Not sure what others have been reading so far, but this article already smells so human, I couldn't imagine anyone mistaken it for LLM output (especially since it is about LLMs and how they fail to work for some of us).
Funnily enough I also consulted three models across different vendors, and all came to the same conclusion (it's very human and very unlikely LLM produced). And yes, I let them all cross reference with older posts which are all pre-LLM era. Takes less than 5 minutes to do so.
I'm not a native English speaker, so I don't really have much of an opinion about LLM writing. It probably doesn't stand out to me that much. Your reply made me curious though, so I asked a bunch of "is this AI writing" websites about this article, and they landed on about 7-8 % AI writing, so pretty unlikely.
On the article itself, I found the title quite relatable. When coding agents became a thing, I kept arguing with LLMs before realizing how pointless it is. I really liked the turn the article took, relating the communication breakdown to neurodivergence, and the inherent bias that current models bring.
According to this communication model, there's four aspects to communication between humans. Apart from the factual and the appeal layers, it highlights other aspects, which it calls self-revealing and relationship layers. It is somewhat surprising, and definitely remarkable, that these also matter when communicating with an LLM.
The "Idiosyncratic rater bias" is a bit confusing to me, since it starts already with Bob being an EM, and to counter it is suggested to write from the manager's perspective. So what are managers supposed to do when they are falling for this bias like Bob did? Was either the example or the solution different initially? Maybe it's just the phrasing. (Hey, EM-Bob, try to write like a manager … — Erm, I am one. Duh.)
That's interesting. I'll have Mac access only next month.
The setting `max-width: 100%` actually breaks the design completely, chopping it down to a half.
I have to try this solution[1] to see if that helps to avoid the scrolling.
I think I should just remove the poster, it was initially replacing the GIF, which was megabytes in size. Given that the sizes of the WebP and WebM animations are pretty close I can live without it; all major browsers support both formats anyway.
I cannot help with the caching fun here, quick tested in both Chrome and Firefox. There's probably something strange in the video tag, either by design or by implementation.
Also zola, hugo, and other pretty light weight generators are quite okay. I'm using zola which runs nicely on Win/Mac/Linux. All you need to get started is some theme/template and your blog posts in Markdown. Then enrich it with whatever floats your boat.
I decided to have no (client side) tracking at all, for both usually it requires JS and cookies. And honestly I don't like to have both on my site if not strictly necessary (as a European citizen I'm also quite aware of all the shenanigans one has to do wrt privacy and consent).
Not sure what others have been reading so far, but this article already smells so human, I couldn't imagine anyone mistaken it for LLM output (especially since it is about LLMs and how they fail to work for some of us).
Funnily enough I also consulted three models across different vendors, and all came to the same conclusion (it's very human and very unlikely LLM produced). And yes, I let them all cross reference with older posts which are all pre-LLM era. Takes less than 5 minutes to do so.
reply