Michael Campbell, the founder of Glossika, was featured on RTI's podcast "The Download," hosted by Staś Butler (Part 1 | Part 2). Michael discussed the founding of Glossika, what he hopes to accomplish with the product, and how Glossika contributes to different areas of the world.
The Founding of Glossika
Glossika started off as a method that Michael himself used to study languages. When people started asking him for language learning materials, he decided to turn it into a business. Glossika is different from traditional classroom language learning methods in that it utilizes thousands of sentences with readily-available translations and recordings to help readers naturally acquire grammatical structures, rather than memorizing them by rote.
Communication Across Languages
Glossika's main focus isn't the grammar of one particular language. Instead, Glossika analyzes the underlying syntactic structure across all languages. Michael hopes that this system based on first-order logic will eventually be able to line up the grammatical elements of every language.
As time went on and the business expanded, Glossika launched a new site — Glossika Viva. Michael hopes that this site will give life to languages, and become a community where people that speak minority languages (or any language for that matter) can contribute sentences and recordings of their language. As a result, the community will be at once preserving languages and also creating content for fellow learners.
In the larger scope of things, Glossika aims not only to teach languages but also to help keep endangered and minority languages alive. Viva is just one way Glossika is doing so.
In our globalizing world, nations are finding ways to better connect with global trends and increase their competitiveness. Taiwan (where Glossika is located) is one such country.
In 2018, the Taiwanese government launched the 2030 bilingual nation plan, which aims to improve the English ability of its citizens. The plan includes implementing bilingual documentation and websites for government ministries and creating a more immersive learning environment in public schools. If you are interested in more details about the project, you can learn more about it here.
When asked about the bilingual nation project, Michael said that one key to its success would be to teach English as a tool for communication, with less focus placed on grammar. Right now, schools place significant emphasis on teaching students the vocabulary necessary to pass tests. Michael thinks it is more practical to encourage students to use English to talk about things in their everyday lives. He also thinks it would be effective and relatively simple to begin having schools use English to teach other subjects, like math or science.
Effects of Machine Translation
With the rise of advanced machine translation tools such as Google Translate, many people predict that we won't need to learn languages in the future. If this is the case, it will directly affect companies like Glossika whose business models consist of teaching people languages. However, Michael isn't worried about these translation tools putting Glossika out of business:
"People are always gonna want to learn languages, because first of all, it's identity. It's who you are... Secondly,... being able to speak a language from your own vocal chords is a very satisfying feeling."
Michael also believes that artificial intelligence still has a long way to go in terms of understanding human intent in language. From his personal experience of working in the translation field, even human translators can have trouble agreeing on a translation for a sentence. "You can get ten people in a room who all disagree on a translation," he said.
Additionally, while Google only currently supports a couple hundred languages, there are a total of 7,000 languages in the world plus a few hundred sign languages on top of that. Machine translation tools still have a very long way to go.
The Future of Glossika
These tools aren't all bad, however. There would be situations where these tools might be needed for humanitarian needs. Maybe a customer on a plane only speaks a certain dialect, for example. In such situations, these AI tools would prove to be very helpful.
Michael hopes that Glossika can contribute to these tools and also help to refine the quality of machine translations as well. As Glossika owns a significant amount of language data that cannot be purchased elsewhere, Michael hopes that these properly annotated datasets can be used for machine learning purposes and potentially lead to additional revenue streams for the company.