The Language Data for AI Event of the Year is back in Silicon Valley

It’s been almost three years since we’ve regularly come together all around the world to discuss the state of the industry. In our 15 years of events history at TAUS, we’ve discussed many topics. From localization workflows, pricing models, adopting quality evaluation metrics in our translation workflows, quality standards, and the many different translation technologies that help translators become more efficient, productive and creative. We brainstormed on solutions for the various challenges we came across, and in turn, TAUS started to create our own suite of products and services to address some of these challenges. The world has changed over the past years, and TAUS has changed as well. We’re so pleased to get together again and present the new TAUS, along with the exciting new subsector Language Data for AI, to you at our upcoming Massively Multilingual Conference & Expo in San Jose, CA on 11, 12 and 13 October 2022.

Over the years, together with our speakers, exhibitors and attendees, we’ve worked hard to advance the industry and push for the adoption of machine translation. Now that MT is here to stay, it’s time to talk about how we can do it better. At the TAUS conference, you’ll see that we’ve invited many gurus, experts and thought leaders from the machine translation, machine learning and AI departments to come and share their insights. I’m excited to give a sneak peek of the program in this blog post.

The words “massively multilingual” are not unknown to you if you’re operating in the translation and AI space. In order to go massively multilingual (in other words catering to every last human being on the planet in their own native language), you need two things: algorithms and data. In our opening session of the conference, we’ll be talking exactly about this. Which should be our core focus? The brute force of the massive models or the finesse of good quality data work to tune these models. And on the subject of data: we’ve also invited a variety of engineers to share the latest and greatest in the world of natural language processing in the translation field.

While we’re working on advancing the models and the techniques to put together high-quality and domain-specific datasets, we cannot do without evaluation. How else do you measure advancement? In another panel conversation, we’ll discuss the evaluation techniques for machine translation, whether we’re close to human parity (if we can reach it at all) and ask ourselves the question: is there a limit to MT?

We cannot just talk about science and theory though. It’s important to also hear from the businesses that have MT at the center of their workflow. Most players look at MT as a plug-in or an add-on to their existing translation processes. That’s not how we see the future of translation at TAUS. Various companies will be sharing insights on this explosion of new use cases for multilingual communications that MT has brought about. And of course, we’ll also talk about the trials and tribulations of transforming from a traditional pipeline to a more and more AI-enabled translation workflow. 

In another panel conversation, we’ll be discussing the type of language data that is currently most sought after: speech data. We’ll be addressing the sourcing of speech data, elimination of data bias and much more.

Massively multilingual models, high-quality data work, AI-enabled translation and other forms of language data. These are all elements needed in order to be world-ready. There are almost 8 billion people on the planet speaking more than 7,000 languages. With the multitude of content that’s being produced daily, how do we reach all these people in their native language? All of the topics on the program for the TAUS conference are working towards this goal. But in our World-Readiness Contest in particular, this question takes center stage. In just six minutes, companies are sharing their strategies and ideas in order to achieve world-readiness. And the audience is in charge of picking the best idea!

Naturally, there will also be plenty of time for networking. Catch up with old friends and colleagues and meet new ones at the opening reception on 11 October at the Doubletree by Hilton San Jose. Or join us for an exclusive networking dinner at the idyllic Regale Winery & Vineyards on 12 October, where we’ll enjoy fantastic views, food and live music by the TAUS Haus Band.

If you haven’t registered yet, make sure to save your seat and join us in San Jose, CA in October to participate in this exciting program where we put the spotlight on language data and artificial intelligence.