통역사 및 번역가

Milan (매우 가능성이 높음)이 말합니다.

Machine translation has existed for a few years and neural networks and AI will take it to the next level. As an example, Google Translate does poor with context and slang. ChatGPT on the other hand, was never intentionally taught to translate, but I found it does far better on understanding modern slang and making sure the sentences it generate actually make sense. With interpretation, the biggest difficulty for humans is short-term memory. AI does not have this issue.

Noskaj (기회 없음)이 말합니다.

Ai won’t be able to translate context in some languages and it will over complicate words a native would never say

D (매우 가능성이 높음)이 말합니다.

an interpreters job requires you to have incredibly good short term memory and a large vocabulary that you can access instantly. A human will always show flaws in these regards, when compared to a machine. This in turn leads to a good probability of miscommunication. In my opinion, human interpreters will only be neccessary in high end communication, and yet still it might be better for them to be correcting the work of a computer, than interpreting themselves. The only things that save this job are government requirements for some administrative proccesses, the need to demonstrate wealth in some cases and the cultural differences, that an interpreter can notice in live conversations.

Johanna Ellsworth이 말합니다.

Translators and texters are already scrambling for work at ridiculously low rates today (in 2024). I have seen several quotes made by German translator colleagues that are being "offered" payment of € 0,1 (approx. 1 USD) per word... I especially see the danger of future bland literature, media articles and movie scripts "created" by AI, which will shape future generations into creatures that won't think creatively, artistically and critically. Only that which is imperfect, i.e. slightly off, is art...

Remigiusz (적당한)이 말합니다.

While computer translators are getting better and better they are still far from perfect. They made simple mistakes, often translating too literally. That's why I believe that there will be place for a human overseer, at least for now.

Ildiko (적당한)이 말합니다.

Machine translation develops amazingly fast and well. Mostly based on the already translated texts. I welcome that when using machine translation, I don't have to type the words, the software does it for me.

However, we humans are still needed to correct the sometimes amazingly stupid solutions the machine offers when the text is completely new and there are no precedents. Anyway, I would not advise my daughter to choose this profession, except when she likes text editing...

Kareem (낮은)이 말합니다.

Context, culture and words that have multiple different meanings will be hard to understand for AI. The hardest languages will be unlikely to get replaced by AI

Daniel Valdes (낮은)이 말합니다.

I have been a medical interpreter for some time now and I really don't see our job being replaced soon. Just in one language there are dozens of different dialects and registers which can include a lot of different words to refer to the same thing. You have to read the context around the conversation to make sense of all of the senseless things the low English proficient client says sometimes.

Also interpreters and translators are two very different jobs with different automation risks, I don't think they should be together in one category.

Carlos Fiuza (매우 가능성이 높음)이 말합니다.

Note that translation (converting a written text from one language to another) and interpretation (converting a live speech from one language to another) are different activities.

Both are likely to be automated in the near future, since there are no constraints in this realm that cannot be overcome by AI, sooner or later.

Nevertheless, automation will probably come much sooner to translation (as a matter of fact, to a great extent it already has) than to interpretation.

Jalves (낮은)이 말합니다.

A really good translation cannot be mimicked by AI simply because AI can't distinguish social cues necessary to humor.

Cheesd Pepperoni (적당한)이 말합니다.

translators already exist, and with the rapid advance of AI that we're seeing even today, i think that translators will be almost completely phased out except for a couple sensitive applications here and there

Thomas (낮은)이 말합니다.

Translators will be replaced by machines only if we accept to adopt a very narrow view of language as a code used to clearly communicate a message. Language is nothing like that. Unfortunately, dumb capitalist companies and boffins are convinced that language is similar to coding

Paula (매우 가능성이 높음)이 말합니다.

I didn't think so in 2019 or even 2021. But now, May 2023, with Chat GPT4 and the moratorium on AI R&D I'm not so sure. There will still be post-editing, I think, but I do believe the profession is under threat.

Vovin (매우 가능성이 높음)이 말합니다.

I give it 5-10 years before machine translation is the new normal, it's already happening and will simply be cheaper and faster (not to mention that some clients don't seem to care that much about nuanced, very polished texts). The current translators will be reduced to proofreaders/post-MT editors, for the most part.

Edith (매우 가능성이 높음)이 말합니다.

20 years from now, only the most creative and sensitive translation work (max. 1%) will still be done by people. I expect that 80% of the work will already be done by ChatGPT and the likes in 2025 (especially when taking e.g. 2015 as a reference point, as nowadays already a lot of translations are either done completely by DeepL or Google Translate or post-edited to make them fit for purpose).

Edith이 말합니다.

Yes, and it will happen quickly in most text types and specialities (except for literature, slogans etc.). In 2010, I still translated the subject line of a Finnish government letter word by word using my dictionary before calling our Finnish distributor to get more information. In 2015, Google Translate was already able to give quite a good idea what such a letter was about. In 2020, Google Translate NT and DeepL still have funny mistakes but usually make even texts in small languages easily understandable. And now Chat GPT had arrived; within a few months already as good as DeepL, and I don't doubt it will be as good as 80% of the human translators within a few years. That's good enough for most companies that need translations, so the human translation market will collapse.

Anna (적당한)이 말합니다.

If we're talking about 20 years perspective I believe almost all of the jobs will become automated by that point

Karume (낮은)이 말합니다.

Language interpretation requires a lot of skills, not just word-to-word interpretation. Unless robots or AI can fully comprehend human culture, reason completely, and provide ideas, this will remain a science-fiction story.

It's like suggesting that adult robots or AI will fully replace human sexuality - something that verges on the impossible.

Emilio (적당한)이 말합니다.

Maybe certified translators will survive longer.

Richard (적당한)이 말합니다.

Every language is able to express the same ideas and concepts more or less, so it’s only a matter of time until AI figures out the patterns. However, this may take a while because communication is very complex.