As a writer, I’ve always been afraid of generative AI. I have data proving that thousands of writers are creating AI-generated novels.
I get it - GenAI, including ChatGPT, can produce incoherent content due to its hallucinatory nature. When media factories flood the market with AI content (often edited by mediocre writers), audiences will gradually forget what human-generated content tastes like. When this shift continues for a generation, humanity may forget that writing is a human ability. I expressed my concerns before ChatGPT arrived.
When it comes to programming, however, I see it differently: GenAI and its co-pilot library of tools will likely fuel the growth of independent programmers and startups.
Does the existing system suffer from huge technical debt, unclear documentation, and vague roadmaps? Human programmers will be indispensable. Why? Experienced people usually know how to ask the right questions and whether those questions have viable answers.
It turns out that the same criteria apply to most of what ChatGPT does for independent programmers.
Here are seven things I've tried and tested that can give programmers superpower-like abilities.
1. Implement use cases + workflow (core)
When learning front-end development, I found user authentication to be the most annoying area.
I visited Stack Overflow, but the answers often lack key steps. When I realized this, I had to search through tons of poorly written technical blog posts (or long YouTube tutorials) that first suggested I install 1-2-3 dependencies.
When I started implementing I would find that they were either outdated or my computer had some serious compatibility issues with them.
With ChatGPT, back and forth is a thing of the past.
If I were to try it today, I would just use the following series of tips:
- With Angular, describe