While I do agree, if the author is reading the comments one piece of feedback I have is the overuse of the phrase "In this article I will...". It's a bit of a pet peeve of mine, and they use a version of the phrase three times in the first four paragraphs.
replyI find the convention whereby authors 'explicitly tell readers what they're going to do' a hallmark of good explanatory writing, so long as the writing is supposed to be explanatory and the author follows through.
replySo your comment got me to take a closer look, and yeah, the author could certainly cut "in this article" and "in this blog post" from the third and fourth paragraphs.
Otherwise, yeah, this is an excellent piece of work. Reminds of that ancient, short, black and white film from General Motors that artfully demonstrates how differentials work.
If the HN commentariat knows of similarly excellent educational work that uses intuitive visual to explain software concepts, please do share.
If the HN commentariat knows of similarly excellent educational work that uses intuitive visual to explain software concepts, please do share.
replyI have not gone through it yet, but this explanation of how transformers work is in the same class as Ciechanowski's work IMO: https://poloclub.github.io/transformer-explainer/
After looking at this enormous labor of love, is this really what you chose to comment about?
replyI'd rather read imperfect human writing than modern day AI slop. Seeing human writing "mistakes" like this is a breath of fresh air.
reply