We present an account of an ongoing practice-based Design Research programme that explores real-time AI image generation. Based on three installations, we reflect on the design of PromptJ, a user interface built around the concept of a prompt ‘mixer’. We present a series of strong concepts based on the design and deployment of PromptJ. Later, we cohere and abstract our strong concepts into the notion of Holistic Prompt Craft, which describes the importance of considering all relevant parameters concurrently. Finally, we present PromptTank, a prototype design which exemplifies these principles. Our contributions are articulated as strong concepts or intermediate knowledge, intended to be used generatively by informing and inspiring practitioners and researchers working in this space.

The proliferation of Generative AI (GenAI) has created a need for effective tools and strategies that teach young people how to use and think about these technologies. This paper presents work in progress extending a successful trial of an AI-focused constructionist learning package, which included a proprietary GenAI tool, lesson plans, and supplementary teaching resources. To explore the scalability of this approach, we interviewed education practitioners. In these discussions they reflected on findings from the initial trial to identify potential challenges and opportunities for broader implementation. From these insights, we propose a set of design heuristics for designing scalable GenAI education platforms.

This pictorial presents an ongoing research programme comprising three practice-based Design Research projects conducted through 2024, exploring the affordances of diffusion-based AI image generation systems, specifically Stable Diffusion. The research employs tangible and embodied interactions to investigate emerging qualitative aspects of generative AI, including uncertainty and materiality. Our approach leverages the flexibility and adaptability of Design Research to navigate the rapidly evolving field of generative AI. The pictorial proposes the notion of prompt craft as a productive reframing of prompt engineering. This is comprised of two contributions: (1) reflections on the notion of materiality for diffusion-based generative AI and a proposed method for a craft-like navigation of the latent space within generative AI models and (2) discussing interaction design strategies for designing user interfaces informed by these affordances. The outcomes are presented as strong concepts or intermediate knowledge, applicable to various situations and domains.

1. A key tenet to this paper’s contribution is the assertion that while there are many specialized terminologies relating to more-than-human design – terms including but not limited to posthumanism...

Generative AI technologies demand new practical and critical competencies, which call on design to respond to and foster these. We present an exploratory study guided by Research-through-Design, in which we partnered with a primary school to develop a constructionist curriculum centered on students interacting with a generative AI technology. We provide a detailed account of the design of and outputs from the curriculum and learning materials, finding centrally that the reflexive and prolonged ‘hands-on’ approach led to a co-development of students’ practical and critical competencies. From the study, we contribute guidance for designing constructionist approaches to generative AI technology education; further arguing to do so with ‘critical responsivity.’ We then discuss how HCI researchers may leverage constructionist strategies in designing interactions with generative AI technologies; and suggest that Research-through-Design can play an important role as a ‘rapid response methodology’ capable of reacting to fast-evolving, disruptive technologies such as generative AI.

Growing concerns over negligent or malicious uses of AI have increased the appetite for tools that help manage the risks of the technology. In 2018, licenses with behaviorial-use clauses (commonly referred to as Responsible AI Licenses) were proposed to give developers a framework for releasing AI assets while specifying their users to mitigate negative applications. As of the end of 2023, on the order of 40,000 software and model repositories have adopted responsible AI licenses licenses. Notable models licensed with behavioral use clauses include BLOOM (language) and LLaMA2 (language), Stable Diffusion (image), and GRID (robotics). This paper explores why and how these licenses have been adopted, and why and how they have been adapted to fit particular use cases. We use a mixed-methods methodology of qualitative interviews, clustering of license clauses, and quantitative analysis of license adoption. Based on this evidence we take the position that responsible AI licenses need standardization to avoid confusing users or diluting their impact. At the same time, customization of behavioral restrictions is also appropriate in some contexts (e.g., medical domains). We advocate for “standardized customization” that can meet users’ needs and can be supported via tooling.

