Highlights of LocWorldWide45—our recap is all about the good, the great, and the downright awesome.
There’s nothing better than an event packed with interesting people, a bunch of mind-blowing talks, and plenty of opportunities for networking and business. Luckily for the translation industry, year after year, LocWorld strikes the perfect balance of all three.
This year was no different: On October 19-21, 2021, attendees had the opportunity to meet amazing people, listen to inspiring reflections on how our industry is changing, and catch up with old friends.
All in all, LocWorldWide45 was a great conference, so let’s have a look at some of the highlights—this recap is all about the good, the great, and the downright awesome.
Mergers and acquisitions: It’s all about the people and the culture
David Čaněk, Memsource’s Founder & CEO, took to the virtual stage on Wednesday, October 20, together with Britta Aagaard—Chief Business Officer at Semantix—to address how mergers and acquisitions (M&A) bring about change and the meshing of two cultures to create a cohesive team with the same goal in mind.
David spoke from personal experience: Since founding Memsource in 2010, the company has gone through both being acquired and acquiring themselves (Phrase—in 2021). Britta and David immediately agreed that, while it may sound like a cliché, people really are at the heart of any merger or acquisition.
So, what can CEOs do once the deal is sealed and the ink dries? They shared some insights on how to make integration a success on the basis of shared values and a joint vision forward.
They analyzed how pre-merger due diligence is often significantly detached from the reality of a business—you are merely looking at a screenshot of the business at that moment. The focus is normally on hard data (finance and legal) and not enough on the human factor, so it’s difficult to anticipate what the reality of integrating with another company will be like.
Something that resonated with many in the audience was Britta’s answer to David’s question, “How do you know that integration has been successful?” Britta pointed out how we shouldn’t try to reach the point where “we are all the same.” We should embrace diversity.
We can, however, reach the point where “we are all heading in the same direction as one,” and that’s when the staff talks about themselves as “us,” without even having to mention the company name. Britta called it “the ‘we’ feeling,” and everyone loved it!
The state of machine translation in Asia differs from Europe enormously
The panel discussion “Machine Translation for Asian Languages: Challenges, Solutions, and Innovation” turned out to be particularly thought-provoking. It was moderated by Aleš Tamchyna, AI Research and Development Manager at Memsource.
Applying MT technology to Asian languages is a different ballgame from what we normally experience in Europe. The panel, who brought together three leading technology and machine translation providers from around the world—Rozetta, Tencent, and Pangeanic—shared some surprising insights:
- How slow the implementation of MT has been in the Japanese market as a result of cultural norms: Japanese corporate culture tends to value traditional methods over innovative ones and demand 100% accurate output. However, Japanese businesses have realized that cost and speed play a critical role in global competition, with MT providing a competitive advantage.
- How rapid expansion from Asia into Europe can benefit from machine translation for internal communications: Tudor Toma from Tencent described how he uses MT for anything from daily exchanges with colleagues in Asia to watching live events. He mentioned that he simply wouldn’t be able to do his work without MT.
- How dialectal and regional diversity in Asia is a challenge for MT: We tend to conceive Asia as one homogenous block, but that couldn’t be further from the truth. Chinese, for example, encompasses over 600 dialects, and Japanese has three alphabets.
- The fact that MT can only get better if you feed it data: We need millions of examples to train a neural engine—because Asia is in the early stages of MT adoption (despite Chinese and Japanese providing a great amount of data already), the more we use machine translation in Asia, the better it will get.
- The fact that some language combinations and directions are easier than others, but that translating between two very close languages like Malaysian and Indonesian poses equally big challenges: This is because users expect much better quality between two languages that are very similar.
- How machine translation and translation management systems like Memsource will become more tightly integrated in the years to come as a result of the demand for custom MT models: Bilingual corpora are needed to create these, and the translation memories inside TMS tools is precisely where customers store this data.
AI helps us do things—but also learn things
On Thursday, October 21, David Čaněk shared a panel with Jay Marciano, and Peng Wang. Nora Díaz and Jost Zetzsche moderated it. After a quick run-through of what the incredible Innovation in Translation Summit left them with, the panelists dove into a discussion about AI and how it’s changing the translation industry.
They defined AI as “technology that works on tasks that we used to assume required human intelligence.” They also established that AI is an accelerator of productivity, a faster and better way of completing repetitive tasks.
Most importantly, besides helping us do things, AI helps us learn things. Take the example of Memsource’s machine translation quality estimation, which tells you whether the MT output might need post-editing or is most likely excellent as is.
This is where the panelists veered off into a discussion of whether AI actually does take over jobs: They concluded that what we’re really seeing is job transformation or displacement, with some tasks being replaced and others requiring new skills.
What was particularly interesting about the discussion was the section on the different expectations and levels of acceptance that translators have of AI-powered machine translation. At one end of the spectrum, there are people who ignore the technology completely and stick to their comfort zone; at the other, those translators who choose to work on machine translation post-editing (MTPE). And in between those two extremes, there’s a wide range of tasks that AI can manage that we haven’t truly explored yet, but which we should start embracing.
The bottom line is that we should take a holistic view of translation and value the relationship between AI and humans.
If you’re not strategic, you’re not adding value to your business
The panel with Francesca Di Marco, Christian Redmann, and Lyena Solomon, moderated by Wayne Bourland, was a delightful mix of perspectives on the strategic side of localization.
The three panelists—who lead localization and internationalization efforts at Pinterest, Siemens, and ServiceNow—joined with nine additional industry professionals to create a globalization strategy playbook with their experience of solving strategic problems in their companies. Offering a strategic framework around decision-making, the book pushes leaders of global businesses towards a continuous improvement model and helps them connect real-life situations to a strategic thinking process.
Christian pointed out how localization departments are often viewed as internal service providers that tasks are insourced or outsourced to. However, they have a lot more expertise and value to offer and can become strategic partners to the business.
In a nutshell, localization departments must be aware of the company’s strategic objectives, and companies need to include localization strategy in their business strategy if they want to be truly global.
What LocWorldWide45 has taught us
A common thread that ran through most LocWorldWide45 talks was how the human factor is key to the localization industry.
Whether you’re a translator, software developer, localization manager, UX designer, or even a business owner, at the end of the day, we work with people and for people as enablers of communication.
As long as we keep that in mind, we’re more likely to develop better products, processes, and services that meet the expectations of our users, colleagues, and clients.