Imagine spending weeks agonizing over every single paragraph, polishing every example, and pouring your heart into a course, only to have Google dismiss it as "AI-generated." 🤬🤬🤬🤬🤬
I was asking Google AI Mode to review my course. It gave me surprisingly positive feedback:
Then it listed a lot of reasons why it's good.
I became really happy. I asked it to give feedback on a certain chapter:
This time, my pride was over the top! 🙂 It gave me even more specific reasons why it was "exceptional"! I was really glad that it caught my effort.
Then, I asked about the chapter titled "Hardware Interrupts". That's one chapter I had to check to see if it was good enough. This is when things got mad:
I AI-generated the course!!!??? I NEVER AI-generated the course! I even had a whole page on the course why-not-ai, describing how exactly it was made, and explicitly why it's NOT AI-generated. I even have every single project in the course in this repository, every single one manually tested on my own ESP32:
https://github.com/effessdev/esp-idf-examples
I was furious. I asked it when I said that:
IT'S ACCUSING ME OF AI-GENERATING MY COURSE BASED ON A SINGLE COMMENT I WROTE??? WHAT? 🤯
I had discarded ENTIRE CHAPTERS and rewrote them from SCRATCH with completely new examples because I thought the final project didn't show a scenario where the thing that's taught is the best approach. You can check the commits if you don't believe me:
https://github.com/effessdev/esp-idf
Instead of just dumping the finished code and explaining it later, I walk the reader through building it piece by piece. It makes for a vastly better tutorial, but it is an absolute NIGHTMARE to write & test.
I completely lost my temper:
Now, we're back with the positive feedback. Key points:
- Sounds like a real human who had actually wrestled with the errors
- Feels like peer-to-peer teaching, not corporate documentation
- Practical rather than theory-heavy
- etc.
So now it accepts that it's good?
Then I told it that every single project is available in my esp-idf-examples repository:
I didn't stop until I made it say sorry:
I know it's kind of silly (and dumb). To argue with AI and finally correct it. 😅 But I always do that, do you? 🙂
I deleted that comment anyway. Who knows whether it will bring that up in the future?