Steve Richardson (past president of AMTA – Association for Machine Translation in the Americas) and Paco Guzmán (Research Scientist Manager at Meta AI) both gave supporting statements about the future of translation. They were unable to attend the NextGen conference as they were attending the TAUS conference on the same weekend. Their statements and bios are below.
Statement from Steve Richardson: “The advent of ChatGPT and other large language models is yet another inflection point in the evolution of technologies relevant to the translation industry. The previous point was the transition from statistical machine translation (SMT) to neural machine translation (NMT), which began around 7 years ago. With the transition to NMT, many claims were made, as they had been over many prior decades, that machine translation would obviate the need for human translators. But the reality is that the number of opportunities for human translators has continued to rise, driven by the ever-expanding world economy. To be clear, the volume of translations produced by NMT systems has also continued to increase, surpassing many hundreds of billions of words every day. At the same time, the need for polished, well-crafted human translations (including human post-edited machine translations) has increased as a smaller but important percentage of the total worldwide translation demand, and this need will undoubtedly continue to grow into the foreseeable future. It will be essential, however, for translators to continue to adapt and learn to use the latest technologies to their benefit, thus ensuring that their relevance and levels of productivity keep pace with the accelerating translation marketplace.”
Steve Richardson’s Bio: Steve Richardson is an associate professor of Computer Science at Brigham Young University. He recently completed serving for four years as the president of the Association for Machine Translation in the Americas (AMTA) and continues involvement as counselor for AMTA. Steve has been involved in MT R&D for almost five decades. His prior experiences include: manager of Machine Translation and Translation Systems at the Church of Jesus Christ of Latter-day Saints, partner researcher and manager of the Machine Translation Group at Microsoft Research, senior programmer at the IBM Watson Research Center and ASD Bethesda lab working on NLP and MT, and researcher at BYU’s Translation Sciences Institute. He has BS and MA degrees in Computer Science and Linguistics from BYU and a PhD in Computer Science from the City University of New York. He has served on numerous advisory boards and conference program committees. He is the author and/or editor of various books, conference proceedings, and research papers, as well as numerous patents in NLP and MT.
—
Conversation between Paco Guzmán and Alan Melby:
Paco: Sometimes, it’s not enough to simply translate at the lexical or even the semantic level. You need to explain and localize the knowledge in a way that goes beyond a simple one-to-one mapping. There are instances where you must provide a comprehensive explanation of what something is and its role in society. This ensures that the person receiving the message can truly understand it.
Alan: So in such use cases, who needs to be involved in the translation process besides a computer?
Paco: When you look at what we did with NLLB (No Language Left Behind), you’ll notice we had a strong presence of ethicists and linguists involved in the development of the project. It wasn’t just technologists and statisticians, because translation is a human endeavor, you know. You are fulfilling a human need for communication, and at the end of the day, humans will be the end users of translation. So, we need to take a more holistic approach to the development of AI systems rather than merely training and releasing models.
Alan: I love the way you put it. Thank you so much! Let’s end on that positive note.
Paco Guzmán’s Bio: Paco is Research Scientist Manager supporting translation teams in Meta AI (FAIR). He works in the field of machine translation with the aim to break language barriers. He joined Meta in 2016 and has co-led several initiatives (e.g. SeamlessM4T, NLLB , FLORES). His research has been published in top-tier NLP venues like ACL and EMNLP. He was the co-chair of Research at AMTA (2020-2022) and Ethics co-chair at EMNLP 2023. He has organized several research competitions focused on low-resource translation and data filtering. Paco obtained his PhD from the ITESM in Mexico, was a visiting scholar at the LTI-CMU from 2008-2009 and participated in DARPA’s GALE evaluation program. Paco was a postdoc and scientist at Qatar Computing Research Institute in Qatar in 2012-2016.