Poll: How often do you use DeepL or other machine translation tools?

jyuan_us wrote:

It would be even better if your MTPE rate is equal to your regular translation rate, while your translation speed has more than doubled.

Roy Williams wrote:

Just because those jobs exist, it doesn't obligate you to take them. DeepL is nothing more than another tool, which if used correctly can be helpful. And if you're going to do MTPE, it's best to charge by the hour at a rate worth your time. Anyone who actually cares about the content they put out will be willing to pay.

Lieven Malaise wrote:



It would also be better if my clients not only pay me in money, but also offer me a car, a house and an all-in 30-day luxury cruise around the world per job I do for them.

A client that pays your translation rate for an MTPE job prepared and specifically offered by them is a client that gives away money for no reason. Let me know how many of such clients you can find.

jyuan_us wrote:


Maybe MTPE in your language pair works much better than mine. If that is the case, some of your peer translators would probably give a 50% discount going forward to beat your 75% discount. Your client will save even more while getting the same quality, let alone giving you an all-in 30-day luxury cruise around the world.

Novian Cahyadi wrote:



Anyone who actually cares about the content they put out will not use MTs to begin with. Let's be honest here, the only reason an agency would use MTs is to cut corners. Cutting corners and quality-oriented mindset don't belong in the same boardroom.

And seeing how some translators are glorifying MTs as the next best thing since sliced breads (special thanks to our colleagues for pushing this p̶r̶o̶p̶a̶g̶a̶n̶d̶a̶ b̶r̶a̶i̶n̶w̶a̶s̶h̶i̶n̶g̶ professional recommendation, Justin and Lieven), in the not so distant future, all translation assignments are going to be MTPE jobs. Eventually, you'll be obligated to take them.

And before anyone's wondering, yes, I've tried MTs to aid my translation tasks. But no, I don't find MT outputs to be particularly helpful. I work with movie scripts and marketing texts. During the time I was using those jaw-dropping MTs, in most cases I needed to overhaul the entire thing because the results from MTs were so horrible.

[Edited at 2022-05-12 16:58 GMT]

Roy Williams wrote:

I don't have any stats but It doesn't seem that there are many translators who think that highly of MT or deepl in particular. The danger is from non-translators who use and blindly trust it like Star Trek's universal translator.

Novian Cahyadi wrote:

Who happen to be the end-clients. Probably not a relevant nor an interesting fact to everyone, but I've come across several foreign sellers in our local online marketplaces. I tell you, their product descriptions were blatantly the results of an MT output. Granted, most of them were just selling the goods in bulk without really knowing what they were selling.

[Edited at 2022-05-13 07:54 GMT]