With conflict persisting to hinder education for countless Ukrainian high school and college pupils, many are resorting to online options, such as MIT OpenCourseWare, an initiative within MIT Open Learning that provides educational materials from over 2,500 MIT undergrad and graduate courses.
For Ukrainian high school senior Sofiia Lipkevych and her peers, MIT OpenCourseWare has offered significant chances to enroll in courses across essential subjects. However, while numerous Ukrainian students are studying English, many do not possess a sufficient grasp of the language to fully comprehend and utilize the often highly technical and intricate OpenCourseWare materials and content.
“In my school, I witnessed how language obstacles barred numerous Ukrainian students from accessing top-tier education,” states Lipkevych.
She managed to confront this issue as a member of the Ukrainian Leadership and Technology Academy (ULTA), launched by Ukrainian MIT students Dima Yanovsky and Andrii Zahorodnii. During the summer of 2024 at ULTA, Lipkevych collaborated on a browser extension that translated YouTube videos instantaneously. Given that MIT OpenCourseWare was a primary source of instructional materials for students involved in ULTA, she was motivated to translate OpenCourseWare lectures directly and to make this translation broadly accessible on the OpenCourseWare website and YouTube channel. She reached out to Professor Elizabeth Wood, the founding director of the MIT Ukraine Program, who connected her with MIT OpenCourseWare Director Curt Newton.
Though there had been some translations of MIT OpenCourseWare’s educational resources available since 2004, these initial translations were done manually by various global partners, lacking the advantages of the latest artificial intelligence tools, and over the years, the programs could not be sustained and were shut down.
“We were delighted to have this contact with ULTA,” remarks Newton. “We’ve been longing for a dynamic translation community, and we are excited to witness a ‘phase 2’ of translations developing.”
The ULTA team chose courses to translate based on the interests of Ukrainian students, concentrating on foundational subjects that serve as prerequisites for advanced studies — particularly those for which high-quality, Ukrainian-language resources are limited. Beginning with caption translations on videos of lectures, the team has translated the following courses to date: 18.06 (Linear Algebra), 2.003SC (Engineering Dynamics), 5.60 (Thermodynamics & Kinetics), 6.006 (Introduction to Algorithms), and 6.0001 (Introduction to Computer Science and Programming in Python). They also collaborated directly with Andy Eskenazi, a PhD student in the MIT Department of Aeronautics and Astronautics, to translate 16.002 (How to CAD Almost Anything – Siemens NX Edition).
The ULTA team devised various tools to help dismantle language barriers. For MIT OpenCourseWare’s PDF content accessible through the ULTA program, they developed a specialized application that employs optical character recognition to identify LaTeX within documents — such as problem sets and other materials — and then utilized a few large language models to translate them while ensuring technical precision. The team created a glossary of technical terms utilized in the courses alongside their corresponding Ukrainian translations to ensure the terminology was both accurate and consistent. Each translation also undergoes a human review to further guarantee precision and high quality.
For video content, the team initially established a browser extension capable of translating YouTube video captions in real-time. They ultimately partnered with ElevenLabs, implementing their sophisticated AI dubbing editor that maintains the original speaker’s tone, tempo, and emotional delivery. The lectures are translated in the ElevenLabs dubbing editor, and subsequently, the audio is uploaded to the MIT OpenCourseWare YouTube channel.
The team is in the process of finalizing the audio translation for class 9.13 (The Human Brain), instructed by MIT Professor Nancy Kanwisher, which Lipkevych notes they selected for its interdisciplinary aspect and interest to a diverse array of learners.
This Ukrainian translation initiative underscores the transformative capability of cutting-edge translation technologies, building upon a 2023 MIT OpenCourseWare experiment using the Google Aloud AI dubbing prototype on various courses, including MIT Professor Patrick Winston’s How to Speak. The advanced functionalities of the dubbing editor utilized in this project are paving the way for a significantly broader spectrum of language offerings throughout MIT OpenCourseWare materials.
“I anticipate that in a few years, we’ll reflect back and see that this was the moment when OpenCourseWare truly became usable for the entire world,” states Newton.
Community-driven language translations of MIT OpenCourseWare materials exemplify a high-impact instance of the strength of OpenCourseWare’s Creative Commons licensing, which permits everyone the ability to modify materials to meet their specific needs and redistribute those modifications globally.
While there currently isn’t a method for MIT OpenCourseWare platform users to quickly determine which videos are available in various languages, MIT OpenCourseWare is working towards integrating this capability into its website, along with increasing its offerings in additional languages.
“This project encompasses more than mere translation,” remarks Lipkevych. “We’re empowering thousands of Ukrainians to acquire skills that will be vital for the country’s eventual reconstruction. We’re also optimistic that this model of collaboration can be extended to other languages and institutions, creating a template for making high-quality education accessible on a global scale.”