Module # 5 is based on Design Consideration # 8:
For the purposes of this learning circle, we will take some liberty by reframing this competency:
The shift from “critical consumers of AI” to “critically engage with AI” better reflects our goal of engendering agency. Indeed, agency is beyond “consumption.” In fact, the course actively problematizes the separation between “providers” and “consumers” of AI. The title of the final module reflects this problematization: “WE” are AI. So far, the preceding modules have provided the competencies necessary for engagement. The goal of this concluding module is to demonstrate engagement (through the ADS activity) and then provide a roadmap for engagement moving forward (through the AI Actions).
The individual warm-up is structured to get learners thinking about engagement and teases the AI Actions document that is presented later in the module. This foreshadowing probes what (if any) sense of agency the group possesses at this point. Facilitators could refer back to this discussion during the closing discussion as a way to structure these early thoughts on engagement.
The course continues the conversation on agency and engagement with a video that stresses agency and directs it specifically on automated decision-making systems. The video defines ADS and focuses especially on its use in hiring. Then, it turns to local efforts to regulate ADS. Throughout, the video references previous concepts (features, classifiers, data etc.) that all require human judgment. Thus, the video argues AI is socio-technical but beyond that fact, AI needs to be rooted in people in the sense of broad, diverse engagement in its design and use. To this end, the video ends with a call to action via the visual concept of “Tech Rooted in People. ”
The group activity attempts to model the call to action from the video through an activity on ADS used in Child Welfare Services. The activity is intended to model civic participation and deliberation around ADS. In reality, this kind of debate rarely happens which should be emphasized. After the debate, the facilitators can revisit the local ADSs mentioned in the video and how the use of these systems in government continues to grow. Here, facilitators could reference the NYC ADS task force, the hiring bill, the CTO, the AMPO, and other local civic groups and organizations where these conversations are happening.
The course ends with the AI Actions document: a resource of actions for participants to continue the journey of engaging AI started in the learning circle. The list is intended to be a living document where people can gather information and actions across different cohorts/iterations of the learning circle. Rather than prescribing specific actions or stance, the AI Actions is more modest in that it provides a range of actions for participants to enact their agency as they see fit. By concluding with the AI actions, the course does encourage agency, without defining the exact form that agency should take (e.g., joining a specific organization or adapting political stances (racial justice, decentralizing power)).