Future Museums: Can AI Really Make Exhibitions More Inclusive?

Museums are evolving faster than we think. A recent study examines how technology, digital culture, and political struggles intertwine in cultural spaces. First, the authors revisit critiques of the modernist museum model, often seen as neutral but actually Eurocentric. New currents like New Museology and decolonial perspectives demand more open, less authoritarian museums. Then the focus shifts from the traditional, temple‑like museum to an “interface” museum: digitization, immersive exhibitions, and interactive games. These innovations promise broader access, yet if applied uncritically they can reproduce the same inequalities, hiding diverse narratives. Finally, the paper explores hyper‑connected museums where artificial intelligence steps in. Projects such as Demonumenta and Curationist demonstrate how AI can curate, mediate, and reconstruct collections—only when algorithms are made visible and open to critique. The notion of “meta‑algorithmic curation” proposes turning data from supposedly neutral tools into political devices that amplify marginal voices. In short, the research calls for technology to become a lever for more inclusive, distributed, and politically engaged museums, while constantly questioning digital choices.

The concept of meta‑algorithmic curation stems from recognizing that algorithms are not neutral entities but reflect the choices of their designers. In museum contexts, visitor data, artwork recommendations, or 3D reconstructions are often powered by machine‑learning models trained on historically biased datasets favoring Western collections. To make these systems fairer, “algorithmic auditing” practices are required: systematic analysis of AI decisions, bias identification, and implementation of diversity metrics. Transparency can be enhanced through open‑source interfaces that allow curators, researchers, and the public to tweak selection parameters. A practical example is using generative neural networks to reconstruct incomplete artifacts: if the training set contains only classical art examples, reconstructions will tend to replicate Eurocentric styles, sidelining African or Asian influences. An inclusive approach integrates ethnographic datasets and co‑creates with local communities, ensuring reconstructions reflect a plurality of traditions. Moreover, gamification and immersive experiences must follow universal design principles, avoiding cognitive or sensory barriers. Adopting open standards for metadata, such as CIDOC‑CRM, facilitates interoperability among collections and enables diverse actors to contribute to a shared narrative. In short, the key to fairer digital museums lies in combining algorithmic audits, diversified datasets, co‑design with communities, and open data standards.