To understand the range of member experiences within the three open AI communities, we have
conducted interviews with community leaders to map user journeys for key points in a member’s
experience with a community: discovering, joining, contributing, and leading.
Discovering Engaging with a community begins with discovery, a growing challenge in AI where
new initiatives are emerging rapidly. For all three communities, the influence of prominent members
and the reputation of organizations supporting the groups (e.g. HuggingFace, The Alan Turing
Institute, and Mozilla Foundation) played important roles. However, in order to expand reach beyond
existing networks, the communities applied tactics to diversify their membership. BigScience’s
founding team took steps to outreach outside their geographical and professional domains through
situated events like local data hackathons. The Turing Way seeks out collaborations with open science
organizations around the world to build off each other’s work and connect their communities. The
MozFest TAIWG is made up of engaged members from the Mozilla Festival Community and the
Mozilla Developer Community. To expand our reach, they have also invited participation from local
partners in cities where the Mozilla Festival is held.
A common draw for new members is in a shared challenge and space for addressing it. Whether it’s
an openly developed LLM, improving research culture, or building trustworthy AI, the communities
center on a direction of AI work that may not be accessible elsewhere for aspiring members.
Joining Joining is an important point in a user journey that can be defined by the barriers presented.
A comparison can be made to the process of joining an academic or industry AI research lab, where
an individual must undergo years of accreditation and a gauntlet of interviews to even be considered
for entry. This process is a major limiting factor, and as AI becomes more impactful in everyday life,
this barrier to entry exacerbates the power dynamic between AI producers and everyone else.
In contrast, each of the three collaboratives is open to any interested participant who typically find out
about the community and join through digital doorways such as an online channel (e.g. email listserv,
Twitter), collaboration space (e.g. Slack), or video call meeting or event (e.g. Zoom). Because these
channels are available to anyone with access to these online resources, this means many more people
in different time zones, backgrounds, and skill levels can enter.
While lowering the barrier to entry is the first step to the joining process, it is not enough to facilitate
active participation. In our conversations with community leaders, they shared that the act of “lurking”
is common practice and not something to be stigmatized Chen and Chang (2013). Whether it’s
listening in on meetings or consuming and reacting to content on Slack, this behavior is characteristic
of digital spaces and offers an easy, safe way to explore the community before contributing.
Contributing After a member has passed through the digital doorway to join a community, the
next step in their user journey is to begin contributing. Community leaders remarked on how the
kinds of available "jumping in" points varied by the stage of the community. This was particularly
important for BigScience and the TAIWG which have hard deadlines associated with working groups,
which members join based on their skill sets and interests. Because the core activity for The Turing
Way is co-writing, members can easily join at any point in time. However, the process of contributing
3