing a question so it can be answered, noticing when a number doesn't make sense, and being honest about how certain you are. Tools can run the model. You decide whether the question was right and whether the answer is believable.
A good habit is to ask every result three questions. Does it pass a rough sanity check? How would I know if it's wrong? How uncertain is it? That last one is the subject of my new companion piece, Honest About Uncertainty, a plain-language guide to prediction intervals.
The creative process is a gift too
For me, building something, whether it's a model, a proof, a song, or a small game, is one of the places I feel most connected to God. There's a quiet joy in the moment an idea finally clicks. I don't want to hand that moment away. I'd rather use tools that let me have more of those moments, and reach further when I do.
That's how I try to use AI: as a way to spend less time on the tedious parts and more time on the parts that require a person, the questions, the meaning, and the care.
We grow faster together
None of us grows our talents alone. I learned from teachers, classmates, colleagues, and people I've never met who posted clear explanations online. If you're learning code, math, or data science, find people to learn with. Share your work in public, even when it's small. Ask questions. Answer someone else's.
If you're working on something and want a second set of eyes, I'd love to hear about it. And if anything on this site has helped you along the way, the support options below are how these articles, games, and music stay free for everyone.