The Living with Machines Report
Living with Machines came to an official end on 31 July 2023. It has been a genuine privilege to work with this team over the last 5 years. As part of our conclusion phase we have created a report which lists some of our many achievements, and reflects on our journey as a project.
Here’s an excerpt of the Introduction that I wrote for the report.
It is with great pride that I write this end of project report, as well as some sadness. When the other investigators and I set out the vision for this project in 2017, we had some big dreams. Living with Machines was imagined at once as a data-driven history project, and a historically-informed data science project. Our object of interest was an earlier moment of huge technological upheaval: the coming of the machine age in the nineteenth century. One consequence of industrialisation in Great Britain was an explosion in the creation and collection of information. Our aim was to show how, thanks to decades of digitisation, we were now in a position to leverage this information as data through computational means. We proposed to develop innovative computational approaches to facilitate a new kind of history that would allow us to tell the stories of the impact of mechanisation on the lives of ordinary people.
But that historical aspiration was just one dimension of what we hoped to achieve. More fundamentally we wanted to develop computational models, tools, code, and infrastructure that would be transformative to the future study of cultural heritage collections. For a project to achieve such broad ambitions, we needed a diverse set of skills and expertise. Over its lifetime, the project has brought together historians, data scientists and research software engineers, curators and library professionals, computational linguists, digital humanists, visualisation experts, literary historians and an urban geographer. As such, Living with Machines was a bold experiment in radical collaboration across disciplinary areas and professional spheres. The experience of building a team and united project vision from these disciplinary parts has been an important experience in its own right, and reflection on this process has led to the publication of our first project book, Collaborative Historical Research in the Age of Big Data: Lessons from an interdisciplinary project (Cambridge University Press, 2023). It is important that others can benefit from our experience, and our challenges.
Building a team such as this means that we not only draw expertise from a breadth of different communities, but develop a very different set of outcomes from those normally generated by history or data science projects. We are just as proud of the open source code and tools, and our databases and language models, as we are of our historical interventions. As our statistics and case studies will show in the following pages, we have been immensely productive in this space, creating 45 public Github code repositories to date, as well as several more polished tools that are being taken up by the community, such as MapReader, DeezyMatch, and T-Res. Through these tools we have created textual data from newspapers, visual data from maps, and tabular data from census returns, as well as digitising 488,000 additional newspaper, and over 15,000 new map sheets, in addition to other material such as the Road Acts and Micthell’s Newspaper Press Directories.
Thanks to working with the British Library, we have also been able to engage with the public in sustained ways throughout the project, not just at its end point. The project exhibition at Leeds City Museum – Living with Machines: Human stories from the industrial age – attracted over 42,000 visitors. The exhibition incorporated key themes and early outcomes from the project. Exhibition themes were also informed by discussions with the public during crowdsourcing projects about the impact of machines, and work on OS Maps was represented by an animation showing changes over time around Leeds. Crowdsourced material was also incorporated into two interactive visualisations. Beyond the exhibition, crowdsourcing has formed a central plank of several pieces of research on the project, including analysis of the language of mechanisation, and the study of industrial accidents – both of which will be featured in the project’s second book, Living with Machines: Computational Histories of the Age of Industry (in progress). We are very pleased that our research has also reached the public through the development of our work on OS maps into a story run by The Economist in April 2023.
One of the aims of UKRI’s Strategic Priorities Fund, which supported this project, is to ‘drive an increase in high quality multidisciplinary and interdisciplinary research and innovation’. This is not simply about delivering some multidisciplinary and interdisciplinary research outputs, but driving a change in research culture beyond the project – for us, in the space of digital history and the use of cultural heritage collections. Where many endeavours fall short is with the attitude ‘if you build it, they will come’. New methods, tools and datasets are of no use if nobody knows about them. In our final phase of the project we have been focusing on ensuring the legacy of our collaboration. As well as making our data and code open for reuse, we are seeking to develop communities of users around our most important tools and methods through blog posts, workshops, in-person and published tutorials. We have also encouraged the uptake of our methods and data through the award of six ‘digital residencies’. These residencies are small fellowships or project awards designed to enable work around one of our datasets or tools. The fruits of their labour will be reported in due course on our project blog (https://livingwithmachines.ac.uk/latest/). Finally, we are maximising our impact by creating a 10-part online docuseries. These short videos help us communicate with a broader audience about what we have been up to, why, and how our work might be useful to them.
More broadly, we believe we have shown that library, arts and humanities scholars not only deserve a place at the table with those designing the future of AI and data science, but also have skills and knowledge that are vitally needed.
If you want to find out more, you can download the full report, and you can remain up to date with future publications and outputs via our mailing list Subscribe here