Future Audiences/FAQ
There are many things that should be improved for our current users. Why devote WMF resources to "Future Audiences" now?
editFor most of Wikipedia's existence, we have relied on a symbiotic relationship with search engines to bring readers and new contributors to our projects. However, the last few years have brought major changes in technology and user behavior.[1][2][3] People are increasingly consuming information in new places and ways (short-form video, conversational AI, and more). If we focus 100% of our attention only on serving existing readers and contributors on our projects, we risk losing relevance with more and more readers and potential new contributors if these trends continue.
Our approach is to monitor and learn more about these trends and our role in the possible futures they will bring in a resource-minimal lightweight way, without investing substantial resources into development and maintenance of new products. We have allocated about 5% of WMF's Product & Technology Department resources to Future Audiences work. (In comparison, work on "WikiExperiences" – dedicated to better serving current editors, moderators, etc. – represents about 50% of Product & Technology's resources.)
What is the different between an experiment and a product? The ChatGPT plugin seems like a new product, so why are you calling it an experiment?
editThere are three high-level stages in the lifecycle of a new project or product – Future Audiences is set up to work on the first two stages and make recommendations for the third stage:
- Experiment: prototype/proof-of-concept that uses minimal resourcing/coding (e.g., a side-project for one developer for a limited period of time), used to learn more about a particular technology or user behavior. Not meant to be maintained long-term in its current state. All Future Audiences experiments, including the ChatGPT plugin, start in this category.
- Iteration: if there is more to learn from the initial prototype/proof-of-concept, it may continue to be used for further learning – for example, to test a new use-case, or to test engagement from a different audience. This is still not a fully-committed new product. Some Future Audiences experiments may be iterated on, but this must be done within the initiative's budget, e.g., still limited to 5% of Product & Tech's total resource allocation.
- Full product a team/set of resources is allocated to implementing a stable version of the prototype and continuing to maintain it. Future Audiences may recommend that an experiment is scaled to a full product in the next fiscal year – the rationale for this will be shared with the Board of Trustees and community as part of the normal process for community input on the draft Annual Plan.
Why run experiments on commercial platforms? Doesn't this run counter to our free knowledge mission?
editFor the past two decades, the vision of Wikimedia projects has been to imagine a world in which every single human being can freely share in the sum of all knowledge.[4]
In 2023, among the most popular digital platforms in the world – websites[5] and apps[6] where millions of people globally spend time every day – Wikipedia is one of the only non-commercial platforms. To date, our ability to remain on the list of popular digital platforms has depended in part on Google Search (a commercial platform), traffic from which accounts for about 75% of all Wikipedia reading sessions.[7] With the search ecosystem changing (Google Search is losing traffic to new social apps[8] and conversational AI platforms[9], and has been introducing more rich media results-page features that reduce the prominence of links to sources like Wikipedia[10]), it is important that we better understand how to spread awareness of and attract new audiences to Wikimedia projects, regardless of where or how they like to get knowledge.
How do we make sure we stay realistic about the future and not fall prey to hype/fads?
editFuture Audiences is not set up or resourced to make big investments in new technologies (no matter how hyped-up they may be) – our goal is to learn about a wide variety of technologies and platforms (quickly and cheaply) and provide data back to the movement on where we think the future may be heading. We are also closely tracking key indicators like readership and contribution trends on our and other popular community-generated content platforms – to date, we have not seen major negative impact from AI and other technology trends, but we are aware that some other community-driven platforms may be seeing serious impact,[3] so we remain measured but vigilant on how important these technologies may be (both in the positive and negative sense) to our movement.
I am [excited about/afraid of/upset at] conversational AI. Why [is/isn't] the Wikimedia Foundation doing more to [encourage/discourage] this technology?
editOur understanding based on an open community call with over 100 Wikimedians from a variety of projects and languages on AI earlier this year[11] is that all of these views (interest in, active usage of, skepticism of, anger at new conversational/generative AI products and applications) and more are currently represented within our movement, and there is as yet no global consensus on whether and how AI can or should be used on our projects.
The Wikimedia Foundation view of conversational/generative AI specifically is that we (Wikimedians, Mediawiki software developers, and WMF staff) have developed and used machine-assisted tools and processes on our projects for many years, and it is important to keep learning about how recent advances in AI technology might help our movement; however, it is equally important not to ignore the challenges and risks that commercial AI assistants may bring not just to our model of human-led knowledge creation and sharing, but to the entire ecosystem of digital knowledge.[12]
Will we run experiments on our projects/with our current users?
editThe focus of Future Audiences is just that – finding ways to engage future audiences (those who are not currently familiar with or consistently visiting our projects) to share in our mission. That said, we would like explore whether offering a different knowledge search and/or reading experience – e.g., rich media and/or AI article summaries – would make Wikipedia a more enticing destination, especially for younger audiences. Given our resourcing, it is likely that this work will be implemented as a prototype that does not change the experience for current audiences (though, similar to the ChatGPT plugin, current readers and editors will be free to preview/play with it, as well).
How can I stay informed and provide input on Future Audiences experiments?
edit- Weekly: We will endeavor to keep this page updated with the latest on current and new potential experiments and answer any questions on various Future Audiences talk pages.
- Monthly: We will host monthly open office hours on Zoom. We will publicize the meeting dates/times here and on popular community channels. Anyone who is interested in following along with this work is welcome!
- Quarterly: We will post slides from Quarterly Learning sessions that include a higher-level analysis of the latest learnings, trends, and recommendations, here and on Commons.
References
edit- ↑ "For Gen Z, TikTok Is the New Search Engine". The New York Times. September 16, 2022.
- ↑ "Google it? People now are searching with TikTok or Reddit.". The Washington Post. July 20, 2023.
- ↑ a b "The Fall of Stack Overflow". July 26, 2023.
- ↑ "Wikimedia vision statement".
- ↑ "Top Websites Ranking". July 1, 2023.
- ↑ "The most popular app in the world revealed". April 5, 2023.
- ↑ "Searching for Wikipedia: DuckDuckGo and the Wikimedia Foundation share new research on how people use search engines to get to Wikipedia". September 23, 2021.
- ↑ "Google Search is in danger of losing Gen Z. Here’s how the company is responding". July 13, 2022.
- ↑ "Google-parent stock drops on fears it could lose search market share to AI-powered rivals". April 17, 2023.
- ↑ "Google is rethinking search because TikTok and podcasts are taking over the internet". September 30, 2021.
- ↑ "2023-2024 External Trends community call notes".
- ↑ "Wikipedia’s value in the age of generative AI". July 12, 2023.