This paper describes our experiences related to a culturally situated Artificial Intelligence (AI) education workshop for urban farmers. The farmers explored the concept of AI and its implications in the context of their own farm. They engaged in hands-on activities, including using traditional sensemaking practices in conjunction with sensor technology to collect contextual data. They then used a tangible educational tool: a corkboard with push pins and images to build tactile AI models based on their farm data. Throughout this process, they discussed their hopes and desires regarding AI in farming, and their concerns about AI technologies. The perspectives of the urban farmers reveal their preference for AI systems that are contextual, integrate community values, Indigenous knowledge, and environmental concerns, and are rooted in community ownership of data. Our report provides a starting point for conducting future workshops that involve 'critical participatory design' of AI technologies to promote AI literacies rooted in the community.

Tactile guiding surfaces in the built environment have held a contentious place in the process of navigation by people who are blind or visually impaired. Despite standards for tactile guiding surfaces, problems persist with inconsistent implementation, perception, and geographic orientation. We investigate the role of tactile cues in non-visual navigation and attitudes surrounding guiding surfaces through a survey of 67 people with vision impairments and ten interviews with navigation and public accessibility experts. Our participants revealed several opportunities to augment existing tactile surfaces while envisioning novel multimodal feedback solutions in immediately relevant contexts. We also propose an approach for designing and exploring low cost, multimodal tactile surfaces, which we call navtiles. Finally, we discuss practical aspects of implementation for new design alternatives such as standardization, installation, movability, discoverability, and a need for transparency. Collectively, these insights contribute to the production and implementation of novel multimodal navigation aids.

3D printing offers significant potential in creating highly customized interactive and functional objects. However, at present ability to manufacture functional objects is limited by available materials (e.g., various polymers) and their process properties. For instance, many functional objects need stronger materials which may be satisfied with metal printers. However, to create wholly interactive devices, we need both conductors and insulators to create wiring, and electronic components to complete circuits. Unfortunately, the single material nature of metal printing, and its inherent high temperatures, preclude this. Thus, in 3D printed devices, we have had a choice of strong materials, or embedded interactivity, but not both. In this paper, we introduce a set of techniques we call FiberWire, which leverages a new commercially available capability to 3D print carbon fiber composite objects. These objects are light weight and mechanically strong, and our techniques demonstrate a means to embed circuitry for interactive devices within them. With FiberWire, we describe a fabrication pipeline takes advantage of laser etching and fiber printing between layers of carbon-fiber composite to form low resistance conductors, thereby enabling the fabrication of electronics directly embedded into mechanically strong objects. Utilizing the fabrication pipeline, we show a range of sensor designs, their performance characterization on these new materials and finally three fully printed example object that are both interactive and mechanically strong -- a bicycle handle bar with interactive controls, a swing and impact sensing golf club and an interactive game controller (Figure 1).

In rural parts of the developing world, spreading awareness about critical issues in health, governance, and other topics is challenging and costly. Traditional media such as print, radio and TV each have limitations and offer little guarantee that new information is absorbed or retained by the target population. This paper describes Learn2Earn, a system that leverages mobile payments to bolster public awareness campaigns in rural India. Users call an Interactive Voice Response (IVR) system, listen to a brief audio tutorial, and take a multiple-choice quiz to check their understanding. People who pass the quiz receive a mobile top-up (about $0.14) and have the opportunity to earn additional credits by referring others to the system. We describe a pilot deployment of Learn2Earn in rural India that spread via word-of-mouth to over 15,000 people within seven weeks. Usage was concentrated among young men, many of them students. In a mixed-methods study, we draw upon call logs, electronic surveys, qualitative interviews, and other sources of data to suggest that Learn2Earn could be an effective way to build awareness about important topics.

Email is more than just a communication medium. Email serves as an external memory for people---it contains our reservation numbers, meeting details, phone numbers, and more. Often, people need access to this information while on the go, which is cumbersome from mobile devices with limited I/O bandwidth. In this paper, we introduce WearMail, a conversational interface to retrieve specific information in email. WearMail is mostly automated but is made robust to information extraction tasks via a novel privacy-preserving human computation workflow. In WearMail, crowdworkers never have direct access to emails, but rather (i) generate an email filter to help the system find messages that may contain the desired information, and (ii) generate examples of the requested information that are then used to create custom, low-level information extractors that run automatically within the set of filtered emails. We explore the impact of varying levels of obfuscation on result quality, demonstrating that workers are able to deal with highly-obfuscated information nearly as well as with the original. WearMail introduces general mechanisms that let the crowd search and select private data without having direct access to the data itself.

For visually impaired users, making sense of spatial information is difficult as they have to scan and memorize content before being able to analyze it. Even worse, any update to the displayed content invalidates their spatial memory, which can force them to manually rescan the entire display. Making display contents persist, we argue, is thus the highest priority in designing a sensemaking system for the visually impaired. We present a tactile display system designed with this goal in mind. The foundation of our system is a large tactile display (140x100cm, 23x larger than Hyperbraille), which we achieve by using a 3D printer to print raised lines of filament. The system's software then trades in this space in order to minimize screen updates. Instead of panning and zooming, for example, our system creates additional views, leaving display contents intact and thus supporting users in preserving their spatial memory. We illustrate our system and its design principles at the example of four spatial applications. We evaluated our system with six blind users. Participants responded favorably to the system and expressed, for example, that having multiple views at the same time was helpful. They also judged the increased expressiveness of lines over the more traditional dots as useful for encoding information.

Crowdsourcing is a relatively new model of labor where both the workers and work providers are experiencing its growing pains. A dominant platform that implements this model of labor is Amazon Mechanical Turk (AMT). While AMT has evolved over the years, the changes have focused mainly on work providers and have not addressed the problems workers face (e.g. dealing with market volatility and unpaid time searching for work). In this paper we present emph{TurkBench}, a tool meant to provide workers with personalized market visualization and session management. We discuss the design philosophy of the tool, briefly discuss four Turkers' reaction to a demo, and outline future work.

Physical visualizations come in increasingly diverse forms, and are used in domains including art and entertainment, business analytics, and scientific research. However, creating physical visualizations requires laborious craftsmanship and demands expertise in both data visualization and digital fabrication. We present three case studies that illustrate limitations of current visualization fabrication workflows. We then present MakerVis, a prototype tool that integrates the entire process of creating physical visualizations, from data filtering to physical fabrication. Design sessions with three end users demonstrate how tools such as MakerVis can dramatically lower the barriers to producing physical visualizations. Observations and interviews from these sessions highlighted future research areas, including customization support, using material properties to represent data variables, and allowing the reuse of physical data objects in new visualizations.