Back
Jaroslaw Kutylowski
Founder and CEO, DeepL SE

Jaroslaw Kutylowski (DeepL) im Interview über Künstliche Intelligenz (KI) Börse im Ersten 21.02.2025

🎥 Feb 21, 2025 📺 German TV Highlights Official ⏱ 9m
In this exclusive interview on February 21, 2025, DeepL CEO Jaroslaw Kutylowski discusses the future of artificial intelligence, ...
Watch on YouTube

About Jaroslaw Kutylowski

Jaroslaw Kutylowski, CEO of DeepL, appeared on the Big Technology Podcast on July 8, 2026, to discuss the rise of specialized AI models. He argued that purpose-built models can offer better accuracy, lower latency, and reduced costs compared to large general-purpose systems, and noted that companies are increasingly using model routers to select the appropriate AI for each task. Kutylowski also highlighted real-time translation as a tool that could help businesses expand across borders, and described voice as the next frontier for AI. Kutylowski stated that AI translation tools like DeepL can reduce the upfront investment needed for companies to enter new markets by handling documentation, sales communication, and customer service in multiple languages. He described the ability for every person to talk to another person in the world as a "beautiful application of AI" that is worthwhile from both a business and human perspective.

Source: AI-verified profile updated from Jaroslaw Kutylowski's recent appearances. Browse all interviews →

