| Σελίδες για το θέμα: < [1 2 3] | People answering questions with AI output Αποστολέας σε συζήτηση: Zea_Mays
|
|---|
expressisverbis wrote:
Thomas T. Frost wrote:
Askers can restrict their questions to natives or users who work in certain languages pairs. If they don't do that, they may get all sorts of unwanted entries.
That is not quite true. I have posted a question today, and restricted it to natives only. I got an answer from that same member with tonnes of languages... she is not a PT-PT native.
You're so lucky to get answers from people who are so competent that they are native in 50 languages or whatever the number is.
Joking apart, it seems the restriction doesn't actually restrict. Here is the help information:
2.11 - Can KudoZ askers direct their questions to specific site users?
Yes. When posting a KudoZ question, askers are given the option to restrict their KudoZ questions by 1) profile information (native language, language pair and field) and 2) membership status. If the asker selects the option "Only ProZ.com members may answer" when posting his/her question, the "Answer" button will not be visible to non-members and the following message will show to them:
The asker has directed this question specifically to ProZ.com users who:
... are ProZ.com members (You do not appear to be) Join ProZ.com >
If the asker specifies a restriction connected with native language, level of expertise in the specified language pair or level of expertise in the field, those that do not meet these criteria will be informed of this, but still invited to post an answer:
The asker has directed this question specifically to ProZ.com users who:
... have expressed interest in this pair (You have)
... report a "specialty" level of expertise in the subject field (You have not) [Edit]
... are native speakers of Tonga (Nya) or Nzima (You have not reported that you are) [Edit]
Although you do not meet the criteria, you may suggest a translation if you like
So my suggestion is no solution at all, I'm afraid.
Maybe it's better just to use AI. Google's Bard is now available and does some things better than ChatGPT, not least since the former is connected to the internet and the latter is not. I read that Claude should be even better, but for the time being it's only available in the US and the UK.
| | | | Zea_Mays 
Ιταλία
Local time: 09:11
Αγγλικά σε Γερμανικά
+ ...
ΞΕΚΙΝΗΣΕ ΤΟ ΘΕΜΑ | Bot involved? | Jul 17, 2023 |
expressisverbis wrote:
I have posted a question today, and restricted it to natives only. I got an answer from that same member with tonnes of languages... she is not a PT-PT native.
Despite everything, I still managed to smile when I saw the answer because she has no idea what false friends are in Portuguese and French.
AI is useful when it falls into the right hands, if it falls into the hands of opportunists the result is this one.
I can use it for doubts and questions of my own and if I really need it (for example, in case clients can't provide me with an answer), but I would never do it to try to answer colleagues with the sole purpose of collecting points or for whatever reason
The question is also, why an INDIVIDUAL can state to translate to and from HUNDREDS of languages (how long did it take to enter all those combinations, btw?).
The person in question is linked to a translation agency, where having a large number of combinations is common instead.
| | | |
Angie Garbarino wrote:
Maria Teresa Borges de Almeida wrote:
In Portuguese Ai is an interjection that expresses pain or joy…
Ai que dolor! (che dolore)
Ai que dor! (PT(pt)...
| | | | expressisverbis
Πορτογαλία
Local time: 08:11
Μέλος από 2015
Αγγλικά σε Πορτογαλικά
+ ...
| Language combinations | Jul 17, 2023 |
Zea_Mays wrote:
The question is also, why an INDIVIDUAL can state to translate to and from HUNDREDS of languages (how long did it take to enter all those combinations, btw?).
The person in question is linked to a translation agency, where having a large number of combinations is common instead.
Angie mentioned before here that being an agency the language pairs are restricted to 40 if my memory serves me well.
The question is that even if she is asked to restrict that long list, she will still answer Kudoz questions in every language pair she can, so I don't think it's worth alerting the staff or even discussing this matter here.
On the other hand, it is obvious to any translator who values their profession that this person makes serious errors, she does not know grammar, misleads colleagues and even her suggestions are too literal when it comes to highly technical terms that even I myself have difficulty with.
I understand this can't be easy to manage by whoever is responsible, but I believe something should be done because today it's her, tomorrow it will be someone else... and it won't take long to see our grandmas, step-mums, uncles, etc. registering on Proz and translating from/to all the languages in the world.
Maybe I'm overreacting, but I don't fear technology, I fear the use that certain users make of it for their own gain in an unfair, unprofessional and dishonest way.
| | |
|
|
|
Zea_Mays
Ιταλία
Local time: 09:11
Αγγλικά σε Γερμανικά
+ ...
ΞΕΚΙΝΗΣΕ ΤΟ ΘΕΜΑ | How many languages can you be native in? | Jul 18, 2023 |
expressisverbis wrote:
Angie mentioned before here that being an agency the language pairs are restricted to 40 if my memory serves me well.
The question is that even if she is asked to restrict that long list, she will still answer Kudoz questions in every language pair she can, so I don't think it's worth alerting the staff or even discussing this matter here.
