Poll: How do you feel about tech companies using public minority language content to train AI? Thread poster: ProZ.com Staff
|
|---|
This forum topic is for the discussion of the poll question "How do you feel about tech companies using public minority language content to train AI?".
View the poll results »
| | | |
I wonder if translation these days is restricted to AI. Isn’t there anything else to talk about? Or is there a hidden reason behind this persistent repetition? To do away with the Quick Polls? Please don’t…
| | | | Thayenga 
Germany
Local time: 15:00
Member (2009)
English to German
+ ...
AI from here to eternity.
If AI becomes the sole subject of the polls, then I might as well copy my comments and just paste them into the next poll. How boring.
| | | | Hipyan Nopri
Indonesia
Local time: 20:00
English to Indonesian
+ ...
| Native speakers and translators should be consulted | Oct 7 |
I chose the second option.
Indonesia has hundreds of regional languages, many with very little written material online.
I believe native speakers and translators, as the people who know these languages best, should be consulted on how their languages are used and fairly recognized for their contribution.
Without their involvement, there is a real risk that limited or inaccurate content will be treated as authentic usage.
| | |
|
|
|
Ricardo Aries
Indonesia
Local time: 20:00
Member (2020)
English to Indonesian
+ ...
| An Indonesian view: regional languages need consent | Oct 8 |
From an Indonesian perspective this hits close to home. Bahasa Indonesia itself is well resourced, but Javanese, Sundanese, Buginese and hundreds of other regional languages are not, and much of their online content comes from small community sites, local news and religious texts. Using it to train AI without consent or credit takes from communities that already have the least. Done with them, though, the same tools could help keep these languages alive. Consent, attribution and native-speaker r... See more From an Indonesian perspective this hits close to home. Bahasa Indonesia itself is well resourced, but Javanese, Sundanese, Buginese and hundreds of other regional languages are not, and much of their online content comes from small community sites, local news and religious texts. Using it to train AI without consent or credit takes from communities that already have the least. Done with them, though, the same tools could help keep these languages alive. Consent, attribution and native-speaker review should be the minimum. ▲ Collapse
| | | | Daryo
Local time: 14:00
Serbian to English
+ ...
| If it's any consolation to you | Oct 8 |
Hipyan Nopri wrote:
I chose the second option.
Indonesia has hundreds of regional languages, many with very little written material online.
I believe native speakers and translators, as the people who know these languages best, should be consulted on how their languages are used and fairly recognized for their contribution.
Without their involvement, there is a real risk that limited or inaccurate content will be treated as authentic usage.
In principle, one can only agree with you.
But given that the tech bros have form when it comes to not paying any blind of notice to what others have to say ...
If it's any consolation to you even for 'major languages' the nauseously bland and oververbose AI output is hardly 'autentic usage' either.
| | | | Hipyan Nopri
Indonesia
Local time: 20:00
English to Indonesian
+ ...
| Perhaps even higher stakes for smaller languages | Oct 8 |
Daryo wrote:
In principle, one can only agree with you.
But given that the tech bros have form when it comes to not paying any blind of notice to what others have to say ...
If it's any consolation to you even for 'major languages' the nauseously bland and oververbose AI output is hardly 'autentic usage' either.
Thank you, Dario.
I share your skepticism about how much the tech companies will listen.
Still, I think the problem may be even more serious for smaller languages.
For major languages, readers can at least compare AI output with a large body of human writing, but for many Indonesian regional languages, AI-generated text could end up being most of what exists online, and eventually be treated as authentic usage.
That is why I think speakers and translators need to be involved, even if it is an uphill battle.
| | | |
My initial reaction to this question was: Wow! A question about AI!! It's a long time since we had one of those!!!
But then, after reading the contributions from our Indonesian colleagues, I realised this is in fact a rather interesting issue.
Before I read this thread I had no feelings at all "about tech companies using public minority language content to train AI" (once I deciphered the syntax of all that). In order to arrive at some feeling about it, I would need to ... See more My initial reaction to this question was: Wow! A question about AI!! It's a long time since we had one of those!!!
But then, after reading the contributions from our Indonesian colleagues, I realised this is in fact a rather interesting issue.
Before I read this thread I had no feelings at all "about tech companies using public minority language content to train AI" (once I deciphered the syntax of all that). In order to arrive at some feeling about it, I would need to know the answers to two questions: Is anybody actually doing this? If so, who, and how?
Can the people who posed this question provide any enlightenment? (Fingers crossed, but without much hope.) ▲ Collapse
| | |
|
|
|
Thayenga
Germany
Local time: 15:00
Member (2009)
English to German
+ ...
| The answer is simple | Oct 8 |
Philip Lees wrote:
...I would need to know the answers to two questions: Is anybody actually doing this? If so, who, and how?
It's probably AI generated, and by deceiving its creator.
Philip Lees wrote:
Can the people who posed this question provide any enlightenment? (Fingers crossed, but without much hope.)
I didn't know that AI enlightens anyone.
| | | |
Philip Lees wrote:
I would need to know the answers to two questions: Is anybody actually doing this? If so, who, and how?
Three questions. I need to know the answers to three questions: Is anybody actually doing this? If so, who, how, and why?
FOUR questions. There are four questions ...
(With apologies to the Spanish Inquisition.)
| | | | Thayenga
Germany
Local time: 15:00
Member (2009)
English to German
+ ...
Why? Probably because it (they) can. There is no other reason that would make sense.
And thank you for the smile at Monty Pythons'.
[Edited at 2026-10-08 12:16 GMT]
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Poll: How do you feel about tech companies using public minority language content to train AI? | 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 
