Microtalks
Track: Microtalk | M2 | Intermediate | MICRO TALK |
Wednesday, June 4, 2025, 9:45am – 10:15am
Held in: Live 3
Presenters:
Marcus Casal - Lionbridge
Peter Reynolds - memoQ
Giulia Silvestrini - Acolad
Kirti Vashee - Translated
Host: Daniel Goldschmidt
In this session, we will offer several quick and dynamic presentations covering topics and ideas too interesting to ignore!
Live Communication: Where Human Meets AI. Use cases on how interpretation, technology, and AI enhance effective and inclusive interactions. Presented by Giulia Silvestrini, Acolad
The talk will delve into various use cases showcasing the convergence of human interpretation, technology, and AI to enhance internal and external communication. These include scenarios like technical training, product presentations, and customer interactions in call center settings. The focus will be on how advancements in interpretation technologies are facilitating live multilingual communication in onsite, remote, and hybrid environments. The session will also explore a fit-for-purpose approach to meet the diverse needs of stakeholders, highlighting how these tools and solutions help overcome language barriers across different settings.
Key Takeaways:
Live interpretation and language services are increasingly accessible and affordable, expanding use cases and enhancing the ROI of effective, inclusive multilingual communication in real-time settings.
Evaluating the Impact of GPT-4 Fine Tuning on Translation Quality: A Competitive Analysis. Presented by Marcus Casal, Lionbridge
Across industries, AI-curious enterprises are discovering how Large Language Model (LLM) fine-tuning can harness the power of generative AI. In this presentation, learn how a recent study leveraged Fine Tuning for GPT-4 via Azure OpenAI Service to show that fine-tuned GPT models can outperform baseline GPT and neural machine translation models, significantly improving translation accuracy for closer alignment with customer-specific terminology, style, and corporate voice. The results highlight the potential for generative AI to revolutionize quality and personalization, especially when deployed by skilled language professionals who understand how to apply AI to meet dynamic customer needs.
Key Takeaways:
The benefits of GPT-4 Fine Tuning for customizing translation services to better meet customer demands regarding terminology, style, tone, and corporate voice; the difference between automatic methods and human evaluations to assess the translation quality across languages; results show the translation quality achieved with the fine-tuned models far surpassed that of the baseline models and even outperformed the “traditional” Neural MT models.
The Coming Impact of LLM MT. Presented by Kirti Vashee, Translated
LLM MT was acknowledged as the next wave of MT technology at WMT24, thus we should expect a shift to this new approach from 2025 onward. However, there are still several obstacles to widespread adoption. What are the issues that would slow down this transition? How is this transition different from the SMT to NMT shift? Are the skills needed to optimally use LLM MT the same as NMT? What are some of the new translation adjacent capabilities that LLM MT will also allow? This session will cover these questions and other related themes.
Key Takeaways:
Key skills needed to optimize LLM MT; key resources needed to build expertise; the likely timelines for widespread adoption
The Greatest Business Thinker You’ve Never Heard Of! Presented by Peter Reynolds, memoQ
Ideas around the use of psychology and group psychology in business, cross functional teams, conflict solutions and getting things done can be traced back to Mary Parker Follett. One of the most relevant and important business thinkers is one of the most ignored. This Micro Talk will introduce her work and explain why ideas created a hundred years ago are still relevant today.
Key Takeaways:
The core takeaway will be ideas on motivation and how to excel at managing people.
