Start coding at the point of least certainty

Where do you start work on a new project? At your desk, ha! But what do you work on first?

The obvious answer is to start at the beginning.

Build the foundation, write a placeholder page, make a hello world, then add the rest. This works. Easy task, gets the juices flowing, sets you up for future work.

But what if you aren't sure? What if there's a risk your project can't be done? What if you're training monkeys?

Start with that. Eat the frog as they say.

Prototype the least certainty first

My favorite approach is to build a quick and dirty prototype of the hardest part of the project. Like when I started exploring OpenAI and ChatGPT – I didn't start with the routing navigation, beautiful styling, or even a webpage at all.

The first experiments looked like this:

Can OpenAI do what I want? What do API calls look like?

Then I tried out what Async React with NextJS 13 feels like. Can the new approach work for this? Do I like it? What happens when you summarize a whole bunch of papers at once?

Looks like shit but answers my question: this can work.

But why make it hard?

Starting easy builds momentum and that's fantastic productivity advice that works. Ask anyone.

Where it breaks down is training monkeys in a circus.

You know you can build a platform for monkeys to stand on. Plenty exist. You know you can build a flaming hoop for them to jump through. Plenty exist. You even know you can get a monkey.

But can you train the monkey to jump through the hoop? 🤔

Without the trained monkey, you have no spectacle. Nobody's gonna come watch a bunch of monkeys standing around a fire.

All that time spent building the platforms and the flames – wasted. Wouldn't you rather do the hard thing first?

Cheers,
~Swizec

PS: this analogy comes from the CEO of Google X, a research lab, via Quit, a great book by Annie Duke

I didn't even realize that was my approach until she spelled it out across 3 chapters 😆