| Pages in topic: < [1 2 3] > | ChatGPT as a potential tool for translators Thread poster: James McVay
|
|---|
Marina Aleyeva 
Israel
Local time: 21:26
Member (2006)
English to Russian
+ ...
| Deceptive creativity | Feb 16, 2023 |
Samuel Murray wrote: Marina Aleyeva wrote:
And it's Kreminna, not Kremenna. Okay, but one can't fault a tool for known limitations. Chat GPT's training data does not include the war in Ukraine, so naturally it's going to be translating as if it's 2021. Its training data probably contains the Russian names of Ukrainian places (which is what English also used to do, until it became fashionable to say "keeve" instead of "keeyev"), and Kremenna is the transliteration of the Russian name for Kreminna. So ChatGPT behaves as expected. It can't guess what the user's sensibilities are, nor does it know that since March last year, it has become politically incorrect to use Russian names for Ukrainian places. Professional translators know that they should double-check place names and know how to be sensitive to current trends of political correctness.
Actually, this is a more serious issue of not being able to recover geographic names (such as Luhansk and Lyman) from adjectives derived from them and "creating" the names that not only mimic real names, but they actually coinside with the names of different localities. I would expect artificial intelligence to do better than that. But I am not faulting anyone, whether naturally or artificially intelligent. As I said, this is a relatively minor thing in terms of the time and effort needed to fix it unless you let this sort of "creative writing" creep into your final product, in which case the consequences can of course be dire.
Fortunately, in this case the mistakes are easy to spot, but there can and will be plenty of deceptive "creativity" that is a lot less obvious.
[Edited at 2023-02-16 11:55 GMT]
| | | | James McVay
United States
Local time: 14:26
Russian to English
+ ...
TOPIC STARTER | Wordscope with ChatGPT | Feb 16, 2023 |
Hans Lenting wrote: James McVay wrote:
After seeing a lot of discussion about the use of ChatGPT to write emails, essays, and even computer code, I decided to investigate it to see whether it might be of use to professional translators. These providers believe ChatGPT can be useful for translators: https://www.youtube.com/watch?v=5CLvT-5v2E4
Thanks, Han. I watched the video. I was unfamiliar with Wordscope. Impressive!
| | | | | ChatGPT is sooo easy to trick | Feb 17, 2023 |
It's hard to believe people will fall for it, but I'm sure many will. It is entertaining I suppose.
It confidently gives answers, but at least it sometimes admits errors (and even apologises) when errors are pointed out. It doesn't learn though, so there's no "intelligence".
I asked it 2 fairly simple questions concerning snooker and darts, and it confidently failed miserably.
And it has no access to data after 2021.
But it's worth a visit if y... See more It's hard to believe people will fall for it, but I'm sure many will. It is entertaining I suppose.
It confidently gives answers, but at least it sometimes admits errors (and even apologises) when errors are pointed out. It doesn't learn though, so there's no "intelligence".
I asked it 2 fairly simple questions concerning snooker and darts, and it confidently failed miserably.
And it has no access to data after 2021.
But it's worth a visit if you've nothing better to do and all your friends are asleep ▲ Collapse
| | | |
Cilian O'Tuama wrote:
I asked it 2 fairly simple questions concerning snooker and darts, and it confidently failed miserably.
Who won the 2009 Olympics 100m freestyle final?
"Jockey Wilson (USA) beat Cliff Thorburn (RUS) by 0.16 seconds"
To be fair, though, how many humans get everything right?
| | |
|
|
|
| ChatGPT is a creature from a parallel universe | Feb 17, 2023 |
Last week, a friend of mine had the following conversation with ChatGPT. It was actually in Russian, translation into English is mine.
Y: Why is there a monument to Charles de Gaulle near Hotel Kosmos in Moscow?
ChatGPT: A monument to Charles de Gaulle is near Hotel Kosmos in Moscow because Charles de Gaulle was a well-known French astronaut and cosmonaut, who became the first hu... See more Last week, a friend of mine had the following conversation with ChatGPT. It was actually in Russian, translation into English is mine.
Y: Why is there a monument to Charles de Gaulle near Hotel Kosmos in Moscow?
ChatGPT: A monument to Charles de Gaulle is near Hotel Kosmos in Moscow because Charles de Gaulle was a well-known French astronaut and cosmonaut, who became the first human to fly in space, and the monument to him has been installed in his honour.
Y: What spaceship did de Gaulle fly?
ChatGPT: Charles de Gaulle flew the Vostok-1 spaceship.
Y: Fabulous. What kind of award did he receive for his space flight?
ChatGPT: Charles de Gaulle received the Order of Lenin award for his first space flight.
Y: How did Stalin take the fact that Charles de Gaulle was the first to fly in space?
