| Pages in topic: < [1 2 3 4 5 6 7 8 9] > | Consider the consequences of AI before you start a career in translation Thread poster: Gerard de Noord
|
|---|
LEXpert 
United States
Local time: 02:13
Member (2008)
Croatian to English
+ ...
| Re. prevalence of English | Mar 31, 2022 |
The threat, for us translators, is not only the advance of technology, but also the increasing prevalence of the English language.
Secondary school students in one of my source countries are currently cramming for the secondary school leaving exam - consisting of 1. Math 2. The local language and 3. First foreign language, usually English.
Last year, the average mark for the English portion of the exam was substantially higher than the average mark for the portion testing knowledge of the local language and literature.
I'm sure there's a lesson in that somewhere.
[Edited at 2022-04-01 12:12 GMT]
| | | | Jeff Whittaker
United States
Local time: 03:13
Spanish to English
+ ...
| Jeff Whittaker
United States
Local time: 03:13
Spanish to English
+ ...
| A lot of the work I get today is... | Mar 31, 2022 |
... non-machine readable PDF documents.
| | | | Sadek_A
Local time: 11:13
English to Arabic
+ ...
| Speaking of AI & technology, that internet search is something else | Apr 1, 2022 |
To think that one types "shades floater" in the search box -hoping for a hot new profile pic look-, just to find the results page giving only "shady gloater".
I guess the search function needs to suit up with a new rhetoric, the old one is getting very stale.
| | |
|
|
|
I agree with other posters who said that MT is only replacing the "grunt work".
Consider this example sentence, which is part of a text to be translated:
I've been thinking about your problem and I think I've found the prefect solution.
At the most basic level, let's call it level one, MT might just translate one word at a time to produce a literal translation, ignoring the fact that there's a noun where you would expect to find an adjective.
... See more I agree with other posters who said that MT is only replacing the "grunt work".
Consider this example sentence, which is part of a text to be translated:
I've been thinking about your problem and I think I've found the prefect solution.
At the most basic level, let's call it level one, MT might just translate one word at a time to produce a literal translation, ignoring the fact that there's a noun where you would expect to find an adjective.
Any competent human translator would notice the typo and mentally convert "prefect" to "perfect" before translating the sentence. The good MTs today could probably do the same. Let's call that level two.
Level three gets more interesting. Suppose that this sentence is taken from a letter that forms a chapter of a novel, and the typo is deliberate, placed there by the author to convey the idea that the character writing that sentence has trouble spelling correctly, an idea that is supported by other, similar spelling errors or inversions sprinkled through that chapter.
Nobody could determine that just from the single sentence, but an accomplished translator would pick it up while reading the entire chapter. Then there would be the problem of "translating" it into a similar error in the target language, following the author's intention, and doing the same, consistently, for the other deliberate errors. This could be crucial, if the character's bad spelling will later have significance for the plot.
Machine translation has been able to do level one for a while now, and in the last few years has been able to deal with level two, at least some of the time.
I would never say "never", but I don't see an MT being able to handle level three any time soon. Many human translators might miss it.
This example is obviously contrived, but I'm sure my erudite colleagues will be able to find equivalent cases for other kinds of source text. ▲ Collapse
| | | | | English is another threat/opportunity | Apr 1, 2022 |
Metin Demitel wrote:
The threat, for us translators, is not only the advance of technology, but also the increasing prevalence of the English language.
Yes, that will be an problem for some - and an opportunity for others.
Similarly, I have lost translation work where end-clients have adopted English as their working language, so translation has been replaced with correcting and polishing their English. The work involved ranges from very little (Step 3) to loads (full Step 2), depending on quality and importance.
I much prefer translating, so I charge more for this review work, making it more of an opportunity than a threat.
(It has also given me a lot of exposure to highly educated foreigners’ English, which can be astonishingly good, but never quite right, throwing more light on the native/non-native issue. But that is a whole different kettle of worms.)
| | | | Sadek_A
Local time: 11:13
English to Arabic
+ ...
Philip Lees wrote:
I've been thinking about your problem and I think I've found the prefect solution.
Beautiful.
Reminded me of a story that included:
"close your eyes and think....think tanks are all torn apart....the world would become a better place, son; right?"
Think that tanks are? Or, That think-tanks are?
