DeepL Translator

Translation was okay-ish.

I mean, DeepL at the time was incredibly doing incredibly well, well.

Um, and then you had the kind of Google Translate version of that.

But all of the components to do dub were, were not good enough yet.

And so they had a harness that ties together a bunch of other things.

Probably it was DeepL under the, or Deepgram, under the hood with a bunch of optimizations and more products.

Now they're trying to do notes, and they're trying to move more into productivity.

And I think they'll be successful with it, yeah.

To a computer, you are a nightmare.

When you start typing into Google Translate or DeepL, the system does not wait for you to finish your thought.

It cannot afford to wait.

If it waited for you to type a full period, you would get bored and switch apps.

And the phrase was coined by the German software engineer Niklas Grun.

He works for the tech firm DeepL.

And he's with us from Cologne.

Niklas, thank you for joining us on the media show.

So maybe you could take us through how does a live stream turn into other languages that end up on my TV?

Yeah, the best way I think to kind of describe it is like to understand the gap involved because there's a lot going on, right? So the first part of the gap is, okay, when you said, how do I translate? Okay, well, let's try to use, let's say an LLM, right? Like Gemini, for example, or there's other engines out there, DeepL, Google even, right? Okay, well, how do we get that into video, right? That's the other aspect of it.

Well, you can't say, okay, here's Gemini output.

Video itself has certain standards for captioning, subtitling, and dubbing.

While Google Translate provides extensive language support, users should be aware of its data tracking practices.

Alternatives like DeepL Translate are also available for those seeking privacy alongside language support.

Evaluating the M6 chip against the M5 Pro Apple's introduction of the M6 chip in the latest Mac Mini has sparked discussions about its performance compared to the M5 Pro.

While the M6 boasts advancements such as enhanced single-core performance and dual neural engines, it does not automatically outperform the M5 Pro in all areas.

And even many of the foremen would have only some English communication skills.

In those cases, I usually use Google or, in some cases, the DeepL translator apps.

And by and large, there's enough communication ability by mouth, my laser pointer, and communications to be able to get the point across.

Consequently, during the slow times, like lunch breaks or other breaks required by OSHA, I get involved with some conversations, and sometimes they involve religious concepts.

Mm.

Saying, "Oh, what are you gonna use languages for? Everyone, everyone can speak, everyone can open DeepL.ai or..." Knowing, being in, in the course itself, that it gives so much more, I do fear that, especially in the UK where there's been a general trend in the reduction of language choices-

Mm

... at GCSE and as an A level, that it's gonna also wither away at a degree level-

40% is huge.

But isn't machine translation something you can just outsource to Google or DeepL? Why would a language app even need to build its own?

And that's exactly the question their engineering team asked.

The thing is, Duolingo isn't just translating user-generated content on the fly.

My guest today is Amanda Johnson.

Amanda leads talent acquisition at DeepL, one of the world's most accurate AI translation companies, where she's been on a journey to build more structured, data-informed TA function from the ground up.

Before DeepL, she spent time at Uber, where she worked across talent and recruiting operations at serious scale and prior to that, Salesforce, too.

Amanda.

Hello.

I'm really happy to be here.

I'm very thrilled to have you on Hiring on All Cylinders today.

Um, before we jump into our conversation, can you just tell us a little bit about yourself and your role at DeepL?