1 Introduction
The catastrophic consequences of climate change are becoming ever more apparent [
63]. World leaders have committed to the Paris Agreement: a guiding framework for nations to ensure a maximum of 1.5
\(^{\circ }\) C of global warming. Yet meeting its targets requires significant emissions reductions both immediately and into the future—specifically, global net zero emissions by 2050 at the latest. To meet Paris targets, nations will require that
all sectors in the global economy reduce their carbon footprint—with implications for
Information Communication Technology (ICT), both in terms of its own climate footprint [
33], and its possible role in decarbonisation of other sectors [
38]. This raises important questions for the ICT sector: Should ICT be allowed to grow and innovate freely for its stated (but largely unevidenced) role in addressing climate challenges? If not, then what innovation can we afford, for what value to society, and who gets to decide?
ICT devices, data centres and networks all have an environmental impact through the energy they use directly but also significantly via the embodied emissions of producing the technology and its end of life disposal. Estimations of the total carbon footprint of ICT vary between experts [
2,
8,
57] from 1.8% to 2.8% of global greenhouse gas emissions [
33]. Some suggest this impact could increase further to 4–5% of global emissions [
1], yet more optimistic projections suggest emissions could instead be halved by 2030 [
56]. Despite these conflicting estimates, historic patterns show ICT’s emissions have continuously risen, and all these experts believe ICT’s environmental impact will not reduce without a concerted political and industrial effort [
33]. If we assume ICT’s impacts will simply stabilise, then the ICT sector would still need to reduce its carbon footprint by 42% by 2030, 72% by 2040, and 91% by 2050 to be aligned with the Paris accord, and while increases in energy efficiency and adoption of renewable energy sources will help this endeavour, they alone could not be relied upon to enable such extensive emissions reductions [
33]. In fact, recent work calls for a constraint on ICT’s emissions to avoid efficiency-based solutions from currently increasing ICT demand and thus offsetting emissions savings or, in the worst case, increasing ICT’s emissions overall [
85].
ICT’s environmental impacts are well known in research communities such as
Sustainable Human-Computer Interaction (SHCI), with efforts looking to expose these impacts and reduce their severity through interaction design [
4,
12,
13,
69,
83]. SHCI work has also often focused on how ICT can promote sustainability [
27,
37,
50], yet such design research has not supported anywhere near the vast carbon savings we need to address the climate emergency [
48] despite calls for more radical, activist, and political approaches [
29,
49,
79]. While we would not of course suggest it is the role of SHCI or even ICT alone to address such a global challenge, we argue it is no longer acceptable simply to adapt designs to be “more sustainable” while tolerating continued technology growth. Rather, this trajectory needs to be
reversed, with the ICT sector needing to make stark decisions on what (limited) ICT innovation it should prioritise under climate constraints. Thus, we instead propose using HCI expertise as a mechanism to expose tensions in prioritising ICT collaboratively and offer ways forward for HCI to support the ICT sector in resolving these tensions through values-based and target-led ICT innovation for a sustainable future.
In this article, we use an exploratory participatory exercise to begin to explore some of these tensions—uncovering influences, interpretations, and considerations for priority decision-making in ICT innovation.
1 Through this, we uncover what evidence, context, and digital tools the sector may require to be able to make collective choices regarding which valued applications of ICT to continue—
instead of other valued areas of ICT—in a carbon-constrained world. Our goal is
not to arrive at a concrete prioritisation for global ICT innovation or to hint at areas undeserving of their carbon impacts. Rather, our goal is to observe participants’ decision-making processes to
expose the thorny challenges entailed in such decision-making. We discuss issues regarding who, what, where, and how we should make decisions to effectively prioritise ICT innovation collectively and outline implications and future HCI research to support the ICT sector in exposing and resolving tensions for prioritising ICT innovation in line with Paris targets.
