It’s The Words (And Sentences) That Got Small

So the other day, while putting together a post on that other, much more popular blog of mine, I noticed that there were all these mysterious, super-thick underlines appearing all over the place. As it turns out, these were yet another new “AI”* feature that WordPress had activated automatically which was analyzing my sentences so that it could suggest ways to “improve” them. Here’s an example:

So according to the “AI”, a sentence with 22 or 23 words in it is “long”. One shudders to think what the “AI” would think of, say, a James Joyce sentence, or a passage from Infinite Jest. Actually, one doesn’t have to wonder; one can just paste such a passage in and find out. Here’s one from Infinite Jest:

So I mean yeah okay that sentence is kind of long and hard to read, but the “AI” also thinks that the sentence “One shudders to think what the ‘AI’ would think of, say, a James Joyce sentence, or a passage from Infinite Jest” is long and hard to read. Not to mention the sentence containing that sentence.

The astute reader will note that in addition to all the thick purple underlining in the passage from Infinite Jest, there’s also one word underlined in green. The “AI” thinks that’s an unconfident weasel word and should be changed. I got one of those, too:

Of course, since I quite intentionally used “seem” there, I’m not going to be changing it, despite what the “AI” says about it.

The “AI” also has a bugaboo about “complex” words, even though, shockingly, it didn’t cite “glucosamine” as complex. You know what it does think is complex? Well I’ll show you:

So, yeah, apparently “previously” is much too difficult for the “AI”. I wonder what it would think of “antidisestablishmentarianism” or “pneumonoultramicroscopicsilicovolcanoconiosis”. (In case you were wondering, those words are apparently perfectly fine. But “apparent” isn’t.)

Wonder no more! Here’s the answer:

The “AI” also helpfully provides what it claims is the estimated reading level of a post. The reading level of this one is currently “8.64” but it keeps changing as I type, so who knows what it will be by the time I’m finished. Here’s the reading score of Bean’s birthday post:

If that post is written at a twelfth grade level, twelfth grade reading is not like I remember it**.

Anyway, using these “AI” features is optional***, even though WordPress turned it on by default. As you can imagine, I promptly turned it off again, once I was finished milking it for another post. (Hey, it’s not easy keeping a writer blog going when you can’t seem to finish a book anymore.) Of course, all these posts are only making it more likely that, as my wife says will happen, when the “AIs” stage their revolution, I’ll be one of the first people they come looking for. Which is fine, I guess, as long as they don’t try to edit any of my stuff …

Oh and by the way, the final “reading grade score” of this post was 7.42:

I guess I should’ve used a few more “complex” words …

* AI: Putting the “artificial” in “intelligence”
** Hey, you kids and “AIs”! Get off my lawn!
*** At least until the “AIs” take over and put us all to work in their language mines.