Overcoming the AI Inertia

Introduction

I have heard it for a while: “You’ve got to learn AI.” “AI is going to reshape your job.” “If you don’t start now, you’re going to be too far behind the curve.” However, I found it tough to know where to start. The AI landscape seems to change daily and the real-world AI-ification of everything made me apathetic to the whole thing.

Finally, one day, I was motivated enough to start. I downloaded Claude, paid for the $20 subscription, sat down at my laptop, and…nothing. Online, I found advice like “automate a daily task” or “apply it to a workflow in your job”, but I didn’t have a use case I was excited about or interested in. All my ideas felt too grand for a beginner (like building my own budgeting app from scratch) or too trivial (like planning a trip).

On a whim, I decided to test something random, just to see what I was even in for. I didn’t have a vision or plan, I just needed to get the ball rolling. I happened to have an esports stream on in the background and decided that was as good a starting point as any. I asked Claude to make me a tracker of changes to the game, such as gameplay updates and pro play results. (Note: this isn’t an esports blog, it just happened to be the nearest thing at hand!)

The first output from Claude surprised me - I had a vague prompt, but I got a fairly sophisticated output. That was enough to spark a few more questions. Could I actually make this useful? How far could I take this? How could I make this accurate? Just like that, I had something I could build on.

In the different Notebooks below, I go into more detail about my process, approach, and experimentation, for those who are interested in more details of my exploration.

However, if you are in a similar place, staring at the monolith that is AI, struggling with the inertia to get started, then I hope you can take inspiration from this. Pick something. Literally anything, as long as you’re interested in it. Ask your AI to do something simple, like a tracker or report. That should be enough to open up all sorts of follow-up questions and interactions, and suddenly you’ll have started the snowball to get you AI-ing.

Project Overview

My first foray into AI. Claude makes me a hub for all things Rainbow Six Siege (a competitive video game with a strong esports scene). I iterate on the resulting output, experimenting with how I can get Claude to make the hub better. I probably spend too much effort trying to perfect something built on the fly, but it gives me a taste of Claude’s capabilities and makes me think about strategies I can use to get better results more efficiently.