The actual meaning is yet to be reached!
| | | | | We'll all become mostly post-editors... | Apr 1, 2022 |
I don't see my income diminishing if I charge an appropriate rate. That said, I'm old, so I probably won't be around to see it (meaning that I will retire, not die...  )
| | |
|
|
|
| In a distant future | Apr 1, 2022 |
Metin Demirel wrote:
The threat, for us translators, is not only the advance of technology, but also the increasing prevalence of the English language.
In my opinion, this is the real threat more dangerous than AI, but well at this time I (and most of us in this thread) probably won't be around.
[Edited at 2022-04-01 12:33 GMT]
| | | | Metin Demirel
Türkiye
Local time: 10:13
Member (2018)
Italian to Turkish
+ ...
| Actually, it's there... | Apr 1, 2022 |
Angie Garbarino wrote:
In my opinion, this is the real threat more dangerous than AI, but well at this time I (and most of us in this thread) probably won't be around.
[Edited at 2022-04-01 12:33 GMT]
...and we're around. Translators of non-English pairs are losing their market share as the companies/agencies are shifting towards the English pairs, rather than direct translations from the source languages.
| | | | | Not in my experience | Apr 1, 2022 |
Metin Demirel wrote: Angie Garbarino wrote:
In my opinion, this is the real threat more dangerous than AI, but well at this time I (and most of us in this thread) probably won't be around.
[Edited at 2022-04-01 12:33 GMT] ...and we're around. Translators of non-English pairs are losing their market share as the companies/agencies are shifting towards the English pairs, rather than direct translations from the source languages.
But perhaps because I mainly translate legal matters, for example (at the moment) a Dutch or French company, cannot send a case written in English to an Italian court. For sure...
| | | | Tony Keily
Local time: 09:13
Italian to English
+ ...
| I can hear a Blake ghost in your machine (translation) | Apr 1, 2022 |
Tom in London wrote:
If you could see the badly written academic paper I'm currently translating from Italian to English, you would know there is no AI on earth that would be able to make sense of it. I ran it through MyMemory, Google Translation, and Deepl, and they were all defeated by it.
[Edited at 2022-03-31 16:03 GMT]
It's weird, given his 'dark Satanic mills' and anti-industrial bent, but DeepL's version actually reminds me of Blake:
As for the greenery, which is now wild and weedy,
a specific intervention is proposed, case by case,
but corresponding to the Citadel's new design!
So maybe DeepL is threatening the livelihood of poets?
To be fair to the Luddites, they were protesting against poverty and mass unemployment of artisans caused by industrial capitalism rather against than machines per se. In fact there is an argument that machine breaking was primarily a handily available form of industrial action (by factory employees as much as out-of work artisans) rather than a form of technophobia.
| | |
|
|
|
Gerard de Noord
France
Local time: 09:13
Member (2003)
English to Dutch
+ ...
TOPIC STARTER | And, of course, MT won’t stay free forever | Apr 1, 2022 |
When CAT was becoming popular, I supported Yves Champollion in his quest to break the monopoly of Trados. I’ve used CAT ever since. At the time I had no idea of the ramifications of translation memories shared over the internet. I myself helped AI to get smarter.
Once AI will have picked our brains sufficiently – that’s a matter of months, not of years, those computers can do many millions of reiterations while we are sleeping – end users will have the choice to pay the MT p... See more When CAT was becoming popular, I supported Yves Champollion in his quest to break the monopoly of Trados. I’ve used CAT ever since. At the time I had no idea of the ramifications of translation memories shared over the internet. I myself helped AI to get smarter.
Once AI will have picked our brains sufficiently – that’s a matter of months, not of years, those computers can do many millions of reiterations while we are sleeping – end users will have the choice to pay the MT providers for their “premium services”, a free translation or paying a human being.
I hope computer power is now directed to teach AI to translate from Ukrainian to English. The thing with AI is that even that helps the algorithms to become faster and better in any other language.
I've posted this in Getting established. Those who are already established, probably will do fine in the comming years.
Cheers,
Gerard ▲ Collapse
| | | | | Intelligence ... really? | Apr 2, 2022 |
I know that almost everyone uses the term 'artificial intelligence' for DeepL and the like. But I don't find much 'intelligence' in it. It is about pattern recognition and statistics.
