Hello, I’m Crafty Olive.
I have been running an HVAC blog for more than ten years.
A blog isn’t as flashy as YouTube or Instagram, and it rarely makes much money. Even so, I have always liked what makes a blog different: it is plain, unembellished and free of sensationalism.
Above all, blogging fit well alongside my day job. Making videos takes far more time and effort than the content I want to share really calls for, while writing lets me focus on what I actually want to say.
In the past few years, however, advances in AI have brought major changes even for knowledge-sharing bloggers like me.
LLMs such as GPT, Claude and Gemini now explain things more accurately and more clearly than my own posts do. On top of that, high-quality articles and videos made with these tools are rapidly flooding blogs and every other platform.
One day, while writing a post as usual, I decided to have GPT review it. The moment I saw the review, I realized it would be hard to keep running my blog the way I always had.
GPT pinpointed the strengths and weaknesses of my writing with great clarity and made concrete suggestions for improving it. The feedback was fairly objective, and I felt that rejecting it would have been little more than stubbornness.
From then on, one question stayed with me:
Does writing on my blog the way I used to still do the world any good?
In the end, I concluded that I would either have to close the blog or, at the very least, run it in a completely different direction. For a while after that I all but abandoned it, and many of the readers who used to visit regularly drifted away. Perhaps that was only natural.
Recently, though, I think I have found a way to connect with you again as a blogger.
That is to show how someone with domain knowledge in engineering, specifically refrigeration and HVAC, actually puts AI to work. And to be honest, I think I’m one of the people in this field using AI most actively.
Expert communities tend to be closed. That insularity can help refine expertise. But we are now facing an era that expertise alone is not enough to meet.
What matters, I believe, is not to oppose AI but to understand and use it. Ultimately, those of us at the front line of the AI era need to learn how to work with this new tool.
That is the direction I will take the blog from now on. Within the single field of refrigeration and HVAC, I will show from my own perspective how I actually use AI, how far it can help, and where human judgment remains essential.
The first project to demonstrate this is the shrimp-keeping game I mentioned in my last post. Rather than simply making the graphics and UI look convincing, I aim to build it with as much engineering validity as possible, including a heat balance model and water quality modeling.