We should get to the point fairly quickly where “written by AI” is far less relevant than the accuracy and timeliness of what was written…

Were the points made in the article valid? Mostly, it was well worth a read, with or without AI involvement.

Did AI add the typos as well?

You know "load bearing" as an expression existed before AI, right?

Knowing the author through his work, it doesn't surprise me he's using that expression. And quickly searching on the blog, he's used it before.

It's yet another borrowing of the prestige of technical terms -- propagated by LLMs, but in rampant misuse by poseurs even before AI: inflection point, exponential growth, order of magnitude, evolve, viable, ...

> You know that the usage has gone up something around 1000x since AI right?

Seems to be going up and down in usage, if you look at Google Ngram + Trends: https://books.google.com/ngrams/graph?content=%22load+bearin... https://trends.google.com/trends/explore?date=all&q=%22load%...

Seems Google Trends say ~50% "interest over time" around 2004, which Ngrams at that same time has about half the occurrence in books compared to around ~1926.

As the labs adjust their models to become more human, less repetitive and more of the population starts catching up these patterns, we'll see it become less used again, starting the new loop of the repeating cycle.

I'm calling it: if Bret Devereaux is being accused of posting slop, we have officially achieved peak unhinged claims of “this seems like AI”