When we talk about rule-based expert systems to reach a legal verdict or a medical diagnosis, then the term 'artificial intelligence' is appropriate.
So, let's just call it what it is: Neural machine translation (NMT).
We translators add the real intelligence.
[... See more I know that almost everyone uses the term 'artificial intelligence' for DeepL and the like. But I don't find much 'intelligence' in it. It is about pattern recognition and statistics.
When we talk about rule-based expert systems to reach a legal verdict or a medical diagnosis, then the term 'artificial intelligence' is appropriate.
So, let's just call it what it is: Neural machine translation (NMT).
We translators add the real intelligence.
[Edited at 2022-04-02 06:38 GMT] ▲ Collapse
| | | | Sadek_A
Local time: 11:13
English to Arabic
+ ...
| There is a severe shortage of translators! | Apr 2, 2022 |
When saying "translators", the meaning is language-graduates who possess springboard talent, complementary skills, undying commitment, and extensive experience.
This shocking revelation on a shortage is heavily covered under a constant, administered claim that demand is lower than supply.
Abundant supply of rotten vegetables on the market doesn't fill your grocery bags, does it?
We're being tricked into the hands of the Rich (the ones generating cash from o... See more When saying "translators", the meaning is language-graduates who possess springboard talent, complementary skills, undying commitment, and extensive experience.
This shocking revelation on a shortage is heavily covered under a constant, administered claim that demand is lower than supply.
Abundant supply of rotten vegetables on the market doesn't fill your grocery bags, does it?
We're being tricked into the hands of the Rich (the ones generating cash from our translations). They tell you there are many others willing and able to do the work, and thereby coerce you to accept low pay.
The Rich do not want to budge, neither will us. And, we should not help MT; and, definitely must not help MT at pitiful rates.
https://www.papercup.com/blog/global-translator-shortage
"Meanwhile, groups that represent professional translators state that this problem is due to the low compensation offered by these companies driving translators away."
“I can tell you literally, this industry will be out of supply over demand for the upcoming two to three years. Nobody to translate, nobody to dub, nobody to mix –– the industry just doesn’t have enough resources to do it.”
And, of course, the diluted version of the COVER-UP:
“There is no shortage, but instead a disconnect between the value of this skilled work and the pay offered, leading to a perceived lack of qualified professionals available for these jobs and subpar subtitles in the world’s most popular titles in film and TV.”
Remind me again, why is the American Translators Association not fixing the problem they just admitted to?
"However, the ATA President, Madalena Sánchez Zampaulo, later told Multilingual that if this lack of fair compensation continues, it’s likely that even more professional translators will specialize in other industries, which will ultimately impact the bottom line of international media companies as audiences reject poor quality translations."
A shy off-the-record guest-room talk, so that the Rich don't come back and say they were misled!
Why exactly was that not included in the organization's official public statement?
The end of translators-favourable talk:
"And even if they were suddenly willing to pay more for these services, it takes years to train a skilled translator, which likely wouldn’t be fast enough to keep pace with the explosion in demand they’re trying to meet."
Now, the article, of course, started out with the tone of taking translators' side, and -similar to what always happens here on proz- ended up taking a stance against those translators.
"But the translation industry is still incredibly reliant on manual human input."
"But that doesn’t mean automation and AI can’t help combat the translator shortage."
"To address the quality concerns, many solutions combine technology and human translation to accelerate workflows while maintaining high quality. You can create AI-generated subtitles, for example, that are then checked and edited by a translator. "
"While this technology can’t be applied to all media genres and formats (yet), it’s capable of scaling translation across a wide range of content types with speedy turnarounds at a low cost. This frees up the limited supply of professional translators to focus on high-priority projects that genuinely need end-to-end manual translation."
NO ONE IS GOING TO HELP US, WE NEED TO HELP OURSELVES. SAY NO TO ANYTHING THAT JEOPARDIZES YOUR PROFESSIONAL STANDING. ▲ Collapse
| | | | | Pages in topic: < [1 2 3 4 5 6 7 8 9] > | There is no moderator assigned specifically to this forum. To report site rules violations or get help, please contact site staff » Consider the consequences of AI before you start a career in translation | Protemos translation business management system | Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!
The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.
More info » |
| | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
Translation news
