Jump to content
Wikimedia Meta-Wiki

AI Sauna/Project ideas

From Meta, a Wikimedia project coordination wiki
– Click here for instructions

Add a new idea:

  1. Click on Add a new idea
  2. Fill in the basic information and save the page

Edit an existing idea or add yourself to the team in an idea:

  1. Click on Edit ideas above.
  2. Select the idea you wish to edit. Click Edit
  3. In the popup window, make the changes you like. You can add more input options from the left panel by ticking one of the blue checkboxes. Confirm by clicking on Apply changes.
  4. Finally, click on Publish changes... to save the page.
In progress

History of the Basque Country in 100 objects

Demo / Transmedia project A GLAM project to show the history of the Basque Country in 100 objects displayed in museums. We are currently prototiping new transmedia ideas beyond articles, so any proposal to improve the project will be interesting.
Project page Create a project page or add information about an existing page.
Open for collaboration

Governance of a collaborative visual name authority

Software / Governance To apply facial recognition technology to enhance descriptive metadata of heritage photo and video collections, a qualitative reference set of the people to be recognised are key. That's why we want to make a shared/collaborative visual list of person names of public figures in Flanders (Belgium). It will consist of:
  • an overarching list of relevant personal names that are uniquely and unambiguously identified and linked to existing authorities;
  • the creation of a shared reference set of portrait photos of relevant individuals for facial recognition.
This undertaking asks for a good governance on community, data and infrastructure levels. E.g.: who can see/edit/use the data? Which images can go into this visual name authority? Where will the data be stored?
Bart
In progress

Dreaming of workflow for digital humanities

Software / Demo / Design project What keeps you back from unleashing the full power of free tools for digital humanities? Is it the hassle of figuring out the right version of Python or the confusion surrounding your system's specs?


Stop dreaming and join MessyDesk team. MessyDesk is a bridge to services that are behind of technical gap between users and technology in the realm of digital humanities. We are trying to build simple workflow that is transparent and self-documentative.
Ari Häyrinen


Open for collaboration

Designing human-scale permission for machine-scale operations

Discussion / Paper prototype We want to work on how to ensure Flickr members are aware of, and give permission for, their images being used as training data in AI systems. The horse may have left the stable, but we at flickr.org think it's an important piece of ethical friction that has been lacking in loads of social platforms. Whether this is drawings on paper or a conversation, we'd love your input.
George Oates


Open for collaboration

Preference Signals in Open Culture

Paper prototype CHIs have a wealth of material, and data, that is increasingly coming online, through our shared efforts in the Open Culture Movement (CC, Wikimedia, and others.) And so the data of the world's cultural heritage collections is also becoming available for analysis and interpretation and ingestion by AI softwares.

CC is exploring preference signals for creators but their application and interpretation is potentially relevant to CHIs everywhere as well and so we would like to explore that together. As with the Licenses we are approaching preference signals broadly and considering the variety of ways that may make sense to indicate a use is preferred.

Besides the implications of preference signaling in CHI’s we want to explore their implementation given legal certainty or uncertainty. Where it is also relevant to explore is whether CHI’s would be allowed/should be allowed to indicate preferences on behalf of the creators.
Connor Benedict


Open

Hot topics in the Finnish local letters of the 1860s

Experiment / Learning Use topic modeling of other suitable technique to find emerging trends or topical interests in the local letters sent by locals to the newspapers of the time.
TuulaP
Project page Create a project page or add information about an existing page.


Open for collaboration

Generate alt-texts for historical images

Demo Can we make historical images searchable using AI, without relying (much) on the metadata, similar to the prototype by the National Museum of Norway? There are two image sets among the AI Sauna Resources as well as a vector database, and LUMI for running AI models, these could be combined into a prototype.
Osma Suominen


In progress

GenAI for Moroccan Arabic

Experiment / Learning Using GenAI to generate biographies in Moroccan Arabic.
Ideophagous


AI for Wikimedia Commons

Discussion / Demo Experiment with existing AI models to classify, categorize, describe and title Commons images, also to help Wiki Loves Monuments
Ferdinando Traversa


Embodied creative process in the context glassblowing

Liisi Soroush
Project page Create a project page or add information about an existing page.
Open for collaboration

Authorship of political artists

Art and audio
Liisi Soroush
Open for collaboration

Vector database of Senates Justice Department

Demo
Juho Inkinen, Atte Föhr, Mikko Lipsanen, Ilkka Jokipii
In progress

Truths or truisms: How to evaluate AI generated texts?

Unpublished article This is a one-person writing assisted by ChatGPT. We are entering times where Wikimedia and Wikipedia communities are expected to give education on AI-generated texts. How to evaluate these texts?
Heikki Kastemaa
Open for collaboration

Summary of all knowledge

Software I want to create a software component that creates a summary text using all Wikipedia language editions of a topic. I would use this text in the Wiki Loves Living Heritage pages such as the page on Chhau dance
Susanna

AltStyle によって変換されたページ (->オリジナル) /