Technology Has Already Outpaced Us. Where Do I Begin to Catch Up?
← NotesAI is the most consequential shift in how work gets done in our lifetime, and I think about it every day.
For me, it begs the question: "How do I adapt to a reality where most technical problems are a few prompts away from being solved or automated?" It's a scary thought!
What retains value, I think, is good judgment. Being able to define the right problem, design the system that solves it, and know when the output is wrong. That's the layer that doesn't compress. I feel this deeply because I've experienced firsthand how colleagues, my bosses, and leaders in general could sometimes point resources toward the wrong problem.
When you apply AI tools to a clear process, you get speed.
Apply it to lazy prompts and a broken workflow and you get AI slop. Solutions nobody asked for, delivered fast!
The differentiator won't be which LLM you're using. Major models now like ChatGPT and Claude are essentially reaching a point of convergence that makes them generally good at most things with some edge case uses for certain tasks (Claude for general use, ChatGPT for debugging and thoroughness, Gemini for creativity and research). The real value will come from the questions and planning you do before you start building.
The way that I've tried to stay informed and relevant in this age of tech has mostly been doom scrolling social media to see the latest hype and confirm it myself by actually using (and paying for) the latest models. I also read the published documentation from Google, Anthropic, and OpenAI to understand how the models work and what changed from their last release. I believe that's terribly unsustainable, but it's a medium I'm used to.
In the long term, I've decided that the best way to stay "caught up" is to gain deeper expertise in one model and set of tools, and master it. For me, that model has been Anthropic's Claude models and using Claude Code.
While each model has their own way of "thinking" and producing output, it's safe to say that reading about the latest releases isn't going to keep exponentially increasing your output. Instead, I'm focusing my time on building more often, getting better at creating documentation, and sharpening my planning skills.