Pages in topic: [1 2 3 4] > | Poll: How often do you use DeepL or other machine translation tools? Thread poster: ProZ.com Staff
|
This forum topic is for the discussion of the poll question "How often do you use DeepL or other machine translation tools?".
View the poll results »
| | | neilmac Spain Local time: 07:37 Spanish to English + ...
Nowadays I use an MT tool most days. However, I don't regard it as a panacea or 100% reliable, and like to think I use it judiciously, as I would any other tool. For example, I never just run a whole document through it, which some people are apparently quite happy to do with DeepL. Instead, I use it sentence by sentence, or in chunks, and compare the MT result with my own translation. The first time I tried DeepL, I tried it on lists of chemical names and although some of the result... See more Nowadays I use an MT tool most days. However, I don't regard it as a panacea or 100% reliable, and like to think I use it judiciously, as I would any other tool. For example, I never just run a whole document through it, which some people are apparently quite happy to do with DeepL. Instead, I use it sentence by sentence, or in chunks, and compare the MT result with my own translation. The first time I tried DeepL, I tried it on lists of chemical names and although some of the results were correct, others were erroneous, which put me off using it initially. Before things like DeepL and Google translate were invented, around the turn-of-the-century I was also using Systran in a similar way, although it works differently. However, at some point they changed the interface and it became too fiddly and time-consuming to use, so I gave up on it. After that, I went for a few years without using any kind of MT assistance at all. In my opinion, like any other tool, these things are only as useful and good as the person who is leveraging them. ▲ Collapse | | | Very, very occasionally | May 10, 2022 |
I use Deepl very occasionally (when I am tired) as a very rough draft just to speed the typing up a bit. | | | Tom in London United Kingdom Local time: 06:37 Member (2008) Italian to English As it happens | May 10, 2022 |
Teresa Borges wrote: I use Deepl very occasionally (when I am tired) as a very rough draft just to speed the typing up a bit. As it happens I used it yesterday to do the first rough draft of a translation but it's fairly useless. Today I'm working through the draft, rewriting it into acceptable English. It's handy up to a point because it does slightly speed up the initial, messy rough translation, but is only a starting point and depending on the source text, may be more or less useful/useless. As an alternative I often use Mac Dictation to just speak the translation directly from the Italian into English, in a new Word document. | |
|
|
Never ever ever in a million yearz | May 10, 2022 |
Tom in London wrote: As it happens I used it yesterday to do the first rough draft of a translation but it's fairly useless. Today I'm working through the draft, rewriting it into acceptable English. It's handy up to a point because it does slightly speed up the initial, messy rough translation, but is only a starting point and depending on the source text, may be more or less useful/useless. I'm genuinely shocked. It's not that long since you said you'd use CAT over your dead body!! I, however, refuse to get with the kidz. | | | For best results, Spoonfeed | May 10, 2022 |
Despite the improvements of MT, one thing hasn't changed: It works best when you spoonfeed it with small amounts of text (Single terms, 1-2 sentences). For this reason, I only use it when I come across very long or convoluted sentences or to look up single terms that I can't find anywhere else. | | | Alex Lichanow Germany Local time: 07:37 Member (2020) English to German + ...
I do not really use DeepL or any other MT engine for whole translations, but rather to give me an idea of the meaning of long, convoluted sentences. I'll rather take my time to come up with my own translations than have everything MT'd and then post-edit the questionable results. To be fair, at least in EN-DE, DeepL is actually surprisingly good at the horrible non-language that is Legalese, so it helps a lot when translating contracts, EULAs etc. | | | Thayenga Germany Local time: 07:37 Member (2009) English to German + ...
The only time I used DeepL and its kind is when I need a good laugh. Nothings is better than genuine human translation. Maybe one day these translation tools will be good, but that won't be during the time I still have left on this planet.