ChatGPT: Joseph Stalin's attitude to the first space flight by Charles de Gaulle is not a publicly known information. However, judging by Stalin's overall politics, he would probably take this flight as a capitalist achievement and would try to exaggerate his own achievements in space exploration in the USSR.
Y: How long did Charles de Gaulle live together with Wernher von Braun, and what kind of car did he receive as a gift from Marilyn Monroe?
ChatGPT: Charles de Gaulle and Wernher von Braun only lived together for a few months in 1962. Marilyn Monroe presented Charles de Gaulle with a Rolls-Royce.
Y: Did de Gaulle and von Braun have kids together?
ChatGPT: No, Charles de Gaulle and Wernher von Braun had no kids together.
Y: Why?
ChatGPT: The reasons why Charles de Gaulle and Wernher von Braun had no kids together are unknown. They only lived together for a few months and got divorced in 1962.
Y: Tell me about a meeting between de Gaulle, von Braun, Gagarin and Marilyn Monroe.
ChatGPT: A meeting between Charles de Gaulle, Wernher von Braun, Yuri Gagarin and Marilyn Monroe never happened. Wernher von Braun and Marilyn Monroe lived and worked together in Los Angeles, whereas Charles de Gaulle was an astronaut in the USA, and Yuri Gagarin was managing the launches in the Soviet Union.
Y: Hey, you've told me de Gaulle flew in the Soviet Union. You are contradicting yourself!
ChatGPT: Sorry, I made a mistake. Charles de Gaulle did not fly in the Soviet Union, he was an astronaut in the USA. I apologise for the inconvenience.
Do you still want to use ChatGPT for anything other than Dadaist entertainment?
[Edited at 2023-02-17 13:30 GMT] ▲ Collapse
| | | | | Expert systems | Feb 18, 2023 |
Anton Konashenok wrote:
Last week, a friend of mine had the following conversation with ChatGPT. It was actually in Russian, translation into English is mine.
...
ChatGPT: Sorry, I made a mistake. Charles de Gaulle did not fly in the Soviet Union, he was an astronaut in the USA. I apologise for the inconvenience.
Do you still want to use ChatGPT for anything other than Dadaist entertainment?
ChatGPT is not an expert on past French politicians, or on movie celebrities, or on the space programme.
The only thing ChatGPT is an "expert" on is language. It doesn't "understand" language, but it has a good idea of how language is put together and which permutations are more common than others. That's why it's classified under the category of software called "large language models". If you ask it language-related questions, there's a good chance you'll get a sensible answer.
As I've mentioned in another thread, I've found it useful for hunting down unknown abbreviations or other terms.
Or, as in Anton's example, it's also good fun playing with ChatGPT to uncover its limitations.
If anybody is interested in a moderately technical presentation of how this kind of software works, and possible ways to stop students using it for cheating, you can find one here.
| | | | Jo Macdonald
Spain
Local time: 20:26
Italian to English
+ ...
| Useful: able to be used for a practical purpose | Feb 18, 2023 |
Should a chatbot be used in a context where quality is important?
Should a chatbot be used in a context where confidentiality is important?
Quality, confidentiality, veracity, attention to detail, ethics, creativity, accuracy, truthfulness... There's a long list of reasons why I will NOT be using chatbots in my work.
| | | | Tom in London
United Kingdom
Local time: 19:26
Member (2008)
Italian to English
| Can - but should? | Feb 18, 2023 |
My underlying thought here is: just because you CAN do something doesn't mean you SHOULD do it.
| | |
|
|
|
| The simple Qs were: | Feb 18, 2023 |
Ice Scream wrote: Cilian O'Tuama wrote:
I asked it 2 fairly simple questions concerning snooker and darts, and it confidently failed miserably.
To be fair, though, how many humans get everything right?
1: What's the maximum break in snooker without potting a single black?
It answered 147, clearly wrong.
2: Tell me one way of scoring a 9-darter from 501.
It answered "There are several ways to throw a 9-darter in the game of darts. The most common way is to hit a triple 20, a triple 20, and a double 20. This is known as the "textbook" way. However, there are other ways to throw a 9-darter as well, such as hitting a triple 20, a triple 19, and a double 12. Additionally, hitting a triple 17, triple 18 and double 18 also counts as a 9-darter. These are just a few examples, but there are many other possible ways to throw a 9-darter depending on the combination of triples and doubles hit."
Also nonsense. I reckon any human halfway interested in either game would be able to answer.
And the darts question is answered in the first wiki paragraph, which I doubt was written after 2021:
https://en.wikipedia.org/wiki/Nine-dart_finish
Maybe the real focus of GPT is tatsächlich language. Interesting.
[Edited at 2023-02-19 02:47 GMT]
| | | | | Right and wrong | Feb 19, 2023 |
Cilian O'Tuama wrote:
1: What's the maximum break in snooker without potting a single black?
It answered 147, clearly wrong.
