Translated Presents the 2025 Imminent Annual Report

Today we’ve announced a two-year research initiative led by our research center, Imminent. The initiative begins with the release of the 2025 Annual Report, Evolution in Words: Beyond AI, which explores the intersection of language, artificial intelligence, and cultural diversity in a rapidly evolving technological landscape. In 2026, the research will continue with a new report examining how the shift from large language models (LLMs) to multimodal AI systems may reshape this intersection, along with their social, cultural, and technological implications.

The 2025 Imminent Research Report features insights from world-renowned thinkers, including Viktor Mayer-Schönberger (Professor of Internet Governance, Oxford University), Dennis Yi Tenen (associate professor of English and Comparative Literature at Columbia University), Gry Hasselbalch (author and co-founder of DataEthics.eu), and Jannis Kallinikos (professor of Information Systemns at Luiss University and professor emeritus at London School of Economics), along with original studies by Patrizia Boglione (Translated’s VP of Brand and Creative) and Kirti Vashee (Translated’s Tech Evangelist), and research contributions from a network of language and technology professionals working in this field. The report raises a pivotal question: How will the development of AI redefine the evolutionary path of human language and humanity itself?

At the heart of the report is a compelling argument: language is not merely a tool for communication, but a fundamental driver of human culture, identity, and power. As AI technologies increasingly produce, translate, and mediate our words, the implications stretch far beyond efficiency or convenience; they strike at the core of human expression.

Combining deep research, provocative essays, and practical frameworks, Imminent has not only published a report, but also begun to explore a roadmap for an inclusive, multilingual AI future. This will serve as the basis for examining the significant shift in AI’s socio-technical impact. Amid growing global attention on AI, a wave of new reports underscores the fundamental limitations of LLMs. These systems continue to struggle with hallucinations, exhibit a marked inability to handle complex reasoning, and show systemic bias toward highly documented languages. Yann LeCun, Chief AI Scientist at Meta and Turing Award winner (often referred to as the Nobel Prize of computing), has publicly warned that the current trajectory of LLM development, driven by enormous and unsustainable investments in computing power, is not a viable long-term strategy. According to LeCun, the next evolutionary leap in artificial intelligence will require a radical scientific breakthrough, not merely a scaled-up extension of today’s language models.

In the future, these topics will continue to be investigated as part of a new phase of research on artificial intelligence. This phase aims to move beyond large language models toward multimodal models capable of managing diverse data systems, situating AI in space, enabling it to learn from new data arising from environmental feedback, and allowing it to act through robotic systems or other devices. Imminent’s research in the coming year will focus on the consequences of this approach, supported by collaborations between Translated and leading European research centers at the forefront of the field, as part of the DVPS project funded by the European Commission. Journalists and media professionals are invited to explore the full report and engage with its authors and contributors. Interviews, expert commentary, and press materials are available upon request.