Mastodon Feed
brib@bribstodon.xyz ("brib :neofox_floof: :Nonbinary: -> EMF 2026") wrote:
For anyone I follow: I recommend being aware of common signs of AI use, e.g. 404 Media has an AI Poster Wall of Shame and Wikipedia has a set of AI writing tells. I've personally noticed that slop websites tend to like their rounded-corner boxes and often (but not always) monospace fonts. AI code is weird and muddled (but human code can be too), what is much rarer for humans to do is start embedding em/en dashers or arrow characters (→) in the source (this is different from literary em-dashes, a lot of real writers like to use them). Or when a lot of things someone creates have a very similar and very AI-esque style (it's normal for AI patterns to be occasionally used but for them to be used everywhere is a lot more suspect). The biggest tell for me is simply admitting to LLM usage, not necessarily for that particular project but talking a lot about using it in general or finding assisted commits in repos (https://slopscan.ava.pet/ is your friend).
I'm not going to yell at people for boosting these slop projects because it's rude and saps my energy but I do mute/unfollow people and hide boosts for it. It breaks me a little when I see a slop project get popular because I put a lot of work into making my stuff 100% human and seeing other people use an AI to slop stuff out feels as personal as watching someone get away with plagiarism.
Of course I can't tell every AI usage and I can't expect everyone else to, and I'm aware of the issues surrounding false positives (and I wish there were better ways to tell if something is actually human 😭). But please at least do a background check of the stuff you share