Working with AI Agents isn’t just about choosing the right tool—it’s about learning how to guide it. This post explores prompt refinement, system instructions, and iterative techniques that can help developers get better results from AI.

Working with AI agents can be smooth at first—especially when starting from scratch. But things get trickier when you dive into real-world codebases.

I ran a simple experiment to see how well AI models like ChatGPT and Gemini can summarize long YouTube videos. Using subtitle files and a consistent prompt, I compared their ability to extract key topics and timestamps. The results were... mixed.

I continue my vibe coding journey by refining the infrastructure of a simple React app. I cover setting up Yarn, adding CSS Modules, improving date formatting, introducing i18n, and writing Copilot instructions. Most changes were guided by AI, with some manual adjustments along the way.

A hands-on look at building a simple React form using GPT-4.1 Agent Mode with minimal manual changes. I share how the Agent handled common development tasks, small UX improvements, and where close supervision was still needed.

MCP is a powerful piece of tech. The idea of an API for AI unlocks a whole new world of possibilities. While it still feels a bit distant from our everyday workflows, we’re actively exploring how to make it fit—and whether a simpler alternative might work just as well.

OpenAI recently released a series of videos showcasing different use cases for ChatGPT. I wanted to summarize them here, because otherwise I'd have to dig through each video again to find those useful examples—and that would be annoying.

How hard would it be to take a topic and generate a one-page comic out of it?

Let’s talk about simple financial analysis with LLMs. In this post, I’m not focusing on how to manage expenses. Instead, let’s see if we can gain some insights into investment analysis.

We're back with part two of our LLM animation adventure, this time giving Gemini 2.5 a whirl. We'll walk you through how we tweaked prompts to get that cool particle sand effect, and share some thoughts on how Gemini handles code changes step-by-step.

There's been a lot of excitement lately about the coding capabilities of modern Large Language Models (LLMs), so I figured it's high time I checked them out for myself. Will lead to replacement of engeeneers anytime soon? Maybe, maybe not—but either way, I'm curious to explore what these tools can offer right now.