Slator - Language Industry News

  A Wake-Up Call for AI Translation Accuracy Artificial Intelligence (AI) has revolutionized translation in recent years, but Google’s latest warning has raised eyebrows across the language technology industry. According to Google, many AI translation benchmarks may be significantly overestimating performance , creating a false sense of accuracy. This revelation is a wake-up call for businesses, translators, and researchers who rely heavily on benchmark scores to evaluate translation tools. But what exactly is the problem, and how should the industry respond? Let’s break it down. The Role of Translation Benchmarks in AI Development Translation benchmarks are standardized tests used to measure the accuracy and fluency of AI-powered translation systems. They guide: Businesses in selecting the right tools. Researchers in tracking AI progress. Developers in refining models. However, when these benchmarks are flawed or inflated , they can mislead decision-makers , r...

eBay Launches New In-House Large Language Model for E-commerce with Translation Capabilities In a June 17, 2024  paper ,  eBay   introduced its series of  l arge language models  (LLMs), tailored specifically for the  e-commerce  sector. eBay’s New In-House Large Language Model for E-commerce Can Also Translate These models, named LiLiuM 1B, 7B, and 13B , were developed in-house to meet eBay’s specific needs across various applications, including translation, in the e-commerce domain, providing full control over licenses, data, vocabulary, and architecture. The authors said that “these models are meant to eliminate dependency on third-party  LLM s within eBay.” eBay explained that using foundation models like the LLaMA-2 models, which can be accessed and adjusted for specific purposes, poses risks related to licensing, data security, and future-proofing. They noted that these models are generally trained on English-centric data and are quite gene...