The Shadowplay installation facilitates creative and embodied interaction with a generative AI image diffusion model. The work aspires to facilitate a tangible experience of working with generative AI and produces striking, aesthetic, and provocative images exposing the edges of our creative relationships with this new class of technology. Exhibition-goers enter into a generative interplay between the probabilistic uncertainty of AI technologies, the familiarity of light, and the experimental playfulness of using one’s body to cast a shadow. By enabling a tangible interaction with this rapidly-evolving technology, we probe the quality and nature of AI-mediated imaginaries, while also providing a ‘lived’ means to experience the limits of its expressiveness and the inseparable ties to underlying training data. Shadowplay balances a provocative and critical stance on AI and creativity, produces a captivating stream of beautiful and striking imagery, and achieves this through an engaging, playful, and tangible interaction.

This article explores the intersections and resonances between unmaking and more-than-human design. We begin by aligning unmaking with decentering, a fundamental practice in more-than-human design, through their shared movement and materiality. Using Lindström and Ståhl’s notion of the double movement in un/making, we analyze a series of workshops focused on designing with AI, annotating what was un/made and de/centered during the workshops’ activities. Through this analysis, we introduce two key contributions that highlight some opportunities in the diffractive alignment between unmaking and more-than-human design: firstly, the notion of “unmaking-with” as an emergent concept to describe a posthumanist unmaking practice, and secondly, three decentering tactics—situating, materializing, and enacting—that instantiate this practice through design. Finally, we discuss how unmaking can enrich more-than-human design and, conversely, how more-than-human design can help define the epistemological scope of unmaking.

Artificial intelligence (AI) technologies are widely deployed in smartphone photography; and prompt-based image synthesis models have rapidly become commonplace. In this paper, we describe a Research-through-Design (RtD) project which explores this shift in the means and modes of image production via the creation and use of the Entoptic Field Camera. Entoptic phenomena usually refer to perceptions of floaters or bright blue dots stemming from the physiological interplay of the eye and brain. We use the term entoptic as a metaphor to investigate how the material interplay of data and models in AI technologies shapes human experiences of reality. Through our case study using first-person design and a field study, we offer implications for critical, reflective, more-than-human and ludic design to engage AI technologies; the conceptualisation of an RtD research space which contributes to AI literacy discourses; and outline a research trajectory concerning materiality and design affordances of AI technologies.

Video-recording remote interviews can sometimes be necessary or desirable, such as in news broadcasting or documentary-making. However, remote interviews are not currently well-supported by digital tools. Unresolved questions about best practices and the kinds of support needed to facilitate remote interviews have become increasingly relevant since the Covid-19 pandemic. To reflect on these questions and explore the design space for systems to support high quality, remote video-recorded interviews, we conducted an exploratory Research through Design study, drawing on professional media-making techniques, novel interviewing methods and a bespoke intervention: The Interview Box. We provide a detailed summary of our design process and, reflecting on both the successes and failures of our interventions, construct two contributions: technical insights relating to the practical challenges of designing and implementing a remote video interview system, and general insights into the broader interaction design challenges of designing for remote video-recorded interviews.

Everyday interactions with computers are increasingly likely to involve elements of Artificial Intelligence (AI). Encompassing a broad spectrum of technologies and applications, AI poses many challenges for HCI and design. One such challenge is the need to make AI's role in a given system legible to the user in a meaningful way. In this paper we employ a Research through Design (RtD) approach to explore how this might be achieved. Building on contemporary concerns and a thorough exploration of related research, our RtD process reflects on designing imagery intended to help increase AI legibility for users. The paper makes three contributions. First, we thoroughly explore prior research in order to critically unpack the AI legibility problem space. Second, we respond with design proposals whose aim is to enhance the legibility, to users, of systems using AI. Third, we explore the role of design-led enquiry as a tool for critically exploring the intersection between HCI and AI research.

In this paper we report on research exploring the privacy, security and safety implications of children being able to program Internet of Things devices. We present our methodology for understanding the contexts in which children may wish to use programmable IoT, identifying risks that emerge in such contexts, and creating a set of questions that might guide design of such technologies so that they are safe for child users. We evaluate the success of the methodology, discuss the limitations of the approach, and describe future work.

