OUR THOUGHTS

Day 1 of AI ThoughtCon brought together a global audience, with over 1,000 registrants attending via Zoom and YouTube Live. The event, organized by the AI Localization Think Tank, focused on exploring the future of the industry, combining a keynote, a panel discussion, and two focused talks in a 2.5-hour program. The conference opened with remarks from Belén Agulló García, who reflected on the rapid growth of the Think Tank since its creation in November 2024. She highlighted

In this article, Marina defines agentic AI solutions by outlining varying levels of autonomy and complexity, illustrated with concrete localization use cases.

In this piece, Marta explains how to correlate human and automated translation quality metrics using Pearson, Spearman, or Kendall methods.

As part of our Q&A series, this article discusses how translation emerges as a skill from the broad training of LLMs and explores how to further improve their translation capabilities.

This article, Marta walks us through translation quality metrics for LLM (and NMT) evaluation, highlighting the role of human assessment and data-driven approaches in building meaningful, case-specific evaluation strategies.

The evolution of GenAI follows stages similar to those of human cognitive maturation. Where will the development of LLMs take us next?

This article explores the definitions of "quality" as a general outcome and "Quality" as a structured practice in today’s Localization ecosystem, emphasizing the need for intentional frameworks to evaluate translation effectively.