I'm writing this in English.

[hypolite@friendica.mrpetovan.com]

@lain_7 It would be the case if these tools weren't aggressively integrated in everything we use. If there's a murky water tap in your city, then it's up to the users to appropriately filter the output for safe consumption. Now if the murky water tap is forcibly installed in every household because the city received bribes for it, then the responsibility shifts from individual users to the murky water tap provider. Even if users can get somewhat safe water from adequate filtration.

[hongminhee@hollo.social]

@julian@activitypub.space I actually just addressed something close to this in a reply up the thread—might be worth a read!

[hongminhee@hollo.social]

@phnt@fluffytail.org Multilingual posting is something I've actually been thinking about and experimenting with already. I'm building a fediverse platform for software developers called Hackers' Pub, where I've been prototyping exactly this kind of feature. My other project, @hollo@hollo.social, is where I post from right now, and I'd like to bring something similar there eventually—the main obstacle is that most fediverse platforms don't properly render multilingual posts even when they receive them—the protocol already supports it via contentMap, but implementations like Mastodon tend to pick one language arbitrarily rather than presenting all of them.

Outside the fediverse, my blog already works this way: each post is written in multiple languages (Korean, Japanese, and English), not as translations filed separately but as the same post published in all three. It's a pattern I find natural given how I actually think and write across languages.

On Google Translate: I stopped using it for Korean–English a long time ago, for the same reasons you'd expect. These days I mostly use Kagi Translate, which is LLM-based and noticeably better for distant language pairs. The gap between that and older MT is stark enough that I think “machine translation” has started to mean two quite different things depending on which generation of tools you're talking about.

[ddelemeny@mastodon.xyz]

@hongminhee @silverpill Thank you for replying with care, your POV is really interesting.
Have you read the Reg's article about semantic ablation that was shared around some time ago ?

https://www.theregister.com/2026/02/16/semantic_ablation_ai_writing/

This may be part of why other commenters allegedly found a negative difference in your writing using MT, and a concern I have about building fluent writing skills. MT (and autocomplete) collapse possibilities into an average, that's probably correct enough but also probably low entropy.

[julian@activitypub.space]

@hongminhee@hollo.social thanks, good answer :smile:

I would wonder then that maybe you might end up sounding like an LLM, then. Best interject some of your own style later on :wink:

[hongminhee@hollo.social]

@ddelemeny@mastodon.xyz That's a useful framing, and the article is worth reading. The concern about entropy collapse is real—I've seen it happen when native speakers run their own writing through a model and get something smoother but somehow emptier back.

My situation is a bit different, though. The high-entropy original is in Korean. The LLM's job is to carry that across, not to sand it down. Whether it succeeds is a fair question, but the direction of the process matters. I'm not polishing a draft into blandness; I'm trying to get something that exists in one language to exist in another without losing its shape.

Anyway, this has been a genuinely interesting exchange. Thank you for the link.

[hongminhee@hollo.social]

@julian@activitypub.space Yeah, that's why I'm still writing short words myself. My accent won't go anywhere!

[clv1@mastodon.social]

@ZDL @fastfinge @hongminhee Hi, for an example of a contemporary project that's technically better than Esperanto, check Globasa: https://www.globasa.net/eng

[zdl@mstdn.social]

@clv1 @fastfinge @hongminhee I think dice with letters on the sides rolled randomly would be better than Esperanto! 🤣 But yes, this project looks pretty sane from a glance-over.

The choice of agglutinative is interesting. I would personally favour particulate, but agglutinative is a very good second choice.

[caohuak@moon.lonewolf.zone]

@silverpill @hongminhee DeepL or some LLM translation assisants often use hyphens to replace commas or semicolons, so I usually use this feature to confirm if a text is machine-translate.