A 3-Question Framework for Making Ethical Tech Decisions

By The New Stack

Share:

Key Concepts

  • Making the Right Thing Easy: The core idea of aligning ethical practices with organizational values and making them the default, rather than a difficult add-on.
  • Scaffolding for Ethics: Establishing clear organizational values and structures to guide ethical decision-making.
  • Open Science and Data Sharing: The principles of transparency and accessibility in research, with a focus on responsible implementation.
  • Stakeholder Engagement: The importance of involving diverse perspectives (technologists, policy makers, researchers) in ethical discussions and decision-making.
  • Psychological Safety: Creating an environment where individuals feel safe to ask questions, express concerns, and admit uncertainty without fear of negative consequences.
  • Ethical Frameworks: Developing structured approaches to analyze and navigate ethical dilemmas.
  • "Both and Neither" Creative Studio: A new initiative focused on practical implementation of ethics in technology.
  • Duty to Resolve: The ethical imperative to address and resolve problems, rather than letting them persist.
  • "Share Widely and Wisely": A mantra for balancing openness with responsible data handling.
  • Premortem for Ethics: Applying the concept of a premortem (forecasting potential failures) to ethical considerations.

Data and Ethics: Making the Right Thing Easy

This summary details a conversation with Meg Derty, a fellow at the Harvard Medical School Center for Bioethics, discussing the challenges and strategies for integrating ethical considerations into technology development and data management. The discussion, stemming from her keynote at JupiterCon in San Diego, emphasizes the importance of making ethical practices the easy and default option within organizations.

Challenges in Organizational Ethics

Derty highlights that the primary challenge in ethical decision-making often stems from a definition problem: a lack of clarity on what constitutes "right" and "easy." She argues that organizational core values should serve as a scaffolding for decision-making. When this scaffolding is absent or treated as mere "hallway posters" without practical application, ethical issues arise. This lack of established structures can lead to confusion and poor decisions, especially during critical junctures or when teams are unsure of the correct course of action.

Case Study: Jupyter Notebooks in Genomic Research at NIH

A significant example discussed involved a project with 15,000 researchers using Jupyter notebooks for genomic research at the National Institutes of Health (NIH). While Jupyter is designed for sharability and embraces open science principles, the project operated within a secure environment governed by federal policies. This created a tension: while technically possible to share, patient privacy concerns and policy restrictions prevented public sharing.

The process of addressing this involved bringing together policy stakeholders and technologists to find solutions. Derty emphasized the importance of building internal momentum for ideas. The concept of a "publish" button on Jupyter notebooks was framed not just as a technical feature, but as a commitment to open science. This highlighted the interplay between technology, policy, and organizational values, prompting a self-assessment of the extent to which they were truly embodying open science principles. The mantra "share widely and wisely" was introduced as a guiding principle for balancing openness with responsible practices.

Practical Implementation and Stakeholder Engagement

Derty, drawing on her background in policy and coalition building, stressed that successful ethical integration requires more than top-down directives. Bottom-up approaches, where individuals on the ground identify problems, are crucial for large-scale digital transformation projects. Her strategy involved finding like-minded colleagues and collaboratively navigating obstacles ("brick walls").

Regarding technologists' engagement with ethics, Derty believes that while engineers naturally approach problems with a structured, "if this, then that" mindset, discussions about ethics can feel messy and overwhelming. She advocates for structuring ethics and values conversations to lower barriers and make them less intimidating. This involves creating non-punitive spaces where individuals feel safe to express uncertainty or even "say the wrong thing." The concept of psychological safety in engineering organizations, where asking "dumb" questions has no negative consequences, is directly relevant here.

The "Why is it Bad?" Framework

Derty introduced a three-question framework for navigating ethical tensions:

  1. Who is it good and why is it good? (Focuses on the positive intent and beneficiaries)
  2. Who is it for? (Defines the target audience and problem set)
  3. Why is it bad? (This is Derty's crucial addition, inspired by agile's premortem ceremony). This question encourages proactive identification of potential negative consequences and risks, even in emerging fields. It promotes working with "open eyes and open understanding" of what could go wrong, allowing for proactive risk management rather than reactive damage control.

The Nuances of "Openness"

In the context of an open-source conference, Derty addressed the complexities of "openness." Drawing from her experiences with open government data and the open internet, she noted that openness is not neutral. When data falls into the wrong hands or is overexposed, it can perpetuate harm, particularly for marginalized communities. She framed "open data" and "open science" as potentially extractive if not managed carefully.

To counter this, she highlighted the importance of a closed feedback loop in data donation relationships. This means ensuring that participants who contribute data, for example, in clinical trials, also benefit from the outcomes, such as receiving free access to new drugs developed from that data. This approach aims to manage the limitations of openness and ensure it is not purely extractive.

Educating Yourself and Building Community

For developers and technologists seeking to educate themselves on software ethics and data ethics, Derty recommends:

  • Engaging with humanities departments at research institutions (medical humanities, bioethics, philosophy).
  • Being curious and asking questions, as academics are generally eager to share their knowledge.
  • Finding colleagues who are also concerned about these issues to build connections and support networks.

To facilitate these conversations, Derty has launched a new initiative called "Both and Neither," a creative studio focused on running community workshops, hosting events, and socializing practical, easy-to-understand frameworks for implementing ethics in technology work. She is also stepping into a role at the nation's first oncology-focused venture studio, recognizing the profound ethical complexities inherent in cancer research and care.

Conclusion

The conversation underscores that ethical considerations in technology are not an afterthought but a fundamental aspect of responsible development. By establishing clear values, fostering open dialogue, employing structured frameworks like the "Why is it bad?" question, and actively engaging with diverse stakeholders, organizations can move towards making the right thing the easy thing to do, thereby mitigating harm and promoting beneficial outcomes. The initiative "Both and Neither" represents a concrete step towards operationalizing these principles.

Chat with this Video

AI-Powered

Hi! I can answer questions about this video "A 3-Question Framework for Making Ethical Tech Decisions". What would you like to know?

Chat is based on the transcript of this video and may not be 100% accurate.

Related Videos

Ready to summarize another video?

Summarize YouTube Video