Wikimedia Product/Perspectives/Augmentation/Machine Translation

Machine Translation

Summary

Machine translation like any other machine augmentation, could accelerate our vision for the sum of all knowledge to be available to everyone in the world. Specifically, it means work done in one language can be made available in others. This could speed up the transmission of information dramatically, making more knowledge available and freeing contributors to work on original research. At the same time, machine translation from a third party, represents a threat to our model, potentially promoting English dominance, siphoning traffic from our wikis, offering a poor experience, or discouraging global contribution. We cannot ignore this trend, but must adapt to it. No matter how fast humans create or translate content, we will not be able to create or, crucially, update content faster than a competitor who uses machine translation. We need to adopt a proactive, long-term approach to machine translation that serves our users, aligns with our values and supports our ecosystem. If we do this correctly, machine translation could represent the ingredient that allows us to realize our vision of global knowledge sharing.

White Paper

DRAFT

Resources

B. Turovsky, 2081 Found in translation: More accurate, fluent sentences in Google Translate https://www.blog.google/products/translate/found-translation-more-accurate-fluent-sentences-google-translate/

T. Wolverton, 2018 Google CEO Sundar Pichai revealed a jaw-dropping fact about its translation app that shows how much money is still sitting on the table https://www.businessinsider.com/sundar-pichai-google-translate-143-billion-words-daily-2018-7

M. Ranzato, G. Lample, M. Ott, 2018 Unsupervised machine translation: A novel approach to provide fast, accurate translations for more languages https://code.fb.com/ai-research/unsupervised-machine-translation-a-novel-approach-to-provide-fast-accurate-translations-for-more-languages/