What’s Happiness got to do with it?
Happy World Happiness Day! Aaron Copland’s “Fanfare for the Common Man” is the incomprehensible masterpiece that emerged in my mind as I was settling down, rushing to organize the threads in my life that have led me to write these lines. Imagine the sound of Aaron Copland’s fanfares as you’re reading these words. As Terence Blanchard said in a recent NPR story about the piece: “Whenever I hear it, it stops me in my tracks and makes me reflect on the goodness of man”.
What happiness got to do with it? It surely depends on what the pronoun “it” refers to. While indigenous scholars have brought to our attention the devastating impact of habitually using language which refers to Nature as “it”, here we are concerned with “it” as in Information Technology (IT). More concretely, with the widespread use of machine learning, we assume that the IT we’re concerned with incorporates some level of automated decision making. Concerns are what afterall keeps us up at night, asking ourselves what’s happiness got to do with the way we measure the impact of artificial intelligence (AI) on the livelihood of the extraordinary common human and non-human living beings? On this International Day of Happiness, let’s live this question.
Reflections From a Recent Interdisciplinary AI Conference
The most recent ACM Fairness, Accountability, and Transparency of AI conference was about anything but happiness and I think this was precisely why it had everything to do with happiness. Above all, a key takeaway for me was the importance of bringing awareness and sharing about our work while also actively seeking out, supporting, or joining communities of practice and individuals who are also engaged in the questions we’re working to disentangle.
I think that when the emerging organizing framework behind our work is centered on solidarity, radical inclusion, transparency, and responsibility, we are better able to see and catalyze enabling connections. No matter if they are spurious correlation or causations, we need to first become aware of them in order to discern.
The plenary sessions at the conference included in-depth conversations with algorithmic auditors, community organizers, investigative journalists, regulators, and others investigating concrete issues in the Responsible AI space.
New perspectives were discussed on the use of existing metrics frameworks (leveraging administrative data for bias audits while assessing disparate coverage with mobility data for COVID-19 policy as well as the impact of bias preserving vs bias transforming metrics with respect to EU Non-Discrimination Law), designing new metrics (measuring biases in open-ended language generation, investigating the environmental impact of language models, and construct reliability and construct validity fairness metrics), bridging machine learning and mechanism design, policy toolkits for technology audits, and lessons from feminist epistemology. One of the hands-on workshops explored civic empowerment in the development and deployment of AI systems while cautioning against ‘participation-washing’. This specific session was part of the so-called CRAFT (Critiquing and Rethinking Accountability, Fairness and Transparency) conference track centered on reflection and response.
A dear friend, mentor, teacher, and collaborator Laura Musikanski describes the Happiness Moment below. See her book “The Happiness Policy Handbook: How to Make Happiness and Well-Being the Purpose of Your Government” for an in-depth discussion.
The happiness movement presents a new economic paradigm whereby the happiness of people, well-being of communities and sustainability of the planet matter most. The happiness movement prioritizes quality of life and the systems that sustain this quality, with an understanding of economic growth as one of many factors sustaining well-being. Economists Stiglitz, Sen and Fitoussi (2009) as well as Helliwell, Layard and Sachs (2012–2021) emphasise the importance of wider measures of well-being in lieu of or in addition to economic metrics for the purpose of world happiness, well-being and sustainability. Recognizing the importance of the role of metrics and data, the United Nations passed Resolution 65/309 Happiness: Towards a Holistic Approach to Development called on member nations to develop wider measures of well-being for guiding all public policy.
The term happiness and well-being are used synonymously in the happiness movement. Well-being is “people’s living conditions and quality of life today (current well-being), as well as the resources that will help to sustain people’s well-being over time (natural, economic, human and social capital)” (OECD 2018. p.2). People’s well-being is “multidimensional” (Phillips & Wong 2017, p. xxix) and includes economic, social, ecological and personal dimensions. Happiness is measured using subjective (survey, polls. questionnaires) and objective metrics.
The Intersection of AI and the Happiness Movement Paradigm Shift
As a data scientist involved in hands on AI impact assessments, I’ve been very fortunate to be learning about the intersection of AI and community well-being from the perspective of community development experts, investigative journalists, artists, and others.
Most recently, I was involved in the Springer publication Intersections of AI and Community Well-Being Special Issue for the International Journal of Community Well-Being (2020). It provides perspectives related to measurement frameworks, community engagement in AI development, and AI’s role in the protection of community well-being. One of the articles in the special issue speaks to the challenges and opportunities for incorporating community well-being metrics in news recommender systems optimization algorithms. Another contribution points out that while there has been some focus on the use of AI to interfere in national elections, the same behavior occurs at the community level, and must be acknowledged, brought to light, and addressed. Disentangling the perceived challenges experienced by specific communities such as indigenous peoples and informal caregivers, several of the authors seek to examine the ways in which AI is being developed and used currently, and point to mechanism design and methodological insights that can bring about positive outcomes contributing to community inclusion, equity and data justice. Furthermore, a few of the contributors to the special issue embark on a journey to help the readers understand the neurophysiological experience of people in public urban spaces through the use of AI, while others investigate the role of AI in improving community well-being within current models of economic growth. Finally, two contributors also suggest biomimicry as a means for the development of AI drawn from art projects involving a type of fungus commonly known as slime mold (physarum polycephalum).
The Springer Special Issue publication on the intersection of AI and community well-being has been made available for free access through the Happiness Alliance nonprofit which has been foundational in working to bridge the gaps between academic research and community engagement and well-being efforts in practice. Please feel free to access/download the articles (near the end of the page) and reach out to me if you want to learn more or contribute.
Happy International Day of Happiness!