eschew
2 days ago
Seems ironic that a blog post describing GHC's AI policy, from a contributor to GHC, which expresses agreement with GHC's AI policy, itself directly violates GHC's AI policy (points P1 and P4)!
wffurr
2 days ago
I don't understand how this violates point 1, and I assume by violating point 4 you mean this essay is LLM assisted but the unnamed authors (Serokell PR?) don't disclose that. I'm not as certain that this post is AI written - it might just be "written by PR committee voice".
eschew
2 days ago
I admit I don't have ground truth here, but the whole thing reads to me as very clearly AI generated. For what it's worth, Pangram agrees. In my view, this style of blog post does not "prioritise people and community" as "Generated text should not be relayed as a substitute for the contributor’s own judgement".
When the content is AI-generated, it's very hard to suss out what, if any, human judgement was involved.
glhpp
2 days ago
The website itself looks Claude-heavy. The link to the PR team actually links to the developers, all of whom are not native English speakers.
Considering that, the article is just too bland and polished and also goes into too much historical detail, which LLMs like to do.
internet_points
a day ago
> Review is the scarce resource
> Maintainer time is the main constraint.
I mean, yes [the] use articles is difficult for speakers of [the] Slavic languages, but then you normally see it like "we want to build code base", whereas this reeks of LLM voice.