Transcript (35 segments)
I
Interviewer0:00
Jaroslaw Kutylowski. Hello, hello, I'm glad to be here. I'm especially glad that you are here because it took months to find an appointment. You are a busy man. Are business going well, or is the competition making life difficult?
J
Jaroslaw Kutylowski0:17
Yes, I mean, as you said, it's a German company, but it's truly global. That means we have offices all over the world and are present in many markets. So the function is sometimes a bit more difficult. I think AI has always been a very competitive field, and we have always competed against strong competition, actually the biggest tech companies in the world. So that's something we grew up with. It's something we enjoy, but also something that is difficult every day and can cause some stomach aches.
I
Interviewer0:55
Now you are in the comfortable situation that the test results with your translation software are very, very good, leaving many others behind. But the competition is not sleeping; it is also getting stronger.
J
Jaroslaw Kutylowski1:09
That has actually been the case over the last 8 years of DeepL's history. We have always fought a hard battle. We build on understanding very well what our customers need. We mainly build for larger companies, for enterprises. And there it is important to really understand how translation is used, in which use cases, how translation is used, for example, in a law firm compared to a manufacturing company. And we combine that with this very strong tech focus, this very strong focus on research, and try to build the best product for the application. So far, that has been a very, very successful package.
I
Interviewer2:00
Can you then outline why your AI is so much better?
J
Jaroslaw Kutylowski2:06
Yes, I think that actually comes from this application-oriented research and from this approach of understanding the application very precisely, but also from the fact that we have always had a lead in the market. We were one of the first companies to bring AI-powered translation to market in 2017, when this hype was not really foreseeable. And since then we have been able to further expand and strengthen our position in the market. I think as one of the very few tech companies, we are also really involved in academic research, and that provides a lot of advantages.
I
Interviewer2:58
DeepL is constantly expanding its language spectrum. Can you explain how an AI learns a new language? Is it like with children?
J
Jaroslaw Kutylowski3:12
A little bit it is like with children. You look at a large dataset of translated data, but also partially non-translated data, data written in the target language. And AI networks learn overall by seeing very, very many data. In our case, it's these translations. And the model, which at the beginning cannot translate at all, doesn't know the language at all, with each sentence, each paragraph, each document it sees, it becomes a little bit better. And that is a very long process. It is a process that also requires a lot of computing power. At the end, for the last bit of quality, we need to ensure that we as humans, as translators, tell the AI very precisely what we care about, what the quality criteria are. And that is a large part of what we do: the mathematical work, but also the work of how we really teach the network what we as humans want from translation.
I
Interviewer4:17
What new trends are there? Clearly you are constantly expanding the circle of foreign languages, but are there other new trends you are working on?
J
Jaroslaw Kutylowski4:25
Yes, I think something I am very happy about is that it is now actually possible to achieve the same quality of translation that we know from written text, where we have invested a lot in recent years, also in the spoken domain. So we launched DeepL Voice, our new product last year, which allows us to translate meetings in a primarily business environment, but also in areas where customers come into a store and don't speak the language, where the entire conversation can be conducted via the mobile app. And that enables entirely new interactions, and it is a huge technological advancement that we are seeing.
I
Interviewer5:15
When we hear that, wouldn't the logical consequence be that in the future we will no longer speak foreign languages, or that we will no longer need translators? Or let me ask differently: Do you still have translators among your friends?
J
Jaroslaw Kutylowski5:34
Translators are a large part of our user base, and they were actually the first to work with DeepL. It brings them a lot; they can focus on the additional inputs they can provide to make the translation that comes from DeepL a little bit better. When it comes to the future, I think we will all still want to speak foreign languages. I can't imagine, if I am married to my partner, communicating through a phone for the whole life. There are still very, very many reasons why it makes a lot of sense to learn languages.
I
Interviewer6:22
So, tips for young people on career choice? They can't avoid AI.
J
Jaroslaw Kutylowski6:25
So even if AI also creates jobs, I think AI will be very, very important for everyone. So I would recommend everyone to actually invest in it: how to use the tools, how to understand them better, perhaps how to get into the field of AI development. I think Germany and Europe really need strong tech companies in the AI area. We have too few of them. Therefore, I would recommend everyone to actually look into this area.
I
Interviewer6:58
A big topic of discussion right now is fact-checking in artificial intelligence. Does that also affect DeepL? For example, does DeepL translate racist texts, or how do you handle that?
J
Jaroslaw Kutylowski7:13
With translations, we make sure that we translate exactly what the author wrote or what is necessary. There is not much room for moderation there, and I think there shouldn't be, because it's really not about creating new texts or spreading them. It's really about making them accessible in another language. And from my perspective, all texts and all types of texts really have their place. We should enable that for our customers.
I
Interviewer7:45
In the introduction, DeepL is still referred to as a startup. Can you still call it a startup given its size?
J
Jaroslaw Kutylowski7:56
I would still say startup, or I still say it very often. Of course, it only fits about 80% or 60% depending. But I think as a company in the tech environment, you always have to be able to change, you have to be able to keep building new things. So a lot of that startup image still remains.
I
Interviewer8:17
Finally, we'll do a little rapid-fire round. Are you ready for that?
J
Jaroslaw Kutylowski8:21
Yes, gladly.
I
Interviewer8:25
Let's see: AI, chance or risk?
J
Jaroslaw Kutylowski8:29
Definitely.
I
Interviewer8:29
Elon Musk makes you a purchase offer, wants to buy DeepL. Would you sell?
J
Jaroslaw Kutylowski8:34
Oh, I don't know, I don't answer that.
I
Interviewer8:38
Donald Trump applies to you. Would you hire him?
J
Jaroslaw Kutylowski8:45
No.
I
Interviewer8:45
Bitcoin: currency replacement or gambling?
J
Jaroslaw Kutylowski8:50
Gambling.
I
Interviewer8:50
When will AI surpass humans?
J
Jaroslaw Kutylowski8:55
Someday, maybe already. It's also not free.
I
Interviewer8:58
Will AI replace humans?
J
Jaroslaw Kutylowski9:03
No.
I
Interviewer9:03
In 100 years, will the chancellor be an AI?
J
Jaroslaw Kutylowski9:11
Possible.
I
Interviewer9:11
Sunday is election. Not voting is not a good choice for you. Jaroslaw Kutylowski from DeepL, the CEO and founder, was our guest, connected from Cologne. Thank you for the interview.