Beyond Checkboxes: Building Ethics Into Tech

Recently I have been reviewing the IEEE 7000-2021 standard which provides a system engineering standard approach integrating human and social values into traditional systems engineering and design.

We’ve seen too many examples recently of how unethical programming leads to harmful outcomes. A social network’s algorithms enabling manipulation. Biased AI systems hurting disadvantaged groups. Voice assistants promoting negative stereotypes. A tracking app for elderly shoppers designed for "helpfulness" over privacy. Voice assistants taking different cultural approaches to expressing compassion

The root issue is often a failure to properly translate human values into code. Technical teams lack frameworks to intentionally design for ethical outcomes.

The new IEEE 7000 standard on value-based system design offers guidance. Key principles include:

- Involving diverse stakeholders to identify core human values at stake

- Making those values like fairness and accountability top priorities, not afterthoughts

- Translating values into technical and organizational requirements

For example, a team building a voice assistant should engage advisors on gender, culture and child development to elicit relevant values. Then they should prioritize empathy and learning over bare metrics. Finally, they derive requirements like: "Assistant responses should model compassion for users expressing distress."

IEEE 7000-2021 gives technologists tools to build ethics from the start. Our industry must move beyond compliance checklists to place human well-being at its center. Adopting value-based methods is an important step.

What do you think? How can we better inspire programming with core human values?

#ethicaltech #responsibleAI #IEEE7000 #humanvalues #stakeholders #prioritizepeople #techforgood #datascienceethics #AIethics #algorithmethics #codedwithcare #techpolicy #ethicalengineering #privacy #transparency #biasfree #inclusion #accessibility #techforsocialgood #bridgethegap