2023:Program/Submissions/Wikipedia and ChatGPT - XPYVYJ
Title: Wikipedia and ChatGPT
Speakers:
Heather Ford
Heather Ford is a writer, researcher and thinker of internet technologies. Her key work has been for Creative Commons as co-founder of Creative Commons South Africa and Wikipedia in her book about how Wikipedians wrote the story of the 2011 Egyptian revolution (“Writing the Revolution: Wikipedia and the Survival of Facts in the Digital Age” MIT Press). She is the co-founder of a number of digital activist and community groups including Creative Commons South Africa, ethnographymatters.net, InfoCamp and OxDEG (the Oxford Digital Ethnography Group) and she has worked as a Digital Vision Fellow at Stanford University, as a Google Policy Fellow, a Fellow for the Software Sustainability Institute in the UK and a University Academic Fellow in Digital Methods at Leeds University.
Room:
Start time:
End time:
Type: Workshop
Track: Legal, Advocacy, and Risks
Submission state: submitted
Duration: 60 minutes
Do not record: false
Presentation language: en
Abstract & description[edit source]
Abstract[edit source]
Large Language Models (LLMs) like ChatGPT have captured global attention, promising great advances in knowledge generation. Some have identified the possibilities of using LLMs to generate a first draft of an article. Others have highlighted the possible risks of ChatGPT to Wikipedia if it is used to dump large amounts of unverified information on the site. In this roundtable, we'll discuss how Wikipedia might respond to advances in LLMs like ChatGPT and what is at stake in doing so.
Description[edit source]
The aim of this session is to gather community perspectives on the risks and opportunities of LLMs like ChatGPT across multiple language versions of Wikipedia. The session will begin with a 15 minute presentation of what LLMs are and do, what we know about their implications for knowledge integrity and information quality on Wikipedia and what we still need to know. We will then move to producing an annotated list of questions and considerations relating to chatGPT and Wikipedia by brainstorming around three key themes (either in groups or as a single cohort, depending on numbers):
1. A SWOT analysis when considering LLMs in relation to Wikipedia practice across language versions; 2. Current Wikipedia policies that are implicated by LLMs; 3. Current Wikipedia practice in relation to AI and automation that we can learn from.
Participants will leave the session with a more grounded understanding of ChatGPT and its possible implications for Wikipedia, as well as some thoughts about what we still don't know and need to do to ensure that LLMs are an opportunity rather than a threat to Wikipedia's knowledge integrity.
Further details[edit source]
Qn. How does your session relate to the event themes: Diversity, Collaboration Future?
Wikipedians have always thought about how new technologies relate to their ultimate goals of representing the sum of all human knowledge. ChatGPT is no different. This session will garner the local knowledge of Wikipedians to think through the possible implications of chatGPT and LLMs for their projects and to help them better understand what we currently know and need to know in order to deal with the risks that LLMs might present.
Qn. What is the experience level needed for the audience for your session?
Everyone can participate in this session
Qn. What is the most appropriate format for this session?
Onsite in Singapore
Remote online participation, livestreamed- Remote from a satellite event
- Hybrid with some participants in Singapore and others dialing in remotely
- Pre-recorded and available on demand
