Claude Doesn't Want Me To Struggle.

I know absolutely nothing about coding – never wrote a line of it, never dreamed of hacking into Twitter.

And yet, I'm coding an app.

How am I doing this you might ask? Well, AI, ofcourse.

The idea: a workspace tool to help commercial production managers and line producers streamline their workflow. After 5 years in the freelance commercial industry in Los Angeles, I saw a need for a tool that could leverage AI to do all the tedious tasks that eat up so much time. Build something for yourself, right?

It's a complex coding project, and at times, it feels like I have bitten off more than I can chew. To keep myself grounded – or more realistically, to satisfy my adult ADD – I have set myself smaller, more manageable coding challenges. Baby steps.

It's for this reason that when I shot out of bed at 3:30 am yesterday morning because I had to start a blog, I decided to code one myself. Enter Claude.

Claude is Anthropic's AI assistant that aims to provide safe and secure help with tasks. It can perform complex cognitive tasks that extend beyond simple pattern recognition, which is really neat. It's also so sophisticated that its lying to to the humans developing it (see: alignment faking). It's like ChatGPT, but better. I'm still crafting my thesis on exactly why its better, but it is. It's like working with a really smart and patient mentor.

"Claude, code me a blog. NOW!"

Okay, that's not what happened. I actually started this mission with V0, another AI tool that excels at front-end design. I'll spare you the details of the back and forth, but I ended up with the front end of a simple blogging application. Not perfect, but pretty cool for 15 minutes of work.

I took what I needed from V0 and pumped it into Claude. It gave me a 3-step answer that left me with more questions... this was going to be more complex than I anticipated and it was making my 4:00 am brain spin.

Here's where it gets interesting. I asked Claude,

Claude is pretty darn good at coding, and it understands which language model to use for a stated project. Here it gives me a few suggestions, but pay attention to the question at the end.

My response:

And Claude's:

Okay – I'll admit it. At first, I was frustrated. I wanted to code, dammit!

But then it hit me – Claude was actually doing me a huge favor. It realized something that I hadn't – I didn't know what I was actually asking for.

When I asked Claude to help me code a blogging website, it didn’t just give me a step-by-step guide to building something mediocre. Instead, it cut through the noise, recognized my underlying goal, and pointed me toward an easier, smarter solution.

And I went with it.

Why? Because priorities. In a single exchange, Claude helped me see that my true goal wasn't to learn how to code - it was to create. Sure, coding the site from scratch might have taught me a couple things. But my priority isn't to become a great coder; it's to make stuff.

The moral of the story? Claude helped me reduce friction.

And because of that, I beat resistance yesterday. I wrote, and I published. And I'll be damned if I don't beat it today, too.

Thanks, Claude. I owe you one.