Love them or loathe them, large language models (LLMs) like ChatGPT and Google Gemini have the potential to make fundamental changes to the way we work—when used the right way.
the real sweet spot is in what I call the “60-percenters.” These are tasks where I already know 60% of what I need to complete the job. In these cases, using ChatGPT can save me hours.
I’ve always "dabbled" in coding, but I never had the flair or skills to progress beyond the basics. That all started to change when I began leveraging LLMs to backfill my knowledge gaps.
Now, many people have tried the "build me an app that does X" approach, often with mixed results.
Why? Several key reasons come to mind:
Lack of Detail: Anything more complex than a simple task requires precision, and if your request is vague, the outcome may fall short.
Knowledge Gaps: Without an underlying understanding of what you're asking for, it can be difficult to evaluate whether it’s working correctly.
Hallucinations: Sometimes, LLMs just make things up—that’s a fact we have to live with.
For me, the real sweet spot is in what I call the “60-percenters.” These are tasks where I already know 60% of what I need to complete the job. In these cases, using ChatGPT can save me hours.
What might have taken me 2-3 hours on my own can now be done in 20 minutes. No exaggeration—it’s completely transformed the way I work!
And it doesn’t stop at coding. Even if you aren’t building apps, LLMs can assist with a wide range of tasks:
Writing complex Regex
Crafting XPATH or CSS selectors
Writing SQL queries
There’s a massive opportunity here, not just to get more efficient but also to learn in the process. That said, there are a few caveats. You should never share personal or sensitive information with these models—like API keys or email addresses. And, as tempting as it is, don't get lazy! To get the real value out of this co-pilot approach, always ask for explanations on concepts you don’t understand.
Will this method replace a seasoned developer or engineer? No, probably not. But does it make me more efficient and help me tackle more complex problems? Absolutely!
So, have you tried using LLMs as a “co-pilot”? Did it transform your work, or did it fall flat?
Bình luận