2 Related Work
ICT research communities have formed various and interlinking pathways that address environmental challenges and ICT’s role within this. Examples include considering sustainability within the development of ICT through sustainable software engineering (e.g., References [
7,
73]), developing methodologies and assessments for the carbon impacts of ICT (e.g., References [
9,
23]) (i.e., green ICT literature [
44]), and using ICT for responding to challenges in environmental science through environmental informatics [
34,
44,
68]. The SHCI community has also been a significantly contributing domain to the topic of ICT and environmental sustainability since its establishment in 2007. SHCI takes a more socio-technical and interdisciplinary approach to sustainability, exploring the interactions between humans, ICT, the environment, and the responsibility of ICT design for sustainable futures; given this focus and its methodological approaches, SHCI provides an effective framing for exploring tensions in collaboratively prioritising ICT.
As an overview of this area, SHCI researchers have been investigating how ICT can be designed to minimise its environment impacts (sustainability
in design [
58]) or be applied to make other activities more sustainable (sustainability
through design [
58]). For sustainability in design, research has explored the environmental impact of ICT [
4,
55,
70,
83], addressing device obsolescence [
71] and offering principles for sustainable interaction design [
12,
13]. In sustainability through design, research has included the following: making resource consumption more visible to users with the hope of reducing this consumption, e.g., for energy and water [
52,
76]; utilising dystopian Virtual Reality to engage citizens with climate change and promote pro-environmental behaviour [
31]; exploring the role of contextual food data [
54] and self-tracking technologies [
22] to encourage sustainable food practices; and the development of information systems for sustainable communities, with calls for such services to better embody these communities’ values within their design [
62].
While this work has merit, there have been several critiques of SHCI that suggest a need to broaden its focus beyond individual behaviour [
17,
48], with Knowles et al. highlighting that SHCI focuses too much on
“minimal impact” solutions that fall far short of what is required by the climate emergency [
49, p. 3593]. Rethinking SHCI, Baumer and Silberman [
6] propose that sometimes it may be best
not to design, e.g., if a non-tech solution is more appropriate or if the ICT does more harm than good. Similarly, Pierce [
66] introduced the concept of
undesign, suggesting that there is opportunity for HCI designers to negate ICT by its own design. Moreover, while there have been suggestions for SHCI and “ICT for Development” communities to create ICT that help people prepare and adapt to political, social, or ecological change (e.g., from climate change) through collapse informatics [
80], there have been brief mentions of
removing ICT to prepare people for collapse scenarios where the ICT may no longer be able to function [
81].
Adding to these critiques, Hansson et al. [
43] have recently investigated how SHCI research maps to the United Nations’
Sustainable Development Goals (SDGs) [
61]—finding that for the majority of papers that
do map to the SDGs, they focus on SDG 12, “Responsible Consumption and Production,” and only 2 of their total 71 paper corpus linked to the SDG, “Climate Action.” This is despite the urgency of the climate crisis and calls for SHCI to orient around climate change [
48]. Thomas et al. [
79] also point out that SHCI does not directly respond to climate change policies or goals, yet ICT4S research by Pargman et al. [
65, p. 293] argues that the
“concrete nature” of carbon emission reduction targets (specifically the “carbon law”) can aid a trajectory for any sector, society, or time span toward Paris Agreement goals. From recently reviewing SHCI research, Bremer et al. [
16] suggest a shift in focus for SHCI toward
“Green Policy Informatics,” enabling the community to develop digital systems that support climate policies.
In this article, we directly respond to SHCI’s critiques and ICT’s role in the climate crisis and explore a new way in which HCI can promote political action, shape computing outputs, and support climate policies (cf. References [
16,
48]): By using a HCI-based exercise to expose the tensions and considerations in collaboratively making decisions on priorities for ICT innovation. With this as a focus, we do not aim to establish prioritisation decisions on ICT innovation but rather explore the role of HCI in exposing tensions in such decision-making that the ICT sector needs and uncover how HCI can support the ICT sector in resolving these tensions as it aligns with the Paris Agreement. While prioritisation has been investigated for broader energy research (e.g., identifying priority research questions for energy in cities [
59] or UK energy system changes to support low-carbon business models [
42]), to the best of our knowledge, this is the first study that considers the topic of prioritising innovation in ICT, what tensions exist in such decision-making, and specifically HCI’s role in exposing and resolving these tensions for ICT’s sustainable future.