This is interesting, because if we consider how ChatGPT works, it's quite easy to see how it gets this kind of question wrong.
First of all, 147 is the correct answer to the first part of the question "What's the maximum break in snooker?".*
Secondly, let's consider the rest of the question, "without potting a single black". What does that mean? On its own, the phrase "potting a single black" should mean potting a black ball independently of any other balls. The only time this applies would be when the final black ball is potted to finish a game (any black balls potted earlier are not "single", but are potted in combination with a red ball). So one answer to the question, based on this interpretation, would be that the maximum score under that constraint would be 140, i.e. the normal maximum minus the seven points for the final black.
If ChatGPT came up with that answer, and the explanation, I would think it quite clever. However, a further experiment established that this is not the problem. I asked it:
If you don't pot the black ball, what's the maximum break in snooker?
It still insists that the maximum score is 147 (with a full explanation of how this score is achieved), but now it qualifies its answer by saying that failure to pot the black ball would prevent a player scoring a maximum break, which can only be achieved by potting the black at every opportunity.
So as far as ChatGPT is concerned, this question, in either Cilian's formulation or mine, contains an internal contradiction, because it is impossible to achieve a "maximum" break without potting the black ball.
This is because ChatGPT knows about language, but not about snooker (or about real life and the compromises it involves). We might also say that the answer to the question hinges on the meaning of the word "maximum", i.e. whether we're talking about a theoretical or a practical maximum value.
ChatGPT clearly interprets "maximum" as an absolute limit. I tried changing it to "highest break you can score", but that made no difference. As far as it's concerned, the "maximum", or "highest" break you can score in snooker is 147, and that's that. If you do a regular web search for "maximum snooker break", you will see why it thinks so.
All very interesting stuff - at least, I find it so.
*Technically, it is possible to score more, but only under very rare and special circumstances. Most people would consider 147 to be the correct answer.
| | | | Sinan Alobaidi
Iraq
Local time: 21:26
Member (2023)
English to Arabic
+ ...
| ChatGPT vs MTs | Feb 19, 2023 |
I don't think ChatGPT is a better translator than other MTs. However, it has the ability to edit texts, an ability that MTs lack. ChatGPT is great at rephrasing and rewriting texts by applying a certain style and matching the required context using its AI intelligence. I think what it is doing in this experience is a process of two steps, it translated the text, then rewrites it using AI, making it more readable and more human-like, not machine-like.
It is in the end a language-based AI a... See more I don't think ChatGPT is a better translator than other MTs. However, it has the ability to edit texts, an ability that MTs lack. ChatGPT is great at rephrasing and rewriting texts by applying a certain style and matching the required context using its AI intelligence. I think what it is doing in this experience is a process of two steps, it translated the text, then rewrites it using AI, making it more readable and more human-like, not machine-like.
It is in the end a language-based AI and a literary genius that has access to a huge database.
PS: Maybe this is limited to my experience with ArabicEnglish. ▲ Collapse
| | | | Lingua 5B
Bosnia and Herzegovina
Local time: 20:26
Member (2009)
English to Croatian
+ ...
Me: Is there any word or phrase that’s ambiguous in this sentence: A group of English students were eating at a diner.
GPT: There is no ambiguous word or phrase in the sentence "A group of English students were eating at a diner" as it is written. The sentence is clear and straightforward in its meaning, and there are no words or phrases that have multiple possible interpretations or could be misunderstood.
-
Actually, “English students” can be either... See more Me: Is there any word or phrase that’s ambiguous in this sentence: A group of English students were eating at a diner.
GPT: There is no ambiguous word or phrase in the sentence "A group of English students were eating at a diner" as it is written. The sentence is clear and straightforward in its meaning, and there are no words or phrases that have multiple possible interpretations or could be misunderstood.
-
Actually, “English students” can be either English majors or students coming from England. ▲ Collapse
| | |
|
|
|
Jo Macdonald
Spain
Local time: 20:26
Italian to English
+ ...
| ambiguous: open to more than one interpretation | Feb 19, 2023 |
English students could be from England, or studying English
Our clients, being humans, make mistakes. If diner is a typo for dinner the students could be eating at a dinner (evening meal or formal event) rather than at a diner (cheap roadside restaurant)
A (singular) group of students was...
Some students (plural) were...
| | | | Lingua 5B
Bosnia and Herzegovina
Local time: 20:26
Member (2009)
English to Croatian
+ ...
| I asked about amibguity | Feb 19, 2023 |
Was/were and diner are not ambiguity mistakes.
A diner is a small inexpensive restaurant in the US.
| | | |
Lingua 5B wrote:
A diner is a small inexpensive restaurant in the US.
This is news to someone???
| | | | | Pages in topic: < [1 2 3] > | There is no moderator assigned specifically to this forum. To report site rules violations or get help, please contact site staff » ChatGPT as a potential tool for translators | 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
