During the past year, there has been intensive discussion about how artificial intelligence will change our society. People have experimented joyfully with AI-powered language models and machine learning algorithms such as ChatGPT asking them trivia questions, cheating with them in essay assignments and using them to generate amusing poems and fake photographs. At the same time, the old debate about whether automation will make traditional professions such as professional transcription obsolete has resurfaced.
A lot of these debates are fueled by a misunderstanding of the nature of reporting and transcription work. The popular website willrobotstakemyjob.com states that professions like court reporting and simultaneous captioning face an extremely high likelihood of being automated in the future since “these jobs consist primarily of repetitive, predictable tasks with little need for human judgment”. Anybody who is even faintly familiar with these jobs knows the staggering amount of careful, specialised and context-dependent human judgment that goes into intra-linguistic translation between the spoken and written medium, from deliberating on what to include and exclude from a transcript to editing the outcome to fit the demands, aims and style of the genre whether that be an official court report or subtitles for the hard-of-hearing.
The development of AI-based solutions will probably mean that in many transcription-based professions, the centre of gravity will move from physical activities like typing to more specialist activities like editing. In fact, this process is already well on the way in such professions as parliamentary reporting and live subtitling. However, even though AI does a great job in reducing manual labour, it is still highly dependent on human supervision and intervention in most transcription tasks. This came up quite clearly in an answer provided to me by the AI application ChatGPT itself when researching my article for this issue of Tiro, I asked it about using AI for transcribing spoken interaction. I suspect that this dependency will not change dramatically anytime soon.
That said, there are of course many advantages to be found in adopting and using the latest technology. As with past innovations, we should ride the AI wave and take all the lessons that we can from it to develop our professions. That way, we stay aware of what is happening around us and can adapt to future changes.
This issue provides us with several excellent examples of the not artificial but genuine human intelligence demanded by professional reporting and transcription. Dikla Abravanel writes about including and excluding interruptions in the official reports of the Knesset, Parliament of Israel. Aysenur Acar and Elif Yayla introduce us to the organisation and work processes of the record office of the Grand National Assembly of Türkiye. Luz Belenguer Cortés discusses combining languages, code-switching and code-mixing in live subtitling. Each article demonstrates from a different perspective the intricate human analysis involved in the making of transcripts in different contexts and for different audiences.
Looking into the future should not mean forgetting or abandoning history. Jonaš Vala takes us on a journey through the development of shorthand and stenography in Czech Republic and introduces us to a modern string stenography application ZAVPIS for the Czech language. D’Arcy McPherson examines a recently published report on stenography progress and trends in China. Unlike some predictions, the report foresees a growing need for stenographers in the future and describes initiatives to ensure that there will be enough students and practitioners to fulfill that need.
This edition also includes our first book reviews. Luz Belenguer Cortés introduces us to a new book on diamesic translation, while Mary Sorene checks out a book about the history of shorthand in Australia. We are happy and proud to be publishing these first reviews and we encourage readers to offer us new reviews for future issues!
All in all, it takes only a glance at this edition of Tiro to see that reporting and transcription professions are thriving and moving forward amid the development of new technologies. In his regular column, Carlo Eugeni analyses the different aspects of human-machine interaction in reporting and transcription and inspires us to be conscious of when a process should be considered “human-aided” rather than “computer-aided”. Whatever we call the process, however, it is clear that the human contribution remains an invaluable, indeed essential, part of high-quality reporting and transcription.
Eero Voutilainen is Tiro’s editor-in-chief.
[…] Eero Voutilainen:Will Artificial Intelligence Steal Our Jobs? […]