The new era of media accessibility (MA) and audiovisual translation (AVT) is characterised by the widespread application of automation through artificial intelligence and the appearance of alternative and creative forms of accessibility and translation. Human-machine interaction thus comes to the fore as the driving force behind the present and future development of MA. From the point of view of technology, the worldwide rollout of large language models and new-generation artificial intelligence in the Autumn of 2022 brought about a step change in the MA industry. Automation, which until then had been partially applied, started to become a reality, for instance in the form of fully automatic live subtitles, which now coexist with those made by human subtitlers. As technology and automation gain ground, the role of the human in MA is reassessed. As some translators and access experts become editors tasked with revising the automatic output produced by the latest technology, an increasing number of filmmakers are stretching the MA spectrum by moving away from a highly standardised traditional access into alternative and creative forms of access that are being used, amongst others, by disabled artists to claim a space that have so far been denied to them. The aim of this project is to carry out a qualitative and quantitative analysis of some of the key developments of the human-machine interaction at play in this new MA spectrum.

The first part of the project will deal with the use of live and pre-recorded subtitles in standard MA. Building up on the results of the Qualisub project, which showed the need to speed up the use of the NER model to analyse the quality of live subtitles, and in collaboration with the EU Parliament, we will produce a new, AI-powered software that can help to (semi)automise this process. As for pre-recorded subtitles, we will focus on the decision by key stakeholders such as Ofcom to do away with speed limits in subtitling, which may result in fast subtitles that can be inaccessible for many viewers. Following a descriptive analysis of the current subtitling speed used in the main streaming platforms, and along with partners from Sweden, Belgium and Australia, we will undertake a worldwide reception study to ascertain whether or not these new free-speed subtitles are accessible for different groups of users (the elderly, second language and singing deaf viewers).

The second part of the project will focus on alternative MA, and more specifically on filmmakers, access experts and users. We will build the first alternative MA database, which will help us carry out a quantitative and qualitative analysis of who is applying alternative forms of access around the world and how. In collaboration with the Spanish Film Academy, we will then help to introduce in the Spanish filmmaking industry a new professional figure, the access coordinator, who will liaise with filmmakers to consider accessibility and translation from the beginning of the filmmaking process, instead of at the end, as has so far been the case. This collaboration will result in the development of the first access coordination training course in Spain, which will cover the full MA spectrum, from standard to alternative access. Finally, we will carry out a reception study of one of the works included in the alternative MA database to ascertain how viewers react to the alternative forms of access included in the new MA spectrum.