Intec Annual Report

10 EMERGING TECHNOLOGIES

TO BOOST SPAIN

Featured


  • Javier Ventura-Travesset (ESA): "Europe has many strengths, but action is essential".

    Javier Ventura-Travesset (ESA): "Europe has many strengths, but it is essential to move on to the next generation.

    Orchestrating decisions to achieve the ambitious goals of the Moonlight project, in an environment as complex as that of the European Space Agency, is one of the tasks of Javier Ventura-Travesset, with whom we talk about how to take action and how to drive innovation through demand.

    Further information

  • New Technologies Report (Intec) 2024

    New Technologies Report (Intec) 2024

    The Chair in Science and Society of the Rafael del Pino Foundation prepares the report Ten technologies to boost Spain. Under the supervision of Professor Javier García Martínez, the Chair's team drafts an annual document that identifies the ten emerging technologies that offer a competitive advantage to the economy and the economy of Spain.

    Further information

  • INTEC Report 2024: Taking action in an hour of the brave

    INTEC Report 2024: Taking action in an hour of the brave

    The introduction to the Report on 10 Emerging Technologies to Power Spain (INTEC 2024) puts the emphasis on mechanisms to turn ideas into real transformative action.

    Further information

  • 1. An AI that thinks in Spanish

    1. An AI that thinks in Spanish

    Large language models (also called large language models, or LLMs), have experienced unprecedented growth, especially since the popularisation of ChatGPT in late 2022. These models, which are actually artificial intelligence systems and programmes trained on huge amounts of text, have revolutionised several industry sectors, far beyond IT and computing. The ability of LLMs to perform and automate complex tasks, such as copywriting, translation, summarising meetings and reports, and even customer service, has demonstrated the enormous potential of this technology. With the advent of ChatGPT 4o (short for "omni", Latin for "all"), these models are able to interpret instructions in real time from photographs and audio and video recordings, just as dozens of science fiction films imagined decades ago. However, despite the advancement of translation capabilities, these LLM systems present an intrinsic problem when it comes to handling data and text in other languages, such as Spanish. Having been trained and educated with thousands of gigabytes of information in English, many language models are not as agile expressing themselves in the language of Cervantes, which poses many challenges and opportunities for the development of better systems for multilingual applications.

    Further information