3 Method and Participants
We designed an exploratory participatory HCI study to investigate how ICT researchers and technologists prioritise ICT innovation as a means of prompting structured reflection about the tensions that the ICT sector would need to consider in making priority decisions regarding strategic cuts to its carbon footprint.
2 This involved 20 ICT researchers and technologists
3 engaging in a “prioritisation exercise” in pair-based interviews via Teams. By offering this practical exercise as experience of prioritising ICT innovation, we were able to uncover the tensions in such processes and discuss these in-depth with the participants themselves through the interview questions. The study was ethically approved and participants were informed that the study was motivated by environmental sustainability, noting the need to mitigate ICT’s carbon footprint and to understand what valued ICT innovation people believe we should prioritise and continue under constraints on ICT growth.
Each participant was asked to answer a pre-interview demographic survey covering their age, gender, employment status, highest education level, primary sector of work (academia, industry, or equally both) and their familiarity or past experience of working with notable ICT trends. Participants were recruited via email advertisement and snowballing methods, were required to work or have worked as an ICT researcher or technologist, and were offered a £10 voucher. Studies were conducted in September–December 2020
4 at a time to suit participants. The participants are summarised in Table
1—ranging from 20s–50s in age, 50% female and 50% male,
5 with a variety of industry or academic experience. Participants were well educated, with 14/20 holding a postgraduate degree and all having engaged in further education; this was not explicitly a design of the study recruitment but as a result of snowball sampling.
The interview involved an ice breaker exercise for the participants to introduce themselves and denote what they believe are ICT’s key strengths and weaknesses,
6 the “prioritisation exercise,” followed by a round of discussion questions. The prioritisation exercise involved displaying a grid via video link of 12 broad categories of where ICT is applied: Natural Sciences; Energy; Manufacturing; Logistics and Transport; Society and Community; Education and Skills Development; Health, Medicine and Psychology; Security, Privacy and Defence; Government, Policy and Law; Business and Commerce; Entertainment and Fitness; and Robotics. The researcher provided participants with examples of the type of ICT innovation that may occur in each category, noting that ICT may be used to enhance these categories (e.g., for “Natural Sciences,” using ICT to monitor the natural environment) or that these categories may be used to enhance ICT (e.g., for “Natural Sciences,” creating new materials for ICT development).
Participants were then asked to individually allocate 12 “tokens” to represent what categories of ICT application they believe should be prioritised, exploring what areas of ICT application they deem to be most important and valuable to continue under constraints on ICT innovation. Tokens were used to broadly represent resources put toward those categories (e.g., money, environmental, or workforce resources); 12 tokens were provided so at least 1 token could be allocated to each category, but participants were able to distribute these as they wished (e.g., three on four categories). The participants were asked to describe how they allocated the tokens; the researcher placed their tokens on the grid.
To simulate the requirement to ratchet down emissions over time in line with increasingly ambitious interim targets, the participants were then asked to allocate 12 tokens as a group—discussing their decisions throughout the task. This was then repeated with a depleting number of tokens: nine tokens (75% of total), 6 tokens (50%), 3 tokens (25%), and, finally, one token to identify their most prioritised application domain. Like the individual exercise, participants were prioritising what ICT they deem most important and valuable to continue under constraints on ICT innovation.
The researcher observed the study, taking frequent photos of the application categories grid to record the participants’ priorities, as well as answering any questions or prompting engagement. Discussion questions followed around feedback on the study alongside the participants’ thoughts on how the ICT sector could make collective priority decisions.
