Less than a year has passed since the original Swedish edition of this book was published. It didn’t take long until before people started asking for an English translation. So, in mid-May 2022, Studentlitteratur and I decided to initiate a translation project of the book.
One of the more mature applications of artificial intelligence is machine translation, so the option of using AI for the translation task was on the table from the outset. However, in this book, I drive the thesis that the future of work lies in a well-designed interplay between AI and humans. So, how does a human-machine teaming approach for translation work? In this case, we arrived at the following workflow:
1. The Swedish manuscript was automatically translated using a state-of-the-art machine translation service called DeepL.
2. The AI-translated text was then adjusted by me, the original author, for nuances in the content. I took the opportunity to add a few updates to reflect the latest advancements in the field, and I corrected some of the machine translation service’s errors. I had great help from the Grammarly service in this work. I also added this additional Preface section that you are reading right now.
3. Finally, a native English-speaking professional proofread the adjusted translation focusing on tone and language quality, capturing the nuances that DeepL, myself, and Grammarly didn’t.
In this AI-powered teaming approach, the humans in the process (me and the proofreader) could skip the initial mundane translation task and instead focus on the content and nuances in tonality and style of delivery.
In Chapter 6, I write about the energy footprint of AI systems and the desired practice to of reporting on energy usage. The translation service used in this project (DeepL) is powered by hydroelectric power generated in Iceland.