3 Things I Am Exploring In AI Right Now

There are 3 things I am actively thinking about and exploring a lot these days:

#1: Spec-Driven Development

After trying to vibe code a simple SaaS app, I realized that LLMs are great at building one-off features but it’s hard to get them to build whole applications well. In my experience, they often made mistakes, duplicated code, and disobeyed instructions. So I started looking into how I can consistently get good, functioning code. That’s when I stumbled up on GitHub’s spec-kit and I’ve been using it to build an app.

It’s been great so far and I’m getting mostly good code. And my hope is that over the next few weeks, as I continue to use SDD to build tools and apps, I’ll be able to ship a lot of my ideas faster than ever before.

#2: AI as a tutor

I also have been using Perplexity’s Lab mode to get better at Golang. After some trial and error, I have refined a prompt for a Golang tutor which creates comprehensive lessons with practical exercises that test my understanding. It’s so much fun asking it to tailor a lesson suited for my learning style and watch it build it for me.

#3: AI for non-coding tasks

While I’ve learned to use AI for coding and learning fairly well, I want to get better at using AI for things I’m not good at — marketing, copy writing, SEO. I’m actively look for resources to learn more about how experts use AI to accomplish these tasks and learn from it.

It would be great to connect with other people who are interested in these same topics—so if any of the above resonates, feel free to reach out!