Digikult 2024 or: The digital gap is growing. What do we do next?

Digikult is an annual conference that brings together professionals from across the cultural sector, including archaeologists, archivists, artists, curators, developers, and librarians. Here's a summary of what I listened to, discussed, and reflected on during my train ride home.

Familiar topics, fresh updates

Several familiar themes were explored, some updated and enhanced in new ways: 3D technologies (scanning, modeling, and reuse) in archaeology and art, augmented and virtual reality, and place-based experiences and apps (here's the program to get a sense of different projects). One standout example was "Historiska vandringar", a web app offering self-guided historical tours without the need to download or sign up. It’s a great multimedia guide that reflects the past decade's learnings in user-friendly design.

Another important discussion focused on how digitization and digital heritage are becoming crucial in today's world. As wars destroy monuments and digital infrastructure becomes a vulnerable target for cyberattacks, the importance of preserving heritage in digital form and making it accessible cannot be overstated.

AI, between hype and work companion

No modern conference can avoid the topic of AI, and Digikult handled it brilliantly, cutting through the hype to focus on what cultural institutions can realistically achieve today. Experts like Mia R. , David Haskiya , Niko Anttiroiko , Martin Malmsten , and Anna Foka showcased projects, some of them openly licensed and ready for experimentation. Even if you're not ready to dive in right away, I highly recommend watching their presentations once they’re published by Riksantikvarie?mbetet . These projects offer a glimpse into the changes some institutions are already shaping, even if they seem light-years ahead of the day-to-day operations of others.

First of all, there were a few familiar topics, updated and upgraded in different ways: 3D (scanning, modeling, re-use) in archaeology and art, Augmented and Virtual Reality, place-based experiences and apps (here's the program to get a sense of different projects). A good example is "Historiska vandringar", a web app with self-led guided tours with a history focus that don't force you to download or sign up for something, and whose quality seems quite high. A simply well-done multimedia guide incorporating much of the last decade's learnings. Others showed how digitisation and digital heritage can become crucial in this new time we live in, with as monuments being destroyed in wars, or how our digital infrastructure is at risk to become a low-hanging fruit for cyber attacks that look for simple acts with huge societal impacts.

That said, there are significant hurdles. Many of the advanced tools discussed are still out of reach for most institutions, either locked behind paywalls or requiring basic programming skills that many staff don't have. For example, Riksarkivet Swedish National Archives ' work on handwritten text recognition for documents from the 17th century onward could revolutionize how we search and access archival material. Similarly, Kungliga biblioteket 's use of machine learning to tag and describe literature promises to make collections more searchable. These innovations spark dreams of a future where tedious digitization tasks are automated—but we’re not quite there yet.

The Growing Divide

The digital divide is widening. While some institutions are pushing the boundaries of how we're working today, many others are still struggling with basic digitization tasks. Vast amounts of material remain undigitized, while much of what has been digitized was done in such low resolution or with so little metadata that it requires significant manual work to become useful. I worry that securing funding for this foundational work will become even harder as the gap between projects with shiny new things and the essential groundwork grows. The flashy appeal of AI and machine learning risks overshadowing the need for continued investment in the basics.

Another challenge is that many cultural heritage institutions are locked into outdated collections management systems, which don't evolve as quickly as the technology around us. While stability is important, it means we may be years away from fully integrating AI tools for tasks like image description, data cleanup, and text transcription. Most systems today don’t even have ways to link AI-generated information to object records, let alone flag it appropriately.

In other sectors, AI is often seen as either a miraculous solution or a dystopian threat. As usual, the reality lies somewhere in between, and it depends on your perspective. For medium-sized and smaller institutions, AI could help solve some of the persistent issues plaguing digital collections—problems that we will likely never have the human resources to tackle on our own.

A Call to Collaborate

Mia R. made a point that resonated with me:

"Cultural heritage institutions move at a glacial pace. We shouldn’t worry about moving too slowly. We just need to start, each at our own pace."

She shared how The British Library fosters internal learning through reading groups where colleagues discuss and test new tools together for an hour or two. While large institutions can manage this internally, I believe the cultural heritage sector as a whole should collaborate more effectively.

Things are moving fast, and we might need to rely on each other to build capacity across institutions. It's time to get better at experimenting, testing, failing, and moving forward—together. We excel at presenting polished work at conferences, but we need to improve at supporting each other along the way.

So, here’s my idea: Would anyone like to join me in starting an informal group where we can test, discuss, and recommend AI tools to help cultural heritage institutions with their workflows? It’s open to everyone—whether you’re from a small or large institution. If you're curious, get in touch!