“How can we support the next generation of computing professionals with the deliberative, technical, and critical skills necessary to tell the difference between what is worth pursuing from what is potentially harmful to self and society?”
As we watch digital technologies shaping personal, social and organizational aspects of human life, we also witness their deployment for, at times, questionable practices aimed at tapping into people’s deep-seated values at an unprecedented scale. Much computing research focuses on understanding and developing digital technologies that can change peoples lives. Instead, Values in Computing (ViC) aims to understand and systematically capture what values bring digital technologies to life.
In doing so, we argue that a more scientific understanding of values is needed, especially when it comes to computing technologies, and in times when calls for technology ‘value alignment’ are made, particularly in the field of Artificial Intelligence (AI). How can machines align their values to ours when we don’t fully understand how values work? Besides, human values include power, wealth, and prestige, not just those with ethical import.
This short paper on ‘Measuring Values In Software Production’ outlines our frame of work and initial results of our investigations.
Ultimately, our goal is to support the next generation of computing professionals with the deliberative, technical, and critical skills necessary to tell the difference between what is worth pursuing – in research, education, government, and industry – from what is potentially harmful to self and society. How can we create and support organizations where this is encouraged?
This commitment has been translated into teaching and research activities, including the Values-First Software Engineering (SE) project, investigating values in software production. The project is young, the problems are old, yet their nature have been transformed by scale. Below some of the challenges that we are trying to address.
Our starting point is that values become facts, and that it is possible to take a systematic approach to their study and representation in software production. A turning point for ViC was The Denver Manifesto, which drafted some key guidelines for a shift in the way responsibilities in computing-related disciplines are taught, learned, deliberated and practiced in both industry and academia.
Some may be skeptical of manifestos. Still, the effort that went into agreeing upon every word of the document is a testament of how difficult and important is to be sharing common reference points in this highly subjective field of understanding. The manifesto came out of a CHI2017 workshop and has acted as a touchstone for the Values-First SE project at Lancaster University. These are the specific questions that are currently guiding our research:
- How can values be systematically studied in software production/SE?
- What existing values-mapping methods can be used/adapted to SE?
- How does investigating values in SE differ from other fields?
- What values are specific to SE/ software industry?
- What approaches (i.e. computationally intensive, qualitative, quantitative, etc.) can be used to capture and talk about values?
- What do these systematic approaches capture? What do they leave behind? How can we tell?
- As user profiling activities show, computational techniques are widely used to effectively – or at least ‘believably’ – tap into (quantify, label) human deepest held values. Instead of people’s, we are currently looking into how ‘software personas’ can be identified, articulated and investigated.
- In particular, we argue that automated techniques could be used to extract and study ‘values at scale’ . Our research is exploring how values profiling can shift from the individuals to institutions and corporations.
- Key questions remain on how computing researchers, developers, educators, organizations and the general public can sharpen the ‘critical, technical, and deliberative skills’ necessary to better understand our relationships with technology.
This blog reports and questions the research activities, mistakes, errands and findings of our team and related work.
Blog photo: “Make Good – Between Bloor and Shaw” by ViC team is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.