Introduction: In a recent fireside chat, Robin Ayoub, host of the Localization Fireside Chat, sat down with Konstantin Savenkov, CEO of Intento, to discuss the state of machine translation for 2024. The conversation revolved around the findings of a report published by Intento and e2f, shedding light on the current landscape of machine translation and its future trends. Here are some key highlights from their insightful discussion.
The Evolution of Intento: Konstantin shared his localization story, revealing that Intento was founded with the aim of making AI more accessible for enterprise companies. Initially focusing on the translation industry, they discovered the wide price range and set out to explore the correlation between price and quality. Their first report in 2017 debunked the assumption that higher-priced systems equated to better quality. Today, Intento has evolved into a platform that enables enterprises to configure automatic translation workflows and maintain consistent brand voice across various software systems.
The Proliferation of Machine Translation Solutions: The conversation delved into the sheer number of machine translation solutions available in the market. Konstantin highlighted that major cloud providers often offer machine translation as a loss leader service to attract customers. Additionally, there are dedicated machine translation companies and large language service providers that develop their own systems. The emergence of large language models (LLMs) has further blurred the lines between machine translation and other language-related tasks.
Performance Comparison: Machine Translation vs. Large Language Models, The report showcased the performance of machine translation systems and LLMs. While traditional machine translation systems still prevail, LLMs have caught up in terms of quality. However, LLMs are significantly slower but more cost-effective. Konstantin emphasized that once LLMs become faster, real-time translation and other applications will witness a dramatic shift.
Real-Time Translation and Future Trends: The conversation explored the growing demand for real-time translation in various applications, such as live chats and on-the-fly website translation. Konstantin highlighted the need for strict response time limitations and the challenge of maintaining quality without the option for post-editing. While current technology has limitations, he predicted that once LLMs become faster, real-time communication will undergo a significant transformation.
Future Innovation and Cognitive Computing: The discussion touched upon future trends and innovations in the machine translation industry. Konstantin emphasized the need to remove legacy configurations and systems to allow technology to shape the landscape. He also mentioned the potential of generative AI and the role of cognitive computing, which is still under the radar but holds promise for further advancements.
Conclusion: The fireside chat between Robin Ayoub and Konstantin Savenkov shed light on the state of machine translation for 2024. The report published by Intento and e2f provided valuable insights into the performance of machine translation systems and the emergence of large language models. The conversation also addressed security considerations, real-time translation challenges, and future trends in the industry. As technology continues to evolve, the localization industry can expect further innovation and advancements in the field of machine translation.
Until next time this is Robin Ayoub signing off
Connect with us:
Visit our website: http://www.L10NFiresidechat.com
Join our WhatsApp community: https://chat.whatsapp.com/BYF3cMEVl8U7gKFFqttyMe
Connect with us on LinkedIn: https://www.linkedin.com/company/localization-fireside-chat/
Robin Ayoub: https://www.linkedin.com/in/robinayoub/
Subscribe for more episodes:www.youtube.com/@L10NFiresideChat Thank you for tuning in!
Stay connected for more episodes and insightful discussions on the Localization Fireside Chat!
Robin Ayoub 
Leave a comment