In this paper we explore the motivations for, and practicalities of, incorporating "implications for adoption" into HCI research practice. Implications for adoption are speculations which may be used in research projects to scrutinize and explore the implications and requirements associated with a technology's potential adoption in the future. There is a rich tradition within the HCI community of implementing, demonstrating, and testing new interactions or technologies by building prototypes. User-centered design methods help us to develop prototypes to and move toward designs that are validated, efficient, and rewarding to use. However, these studies rarely shift their temporal focus to consider, in any significant detail, what it would mean for a technology to exist beyond its prototypical implementation, in other words how these prototypes might ultimately be adopted. Given the CHI community's increasing interest in technology-related human and social effects, the lack of attention paid to adoption represents a significant and relevant gap in current practices. It is this gap that the paper addresses and in doing so offers three contributions: (1) exploring and unpacking different notions of adoption from varying disciplinary perspectives; (2) discussing why considering adoption is relevant and useful, specifically in HCI research; (3) discussing methods for addressing this need, specifically design fiction, and understanding how utilizing these methods may provide researchers with means to better understand the myriad of nuanced, situated, and technologically-mediated relationships that innovative designs facilitate.

This paper considers how design fictions in the form of 'imaginary abstracts' can be extended into complete 'fictional papers'. Imaginary abstracts are a type of design fiction that are usually included within the content of 'real' research papers, they comprise brief accounts of fictional problem frames, prototypes, user studies and findings. Design fiction abstracts have been proposed as a means to move beyond solutionism to explore the potential societal value and consequences of new HCI concepts. In this paper we contrast the properties of imaginary abstracts, with the properties of a published paper that presents fictional research, Game of Drones. Extending the notion of imaginary abstracts so that rather than including fictional abstracts within a 'non-fiction' research paper, Game of Drones is fiction from start to finish (except for the concluding paragraph where the fictional nature of the paper is revealed). In this paper we review the scope of design fiction in HCI research before contrasting the properties of imaginary abstracts with the properties of our example fictional research paper. We argue that there are clear merits and weaknesses to both approaches, but when used tactfully and carefully fictional research papers may further empower HCI's burgeoning design discourse with compelling new methods.

The term design fiction was first used in 2005 by Bruce Sterling [18:30] and in 2009 Julian Bleecker built on the idea by combining it with various other characterisations [cf. 1,2,10] and catalysed a step change in design fiction discourse. Since then design fiction has gained significant traction across academic contexts; at symposia and conference events; and through its practice within commercial design studios and industry. Despite becoming a popular way of framing speculative design, the characterisation of design fiction as research approach still remains "up for grabs" [19:22] as it is -- enticing and provocative, yet [...] remains elusive" [7:1]. In 2013 Bleecker remarked in terms of his studios own practice "I don't think we've figured it out" and that studying it, understanding it and trying to devise some of the principles - of what we're calling design fiction - is what we're trying to do? [1]. Adopting a research through design approach [5,6], this doctoral research intends to shed light on the questions raised by Bleecker by researching design fiction, with design fiction.

This proposal is to exhibit the work named Crypto Heater which is part of a design fiction [c.f 1,5,8:30] series intended to explore a near future world in which cryptographic currencies such as Bitcoin [6] have become commonplace. This work opens up space for discussion about the activities of the distributed peer-to-peer network of so-called "miners" that ensure the security of the Bitcoin network and regulate the supply of new currency in the Bitcoin economy. The physical part of the work (the heater itself) is set within a fictional near-future reality. In this reality, Bitcoin has become central to our financial service industry, and "mining" in domestic settings is promoted by the government, as a means of heating our homes and to ensure security of the network. A "story world" is constructed using devices such as promotional materials from the UK Government's Ministry of Crypto Currency; technical specifications; customer testimonials; and the heater itself. The main element of the exhibit is a fully working Crypto Heater prototype. This device is (in the fictional world, and the real world) part of the distributed network of Bitcoin miners. Through computation, it converts electrical energy into cryptographic currency. Uniquely Crypto Heater dissipates the heat energy (a by-product of the computational effort required to be a Bitcoin miner) through a standard household radiator. By offsetting the value of the cryptographic currency produced, against the cost of electricity used, the heater provides subsidized domestic heating.