Gabriel Karandyšovský previews “Global Ambitions: Revolution in Motion,” a five-article sampler. He digs into making AI actually ship through better infrastructure, why incremental wins outlast hype, localization’s “plumber” reality, treating translation as a product feature, and the skills next-gen leaders need across product, UX, growth, and data.

On July 3, 2025, at Translated’s unique “PI Campus” in Rome—eight luxury villas retrofitted into pools‑and‑massages‑equipped offices—The Translated team kicked off Project DVPS (Diversibus Vis Plurima Solvo). Backed by a €29 million EU grant, DVPES unites 70+ top AI researchers from EPFL, Oxford, ETH Zurich, FBK Trento and more to build the next generation of multimodal foundation models.

By fusing language with handwriting, audio, video, sensor data and beyond, and by moving from sequential “chain‑of‑thought” to real‑time “learn‑by‑doing” reasoning, the project aims to shatter today’s context, reasoning and continual‑learning limitations in LLMs—and push toward truly grounded, human‑level AI.

Sébastien Bratières, Director of AI at Translated, outlines “DVPS”, a four-year EU project to build multimodal foundation models with partners including EPFL, Oxford, and ETH Zurich. We track why language tech must move beyond text to “physical AI,” integrating speech, video, handwriting, environment (satellite imagery), and cardiology data so models learn meaning from real-world context.

Amir Kamran and Amir Soleimani from Taus unpack Quality Estimation (QE) as the missing link between machine translation at scale and human-quality outcomes. They frame QE as an automated “second opinion” that flags what can ship, what needs a quick AI touch-up, and what should go to a linguist—shifting human effort to the tricky, high-risk bits.

In this conversation, we explore where this matters most (high-volume content, real-time chat, regulated use cases), how QE reshapes workflows after MT, and why language access still demands human oversight in sensitive domains. They close with what’s next: lighter, task-focused models, more agent-style automation, and a broader definition of quality that favors fitness-for-purpose over perfection.

Eddie Arrieta sits down with Craig Stewart, Director of AI Research at Phrase and one of the architects of the Comet machine-translation metric, to unpack the evolving definition of “quality” in global content.

They explore why linguistic accuracy alone no longer suffices, how downstream signals like engagement and conversion should influence localization strategies, and what it takes to turn AI-driven volume into truly resonant experiences. Craig also demystifies the emerging world of agentic AI, offers advice for today’s linguists, and shares how teams can bridge the gap between language technology and real-world business impact.

Interview by Cameron Rasmusson

In this interview, Giulia Greco discusses her decision to leave corporate localization management to become a consultant — a path that comes with uncertainty, but also offers freedom, agency, and new opportunities for impact.

By Carol Velandia

Executive Order 14224, signed in March 2025, declared English as the official language of United States federal agencies. The author calls for language service providers to publicly oppose the order and to become fierce advocates for language access policies and standards.

By Lucía Gutiérrez Franco and Gabriela Kouahla

The authors explain why mastering glossary creation and management — as well as leveraging the right translation tools — helps language professionals ensure accurate and consistent translations in an increasingly automated world.

By Yasir Tehsin

The author argues that AI voices, in no matter the language, are sorry substitutes for humans when it comes to narrating audiobooks — asserting that they pose a threat to the future of literature.

By Jorge Russo dos Santos and Agustín Da Fieno Delucchi

Unlike traditional localization, which adapts content from a source language, a new methodology developed by Trilogica Global simultaneously creates content in multiple languages and cultural contexts. The authors argue that this approach represents not just an incremental improvement in localization technology, but a fundamental reimagining of global communication.