Programming insights to Storytelling, it's all here.
When the Black Lives Matter movement was in full swing, someone made a tweet along those lines: "Hey, let's help black businesses get online" (paraphrased). This just resonated with me. Well, not just with me, but more than 100,000 people decided to like and retweet it.
When someone shares a genuine frustration with AI like a hallucination, a bias, or a workflow meltdown, like clockwork the replies are always the same: "You’re just not using it right." "Skill issue." "Prompt better."
When you're a small company, the temptation to mimic the processes of large, successful corporations can be strong. The illusion is that these workflows are inherently superior, a secret ingredient to their success. However, the reality is far more nuanced: large companies often adopt complex workflows not because they're the "best," but because they have no other choice.
A couple years ago, a friend told me he came about $3,000 and was asking which crypto he should invest in. We were at the tail end of the Web3 hype, and from the beginning, my answer has always been the same: Don't!
I've tried vibe coding. It did not work for me. I think the problem is that I'm an actual developer, and I feel dread running code I don't understand. As Andrej Karpathy, who coined the term, described it, vibe coding is about fully giving in to the vibes, embracing exponentials, and forgetting that the code even exists. He talks about asking LLMs for "the dumbest things" because he's "too lazy to find it," accepting all changes without reading diffs, and simply copying error messages in until they're fixed.
I was trying to experiment making an animation using the HTML Canvas and JavaScript. After drawing a small image and creating a game loop, I noticed my laptop fans whirring to life. The more I worked on my little animation, the louder my fans became. In fact, I didn't even finish my experiment, and I got a notification for low battery. What was it about my tiny 512x342 animation that made my powerful modern computer choke?
The 2010s was the Wild West of the mobile world. "Mobile-first" was the buzzword, much like "AI-first" is today. Every company, from the biggest social media giants to your local pizza parlor, seemed to be pestering you to download their app. There was a genuine hype train, and everyone was on board. The apps, frankly, were always mediocre, and a far cry from the full functionality of their website counterparts. But the message was clear. If you weren't on mobile, you were falling behind.
When companies publicly announce on LinkedIn that they're embracing AI, what are they actually doing internally? Are they replacing developers with agentic AI that swarms through code, debugging errors and deploying fixes? Are they parsing business requirements, accessing repos, building features, and deploying them before the 9am standup? For outside observers, it's hard to tell. But as a developer, I can assure you there's often a disconnect between what a CTO says and how the day-to-day workflow is actually affected.
Every time I join a new company, I wonder how their code ends up becoming so complex. You’ll see an application that is supposed to do a simple task turn into a monstrosity that everyone is afraid to update. Writing simple software is often the goal, but somehow we go astray. Tl;dr: Dedicate significant time to refinement *without* adding new features.
So, I call myself an AI skeptic. But honestly? It's more accurate to say I’m a technology skeptic in general. Before you picture me churning butter by candlelight, boycotting smartphones, or mailing handwritten letters, let me clarify: skepticism isn’t rejection. It’s a pause, a thoughtful question: Does this actually solve a real problem, or is it just the latest shiny new hype cycle?