I was a bit surprised when I saw how many language combinations you can have as an individual on Proz.
As a bilingual, I know that you can be a native speaker of more than one language, but how many can those actually be? I think 10 would be a reasonable limit, very few people will be native speakers of that many languages.
So my suggestion would be to limit the number of language pairs an individual can have.
I also agree that this is just the beginning of a growing trend.
There's also increasing reaction from peers, adding comments to the AI-generated answers, which helps raise awareness of this behavouir.
| | | | | What the rules say | Jul 18, 2023 |
Site rule 2:
Mutual respect, professionalism and fair play are expected.
Site users are expected to treat each other with courtesy, whether posting publicly or making direct contact, and are advised to act under the assumption of good faith. Harassment of, or attacks or ad hominem statements on, individuals or groups, of any form, as well as discouragement of another's use of the site, will not be tolerated. No action aimed at gaining unfair adv
... See more Site rule 2:
Mutual respect, professionalism and fair play are expected.
Site users are expected to treat each other with courtesy, whether posting publicly or making direct contact, and are advised to act under the assumption of good faith. Harassment of, or attacks or ad hominem statements on, individuals or groups, of any form, as well as discouragement of another's use of the site, will not be tolerated. No action aimed at gaining unfair advantage in KudoZ, the directory or elsewhere, whether taken alone or as a group, will be tolerated.
The rules say nothing about the use of AI, but if a user goes around spamming a large number of Kudoz combinations with low-quality answers in languages they clearly don't master, it would seem fair to report that user and let the site decide what to do about them. ▲ Collapse
| | | | neilmac
Ισπανία
Local time: 09:11
Ισπανικά σε Αγγλικά
+ ...
Zea_Mays wrote:
Kudoz is obsolete, isn't it?  [/quote]
I still find kudoz useful, and I also still enjoy contributing to it.
| | | | Daryo
Ηνωμένο Βασίλειο
Local time: 08:11
Σερβικά σε Αγγλικά
+ ...
| That's the kind of arguments | Oct 18 |
Thomas T. Frost wrote:
The primary purpose of Kudoz is to help the Asker find the correct answer. If someone manages to provide that faster and better with an AI chatbot than a native, then what is wrong with that? The correct term is not less correct just because it was suggested by an AI chatbot.
That's the kind of arguments used to "prove" that smoking is good for your health. All you have to do is find few very very old people who chain-smoked all their life and you have "proven" there's nothing wrong with smoking.
Using the same kind of statistical sleight of hand you can "prove" there's nothing wrong with using AI instead of your own brain, as occasionally the AI output will be perfectly acceptable.
Only minor problem being, if you don't know already enough about the subject matter, you will have no ways of knowing when AI is "hallucinating" and when the AI output makes sense.
Keep the bots away.
| | |
|
|
|
Daryo wrote:
Thomas T. Frost wrote:
The primary purpose of Kudoz is to help the Asker find the correct answer. If someone manages to provide that faster and better with an AI chatbot than a native, then what is wrong with that? The correct term is not less correct just because it was suggested by an AI chatbot.
That's the kind of arguments used to "prove" that smoking is good for your health. All you have to do is find few very very old people who chain-smoked all their life and you have "proven" there's nothing wrong with smoking.
This sort of argumentation is fallacious and does not stand up to logical scrutiny. First of all, it is well known that smoking harms your health, whereas AI does not affect anyone's health.
Daryo wrote:
Using the same kind of statistical sleight of hand you can "prove" there's nothing wrong with using AI instead of your own brain, as occasionally the AI output will be perfectly acceptable.
The exact same argument applies to every other linguistic resource you can think of: paper dictionaries, online dictionaries, term bases, translation memories, you name it. In each case, we have to decide which result from which resource is the right one, whether or not it is from AI. Did it occur to you that you cannot use AI output uncritically, just as you cannot use any other linguistic resource uncritically? A good researcher will use all resources available – wisely. An amateur will take the first word in the dictionary, as some of us regularly notice when revising others' translations.
Daryo wrote:
Only minor problem being, if you don't know already enough about the subject matter, you will have no ways of knowing when AI is "hallucinating" and when the AI output makes sense.
In that case, it would be better not to answer because if you know nothing about the subject matter, you will not know if the suggestion from any other resource is correct either.
Daryo wrote:
Keep the bots away.
Should we also keep the dictionaries, TBs, TMs and other resources away? Are we only allowed to answer, according to you, if we know the correct answer by heart? Research is banned?
With or without AI, some users post unsuitable answers, as is obvious to anyone who has used Kudoz. In a recent Kudoz question, half the answerers did not understand that 'top heavy' means that the centre of gravity is high and thought it just meant 'very heavy'. They clearly did not use AI, as it would have told them the correct meaning. So how do we deal with users who post incorrect answers not influenced by AI? The problem is the user, not the tool (puns would seem inappropriate here).