Following each interview, the researcher textually recorded the priority decisions from the photos to visualise how priorities changed during the exercise and compared to other groups; and fully transcribed and analysed the audio recordings, using thematic analysis (cf. Reference [
15]). Themes were re-coded as each interview was conducted until findings converged; final themes were discussed with another researcher to ensure the coding was an appropriate representation of how participants approached the decision-making, as well as the tensions and considerations in their prioritisation.
A note on categories used in the prioritisation exercise: The 12 categories of ICT application were created by a preliminary analysis of data from the UKRI’s GtR,
7 a publicly available database containing information about UK Government funded research. This specifically involved analysing 102 digital technology projects associated with major trends that have been noted for their significant energy impact [
33]: Data Science and
Artificial Intelligence (AI), the
Internet of Things (IoT), and Blockchain. We recognise there are limitations in asking our participants to prioritise global ICT innovation based on analysis from a subset of UK-funded projects; however, our focus was on understanding how participants approach decisions rather than the prioritisation outcomes made by participants. We are not making any claims for global ICT prioritisation based on the participant decisions, and the exploratory method is instead used to uncover complexities in decision-making for ICT innovation. In that notion, we could have perhaps even replaced these 12 with a different set of categories, leading to different decision outcomes yet similar challenges for decision-making approaches and influences and HCI’s role in ICT prioritisation. Nevertheless, we provide all details regarding the analysis and formation for the 12 categories as supplementary material for transparency.
7 Conclusion
As SHCI researchers, we recognise achieving positive impact on ICT’s environmental sustainability from the vantage point of HCI is challenging. Not only do we suggest designs that go against traditional HCI narratives, but we also critique our own work for having a too narrow focus (cf. Reference [
17]) and not creating enough impact [
49]. From our prioritisation exercise and discussions with ICT stakeholders, we propose a new area of research for HCI to lead on for responsible and sustainable ICT, comprising of urgently compiling a socio-technical evidence base for the ICT sector to make priority decisions on, understanding how the ICT sector can make priority decisions and societal tradeoffs in a global context,
and designing transparent digital tools that aid the ICT sector in making priority decisions on its innovation in the increasingly limited time there is to meet global climate goals. Through this, we envision the HCI community can utilise their socio-technical expertise and expand its focus with experts in responsible computing to help the ICT sector traverse a new pathway for sustainability, differing vastly from the “ICT status quo” of continuous innovation and growth.
While the stakes are high, we wish to remind readers that the HCI community need not be alone in this: our implications call for transdisciplinary research that encompasses ICT stakeholder involvement, and we simply propose the HCI community is well placed to glue relevant research and associates together. By utilising the socio-technical expertise that HCI embodies, the HCI and responsible computing communities can support the ICT sector in priority decision-making (as highlighted in our implications for next steps in this domain), as well as build upon what we have started through our study to expose further tensions of ICT innovation prioritisation amongst different stakeholders and under varied criteria. As activists, educators and professionals, such communities can then utilise this work to influence others in the ICT sector—particularly industry and academic leaders, as well as policymakers, who can hold the sector accountable for ICT’s environmental impacts.
Given the ICT sector is broad in the innovation it provides, we suggest this avenue for responsible and sustainable ICT should initially focus on supporting the ICT sector in prioritising a subset of ICT innovation under constraints: Freitag et al. [
33] found that the growth and environmental impacts of AI, IoT, and Blockchain (e.g., for AI training [
11], for IoT’s vast expected growth [
75], Bitcoin as an application of Blockchain [
25]) could have profound effects on ICT’s future emissions. This work can leverage and extend research examining the ethical issues in regards to these trends in ICT, such as guidelines of AI [
46], challenges for IoT [
3] and issues of Blockchain [
77]. Focusing on these trends should expose the key tensions for prioritising innovation in the ICT sector as a whole and support the ICT sector to make effective priorities that enable significant inroads in reducing ICT’s carbon footprint to align with the Paris Agreement.