In this video I break down what I think the future of software engineering is going to look like as GPT-style tools keep getting better. We’ve already seen productivity boosters like Copilot helping with autocomplete, plus tools that help with testing, spotting vulnerabilities, debugging, and generating documentation. But the big shift is that the new “app builder” style demos are pushing us past basic assistance and into something that can take you from zero to a working product way faster than a traditional dev cycle. The way I describe what’s coming is “prompt-driven development.” Instead of spending most of our time writing code, we’ll spend more time writing prompts, descriptions, and requirements—and then judging whether the output actually makes sense, is written in a good style, and is safe. That changes the job: software engineering becomes more like consulting, system design, and architecture, where you need domain knowledge to make decisions (like database choices) instead of blindly trusting whatever the model suggests. My main takeaway: if you’re a software engineer, your biggest asset can’t just be “I can code.” Coding should be the default. You need depth—design thinking, patterns, and decision-making—because you’ll never beat AI on raw speed, and the people who survive this shift will be the ones who can hold the quality bar.