It can be tempting, every time a problem is observed, to ask for more rules and bans, just like politicians publishing a steady flow of regulations – until everything is regulated so much that it becomes unworkable. Some people will publish inane answers. They could be filtered out if every Kudoz answer had to be approved by a Kudoz board, but it would ruin Kudoz. Or Kudoz could be shut down. Then there would be no more incorrect answers. We should not make the solution worse than the problem.
| | | |
I've never seen AI-like answers in section of KudoZ, so whatever I can say about it is probably irrelevant. When writing another post in another thread, l noticed, however, that the writing aid here has become a lot better than it used to be (unlike the one that goes with my phone keyboard). So I used it a few times even though reaching for the Tab key is quite a task: it's a small, inconveniently located key on my laptop. I don't know what powers that writing aid, btw
| | | | Daryo
Ηνωμένο Βασίλειο
Local time: 08:11
Σερβικά σε Αγγλικά
+ ...
| A bit more "nonsense" ... | Oct 20 |
Thomas T. Frost wrote:
Daryo wrote:
Thomas T. Frost wrote:
The primary purpose of Kudoz is to help the Asker find the correct answer. If someone manages to provide that faster and better with an AI chatbot than a native, then what is wrong with that? The correct term is not less correct just because it was suggested by an AI chatbot.
That's the kind of arguments used to "prove" that smoking is good for your health. All you have to do is find few very very old people who chain-smoked all their life and you have "proven" there's nothing wrong with smoking.
This sort of argumentation is fallacious and does not stand up to logical scrutiny. First of all, it is well known that smoking harms your health, whereas AI does not affect anyone's health.
Daryo wrote:
Using the same kind of statistical sleight of hand you can "prove" there's nothing wrong with using AI instead of your own brain, as occasionally the AI output will be perfectly acceptable.
The exact same argument applies to every other linguistic resource you can think of: paper dictionaries, online dictionaries, term bases, translation memories, you name it. In each case, we have to decide which result from which resource is the right one, whether or not it is from AI. Did it occur to you that you cannot use AI output uncritically, just as you cannot use any other linguistic resource uncritically? A good researcher will use all resources available – wisely. An amateur will take the first word in the dictionary, as some of us regularly notice when revising others' translations.
Daryo wrote:
Only minor problem being, if you don't know already enough about the subject matter, you will have no ways of knowing when AI is "hallucinating" and when the AI output makes sense.
In that case, it would be better not to answer because if you know nothing about the subject matter, you will not know if the suggestion from any other resource is correct either.
Daryo wrote:
Keep the bots away.
Should we also keep the dictionaries, TBs, TMs and other resources away? Are we only allowed to answer, according to you, if we know the correct answer by heart? Research is banned?
With or without AI, some users post unsuitable answers, as is obvious to anyone who has used Kudoz. In a recent Kudoz question, half the answerers did not understand that 'top heavy' means that the centre of gravity is high and thought it just meant 'very heavy'. They clearly did not use AI, as it would have told them the correct meaning. So how do we deal with users who post incorrect answers not influenced by AI? The problem is the user, not the tool (puns would seem inappropriate here).
It can be tempting, every time a problem is observed, to ask for more rules and bans, just like politicians publishing a steady flow of regulations – until everything is regulated so much that it becomes unworkable. Some people will publish inane answers. They could be filtered out if every Kudoz answer had to be approved by a Kudoz board, but it would ruin Kudoz. Or Kudoz could be shut down. Then there would be no more incorrect answers. We should not make the solution worse than the problem.
You keep mixing what I did say with what I did not say. Plenty of things you can "prove" that way ...
To keep it short: expecting from a bot to serve you on a plate a "perfect digest" of whatever you need is mental laziness, pure and simple. I still haven’t seen a bot that can find better information than what I can find myself. You maybe get some kind of answer quickly, but so far it looks to me as false economy of time.
As for "With or without AI, some users post unsuitable answers" and "if you don't know already enough about the subject matter ...(In that case,) it would be better not to answer" I can only agree. Where did I say anything to the contrary?
What I mean by "Keep the bots away" is that simply copy-pasting the bot's output and presenting it as if it was your work is NOT "answering a question", be it a Kudoz question - or any other question.
@Denis Fesik
It looks like your language pairs were left unaffected by this now terminated experiment:
https://www.proz.com/forum/prozcom_suggestions/363811-please_remove_chatgpt_option_from_kudoz_staff_removed.html
so you never saw any of the creative writing of this "fellow translator": https://www.proz.com/profile/670924
Like these https://www.proz.com/profile/670924?popup=kudoz&show_mode=KudoZ&mode=answered&pair=eng_rus
| | | | | On a positive note | Oct 20 |
we’ve got rid of many basic/non PRO KudoZ questions.
[Edited at 2024-10-20 10:43 GMT]
| | | | | Σελίδες για το θέμα: < [1 2 3] | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » People answering questions with AI output | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| | 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 
