First and foremost, my initial comment was made after spending 30 seconds reading the article. It was a quick judgement and (unfairly) condescending. My initial comment was based in the assumption you were a fly-by-night blogger trying to cram buzz words unto your post to get clicks.
I was wrong. And I apologize.
I have now read more of the post. It IS a document on implementing GPT-3, and it IS a good foundation for people who are not familiar with it. I applaud you for making ML stuff more accessible to HNers. I recently took a crash course in deploying ML models (like GPT-3), and one question I kept in my head was "How can we make this accessible to EVERYONE".
Please keep documenting and democratizing cool models.