Live data from Hacker News

End of Japanese community

support.mozilla.org

361–370 of 820 posts

Re: End of Japanese community

#361
post #11

Earlier quoted context omitted.

Some grievances were vague. It doesn't follow our translation guidelines. What specific guidelines did it not follow? It doesn't respect current localization for Japanese users, so they were lost. What was not localized? Mozilla's response should not be limited to clarifying these grievances. But it could have been all the staff member who responded could do.

These ones? https://github.com/mozilla-japan/translation/wiki/L10N-Guide... Looking through that wiki there seems to be a lot of things that ML would get wrong.

Interestingly, it looks like some of the "bad examples" are precisely the kind of things ML would produce. Others are what non-native speakers would produce (starting with "あなたは" to translate a sentence starting by "you").

I'm sure some translators were using ML before it was integrated, and those guidelines are here in particular to tell them about those problems.

Also, ML is now really good to translate between European languages, but Japanese is very different in its structure so ML from English to Japanese is not as good. I'm sure some people who only know English/French/Spanish/German saw that ML is pretty good, and don't realize that for some other language it just doesn't work.

Re: End of Japanese community

#362

Isn't it fascinating that despite `while true; do claude --yolo` over a weekend being all it takes to port some project across platforms, LLMs completely fall apart when it comes to speaking grammatical and natural Japanese? Free tier Gemini CLI literally writes Android app for me by just endlessly wondering in English. AGI's here. And it struggles with Japanese. How!?

The simple truth is that LLMs are still bullshitting their way through life.

In some fields and languages, this is easier, but as soon as you need the LLM to follow rigorous instructions, it'll fail.

Re: End of Japanese community

#363

Earlier quoted context omitted.

sounds like it's almost an AI response? I can't believe a person responds that way

A few years in corporate managerial role will turn you into a bot.

Not even necessarily managerial. Any large bureaucracy tends to force this kind of speech on the people who constitute it.

Re: End of Japanese community

#364

> Hi Marsf, > I'm sorry for how you and the Japanese community feel about the MT workflow that we just recently introduced. Would you be interested to hop on a call with us to talk about this further? We want to make sure we trully understand what you're struggling with. > My timezone is UTC+7, so it should be easier for us to set up time. Let me know how that sound! What a horribly condescending and patronising resp…

> I'm sorry

good start

> for how you feel

oh.

> hop on a call

...oh.

Re: End of Japanese community

#365
post #18

As a multilingual/multicultural human it’s been pretty weird witnessing what AI translation has been doing to regional languages & cultures on the internet in the last few years. Sure we had machine translation before, but it was still a little off. Now the latest language models get us 99.9% there, so they are judged good enough to deploy at scale. What results is a weird twilight zone where everything is in your la…

I'm a native English speaker fluent in Japanese, recently moved to Japan this year. The one that really gets me lately is YouTube now automatically dubbing over content in Japanese that was originally in English. It's... so uncanny.

It is awful. Its also not so transparent that its being done which feels unethical.

Re: End of Japanese community

#366
post #42

Earlier quoted context omitted.

The ones that are easiest to point to are turns of phrase like “living your best life”, “that feeling when you…”, etc.

I don't think the LLMs are to blame here. Not yet, at least. This is caused by people active in English-speaking communities translating memes literally and spreading them in their native language communities as-is. As the meme spreads, monolingual speakers begin using the same format and eventually they reference it off-line.

Not really. I have observed the same thing in Russian, and no, it's not for expressions that are translated literally.

Having said that, SOTA models got much better at this kinda stuff. They're quite able to write in a way that is indistinguishable from a native speaker, colloquialisms and all, with the right prompting.

But SOTA models are also expensive. Most automated translations are done with something way cheaper and worse.

Re: End of Japanese community

#367
post #317
post #272

Earlier quoted context omitted.

Why is it condescending and patronzing? I read it as a person trying to understand the situation.

I'm not part of Mozilla or any of the communities and I understood the situation by reading the damn post, on the first time. In the follow-up, any words concerning how the person feels, words on how to talk about this further, and wanting to truly understand what he just wrote in plain and simple terms simply reek strongly of "we really won't change anything, we've made our decision, we are disagree with you but we…

To add to this fire. I recently left my 'recently bought by private equity' workplace of 8 years to work somewhere else not-yet-ruined-by-PE (yet..) A major part of my decision to quit was this communication pattern.

The whole organisation was very efficiently structured with two separate layers of managers - those who had actual decision power, and a separate layer whose task was to 'deal with us employees' but no decision power. All communication flowing one way, the same way shit drips (the only resource following 'trickle-down' mechanics). The only time I got into contact with the former level, was after I had put in my resignation; then they suddenly wanted a 1-1 to "see if there was anything they should learn from this" (presumably to sharpen/hone their skills in mistreating the employee level more efficiently in the future).

Re: End of Japanese community

#368

Earlier quoted context omitted.

lmao. This is a "research team and five years" task with current state of LLM.

Yes, but now that's the answer to questions like "how do we deploy AI without pissing off our communities?"

I have your answer: you provide a tool for your translators, you don't unleash a bot that makes changes left and right and creates new pages.

Typically, when a new page is written in English, don't automatically generate a version in all languages. When a translator starts creating the page in their language, provide a button to pre-fill with ML translation if they want to.

And for users, you can display the English version with a message, "this is not translated in your language yet but you can read an ML version if you want".

Re: End of Japanese community

#369
I know that everyone has already given their opinions about what kinds of people are involved and their motivations, but this is really about two fallible humans, one listing grievances and another asking to open a communication channel.

That's it.

Anything else you read into this is going to be fraught with your own coloring based on a hundred words written in text (a notoriously difficult medium to establish emotional communication over).

Regardless of how nice or not-nice the text may sound to the various cultures that have weighed in so far, the right thing to do is talk voice/video and hash out what the problems are, and work together to come up with a solution that will satisfy everyone.

That's what communication is about.

Re: End of Japanese community

#370

Earlier quoted context omitted.

If this were some small bug report, even from someone pretty ticked off, I'd agree. But this is a response to someone who has just announced that they are quitting an organization they had spent 20 years of their life volunteering for, because of the disrespect they have felt from the org. This is not a "hop on a call" moment. This is a "please accept to meet so I can apologize in person, and see if we can repair thi…

I don't know the original author but as a non-native speaker I guess I miss the cultural connotation surrounding "hop on a call". Doesn't seem weird to me, seems apt to ask to escalate the method of communication when a serious issue like this arises. Would you consider it more respectful to deal with this issue just by posting in the thread?

It's a community matter posted in a public place, a non-statement with an immediate attempt to direct it to private conversation reads like trying to avoid attention to your mistakes (e.g. hypothetically you don't have to public admit you didn't do anything to check for guidelines to follow). More vibe-y, it all sounds very corporate, like any PR statement in response to criticism ever, or a manager writing to an employee in a big corp, not "humans working together in a community, and one of the humans is clearly pissed off right now".

Also while it's phrased as a question, it doesn't offer any alternative next step. So a better approach would be writing down the initial questions you have and then offer that you'd be open for a call if the OP prefers that. If they don't, they can immediately engage with your questions, and they are open to everybody else in the community. Whereas right now if they say "no, I don't want to call you" that's all you've given them.

(To be clear I can easily believe the writer of the response is not intending any of that and means well, but that's how it comes across)

Post reply on HN