[Edited at 2022-05-10 12:49 GMT] | |
|
|
Lieven Malaise Belgium Local time: 07:37 Member (2020) French to Dutch + ... As often as I can. | May 10, 2022 |
I use it as often as I can. It's the future and I want to be as fast as possible by the time most of the job offers will be MTPE. | | |
Deepl is the best MT tool I have ever found. By far. It may, indeed, be useless for the translation of SOME kinds of texts (artistic, poetic, fiction...) but for many others it is frankly impressive, and constitutes an indispensable resource. It's a truly remarkable and powerful tool, not only for the breadth of its memory bank, nailing thousands of terms that would require tedious research, but it even comes up with some surprisingly good turns of phrase more often than I ever w... See more Deepl is the best MT tool I have ever found. By far. It may, indeed, be useless for the translation of SOME kinds of texts (artistic, poetic, fiction...) but for many others it is frankly impressive, and constitutes an indispensable resource. It's a truly remarkable and powerful tool, not only for the breadth of its memory bank, nailing thousands of terms that would require tedious research, but it even comes up with some surprisingly good turns of phrase more often than I ever would have imagined. Of course, like any MT, you have to watch it as if it were a child playing with a chainsaw, but one develops that capacity, and it makes fewer gaffes than anything I've come across before. Again, by using them I am able to spend more time on proofreading and polishing style, rather than rattling away at my keyboard. I never cease to be amazed at the widespread reticence to use these kinds of tools, but, to each his own, I guess. ▲ Collapse | | | Alex Lichanow Germany Local time: 07:37 Member (2020) English to German + ... No chainsaw, no problems | May 10, 2022 |
Justin Peterson wrote: Of course, like any MT, you have to watch it as if it were a child playing with a chainsaw, but one develops that capacity, and it makes fewer gaffes than anything I've come across before. Again, by using them I am able to spend more time on proofreading and polishing style, rather than rattling away at my keyboard. I never cease to be amazed at the widespread reticence to use these kinds of tools, but, to each his own, I guess. Quite simple from my perspective: no kid with a chainsaw = no need to watch = more time for myself. Also, PERSONALLY, I am simply faster when translating by myself than when I try to post-edit MT output, most of which turns out to be complete gibberish more often than not. I will give DeepL that it is WAY better than pretty much every other engine I have encountered, but one of the reasons I still do not offer MTPE is that most clients do not use DeepL, instead relying on gibberish engines like the one from Microsoft. Combine Microsoft's "less-than-ideal" MT results with the lowered rates most clients offer for MTPE and I can just as well go live under a bridge. | | |
Thayenga wrote: The only time I used DeepL and its kind is when I need a good laugh. Nothings is better than genuine human translation. Maybe one day these translation tools will be good, but that won't be during the time I still have left on this planet.
[Edited at 2022-05-10 11:28 GMT] *** I am not denying that this may have your been experience with it, but I guess it depends entirely on the source texts you put into it. If you put, for example, a newspaper article into it ... I guarantee you will not be laughing. In fact, your jaw might drop at how good it is. | |
|
|
It depends on the content, of course | May 10, 2022 |
Alex Lichanow wrote: Justin Peterson wrote: Of course, like any MT, you have to watch it as if it were a child playing with a chainsaw, but one develops that capacity, and it makes fewer gaffes than anything I've come across before. Again, by using them I am able to spend more time on proofreading and polishing style, rather than rattling away at my keyboard. I never cease to be amazed at the widespread reticence to use these kinds of tools, but, to each his own, I guess. Quite simple from my perspective: no kid with a chainsaw = no need to watch = more time for myself. Also, PERSONALLY, I am simply faster when translating by myself than when I try to post-edit MT output, most of which turns out to be complete gibberish more often than not. I will give DeepL that it is WAY better than pretty much every other engine I have encountered, but one of the reasons I still do not offer MTPE is that most clients do not use DeepL, instead relying on gibberish engines like the one from Microsoft. Combine Microsoft's "less-than-ideal" MT results with the lowered rates most clients offer for MTPE and I can just as well go live under a bridge. *** Yes, if I see it's not going to be efficient, I simply don't use it. Each job requires a different set of tools. Yes, there are situations where it is pretty useless. It depends on the kind of material you're translating. Luckily, I don't have clients (yet) obligating me to use those tools you mention, or applying discounts based on them. Fingers crossed. So far, so good. | | |
Justin Peterson wrote: Of course, like any MT, you have to watch it as if it were a child playing with a chainsaw As a dad of two kids, I can totally relate! | | |
First, MTPE is boring and brain-numbing. Second, it would be even more boring and time-consuming to fix it in the kind of content I get. Third, I'm sure MT can be helpful. Anyway, the final outcome wouldn't really be mine, but anybody else's. People pay me for what I meself produce personally and wouldn't want something that every living being as well as machines can come up with. From recently, nothing too linguistically elaborate: Source: Move the schmoll* ... See more First, MTPE is boring and brain-numbing. Second, it would be even more boring and time-consuming to fix it in the kind of content I get. Third, I'm sure MT can be helpful. Anyway, the final outcome wouldn't really be mine, but anybody else's. People pay me for what I meself produce personally and wouldn't want something that every living being as well as machines can come up with. From recently, nothing too linguistically elaborate: Source: Move the schmoll* backwards as far as possible DeepL/translator: Déplacez le schmoll vers l'arrière aussi loin que possible Me, after losing all faith in mankind: Reculez le schmoll au maximum DeepL back translation: Move the schmoll back as far as possible Human one, MT zero Philippe ▲ Collapse | | | Pages in topic: [1 2 3 4] > | 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 often do you use DeepL or other machine translation tools? Pastey | Your smart companion app
Pastey is an innovative desktop application that bridges the gap between human expertise and artificial intelligence. With intuitive keyboard shortcuts, Pastey transforms your source text into AI-powered draft translations.
Find out more » |
| Trados Business Manager Lite | Create customer quotes and invoices from within Trados Studio
Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |