Prototyping with LLMs

Did you know that Jesus gave advice about prototyping with an LLM? Here’s Luke 14:28-30:

Suppose one of you wants to build a tower. Won’t you first sit down and estimate the cost to see if you have enough money to complete it? For if you lay the foundation and are not able to finish it, everyone who sees it will ridicule you, saying, ‘This person began to build and wasn’t able to finish.’

That pretty much sums me up when I try to vibe a prototype.

Don’t get me wrong, I’m a big advocate of prototyping.

And LLMs make prototyping really easy and interesting.

And because it’s so easy, there’s a huge temptation to jump straight to prototyping.

But what I’ve been finding in my own behavior is that I’ll be mid-prototyping with the LLM and asking myself, “What am I even trying to do here?”

And the thought I have is: “I’d be in a much more productive place right now if I’d put a tiny bit more thought upfront into what I am actually trying to build.” Instead, I just jumped right in, chasing a fuzzy feeling or idea only to end up in a place where I’m more confused about what I set out to do than when I started.

Don’t get me wrong, that’s fine. That’s part of prototyping. It’s inherent to the design process to get more confused before you find clarity.

But there’s an alternative to LLM prototyping that’s often faster and cheaper: sketching.

I’ve found many times that if I start an idea by sketching it out, do you know where I end up? At a place where I say, “Actually, I don’t want to build this.” And in that case, all I have to do is take my sketch and throw it away. It didn’t cost me any tokens or compute to figure that out. Talk about efficiency!

I suppose what I’m saying here is: it’s good to think further ahead than the tracks you’re laying out immediately in front of you. Sketching is a great way to do that.

(Thanks to Facundo for prompting these thoughts out of me.)