10 Ways Large Reasoning Models Fall Short in AI Translation
Large reasoning models perform well in open-domain AI translation, but struggle with domain-specific terminology, instruction...
AI translation
From enterprise translation AI to automated localization.
Grok Gets Promoted on X to Chief AI Translator
Google has been fired and Grok is now the exclusive AI translator on X. The...
UK Society of Authors Launches Campaign Addressing GenAI in Translation
A 2024 member survey found translators have lost work and wages to AI; in response,...
From AI Translation to AI Localization With Unbabel’s Open-Weight Tower+
Unbabel releases TOWER+, a suite of open-weight multilingual LLMs designed to deliver top-tier AI translation...
New Method Helps Large Language Models Handle Polysemy in AI Translation
Researchers from the Harbin Institute of Technology propose a method that helps large language models...
Microsoft Introduces Phi-Omni-ST for AI Live Speech Translation
Microsoft demonstrates how to extend speech-aware multimodal large language models from speech-to-text to AI live...
AI Live Translation Stars as Key Feature at Apple 2025 Developers Conference
Unlike many of the announcements made, Apple Intelligence-enabled Live Translation was well received and dominated...
OpenAI Doubles Down on AI Live Speech Translation in ChatGPT
A year after the “Italian” live speech translation demo with former CTO Mira Murati, OpenAI...
Images Improve Cultural Accuracy of AI Translations, Research Finds
Researchers from MBZUAI suggest that images help multimodal large language models generate more culturally accurate...
Huawei Uses Large Language Models for AI Translation Only When It Counts
Huawei researchers propose a new method that uses LLMs only when they are likely to...
Alibaba Launches TransBench for ‘Industrial-Scale’ AI Translation Evaluation
Alibaba introduces TransBench, a benchmark aimed at testing how well AI translation systems perform in...
Large Language Models Struggle to Evaluate Long AI Translations, Amazon Finds
An Amazon study reveals a limitation in using large language models to evaluate AI translation...
IIT Bombay Explores Accent-Aware Speech Translation
IIT Bombay researchers propose a new approach to speech-to-speech translation that not only translates speech...
QA-Based Evaluation Takes on Literary AI Translation
A new LLM-based metric builds on the promise of question-answering to evaluate literary translations, capturing...