A 3-question Framework For Making Ethical Tech Decisions

What happens erstwhile 15,000 researchers utilizing Jupyter Notebooks for genomic investigation deed a wall betwixt unfastened subject principles and national privateness policies?

For Meag Doherty, package designer and bioethics fellow, this wasn’t conscionable a method problem — it was an morals problem disguised arsenic a argumentation challenge.

In this On nan Road section of The New Stack Makers, Doherty sat down pinch TNS Editor successful Chief Heather Joslyn astatine JupyterCon successful San Diego to talk her activity launching nan National Institutes of Health’s All of Us Researcher Workbench and why making ethical decisions successful exertion shouldn’t consciousness for illustration navigating a minefield.

When Organizational Values Are Just Hallway Posters

“When things commencement to autumn isolated aliases decisions get made and teams are confused, it’s often because location could beryllium a meaning problem,” Doherty said. This happens, she continued, because galore organizations person ngo statements that beryllium only arsenic hallway posters: aspirational, but disconnected from existent decision-making.

At NIH, Doherty and her squad faced a actual example: Jupyter was built to beryllium shareable and clasp unfastened science, but they’d configured it wrong a unafraid situation constrained by national policies. Sharing was technically possible, but argumentation compliance meant they couldn’t stock publicly, chiefly for diligent privateness reasons.

“We want to stock ‘widely and wisely,’” Doherty said, quoting her erstwhile boss. Getting location meant bringing argumentation stakeholders and technologists together to research what needed to beryllium existent for sharing to hap — building soul momentum for an thought that served some unfastened subject values and diligent protection.

3 Questions to Simplify Ethical Decisions

Drawing connected her fellowship successful bioethics astatine Harvard Medical School, Doherty introduced a deceptively elemental three-question model for ethical decision-making:

1. Who is it bully for?
2. Why is it good?
3. And, astir crucially, why is it bad?

That 3rd mobility matters most. “Working successful these new, emerging fields, it’s arsenic important to region and say, wherever could this spell wrong?” she said. “That doesn’t mean you person to stop, but I want to promote group to do that pinch unfastened eyes and unfastened knowing of what could spell wrong, and past negociate against it.”

It’s nan ethical balanced of an Agile pre-mortem, and it useful because it lowers nan somesthesia astir conversations that tin consciousness overwhelming. “Talking astir morals and values tin consciousness messy,” Doherty noted. “People perceive ethics, and they think, ‘OK, she sounds for illustration [the] ineligible [department], I’m not talking to her.'”

Why ‘Open’ Isn’t Always a Good Thing

Perhaps astir provocative for a keynote speaker astatine a convention astir unfastened root technology, Doherty challenged nan presumption that openness is inherently good. “Open information tin mean extracted data,” she said. “Open subject tin consciousness extractive.”

When information gets into nan incorrect hands aliases is overexposed, information donors — who are often from marginalized communities — tin beryllium nan ones to carnivore nan brunt of harm. To antagonistic this, she suggests closed feedback loops wherever information donors spot tangible benefits, for illustration objective proceedings participants receiving free entree to narcotics developed from their contributions.

In nan aftermath of her clip astatine NIH, Doherty is launching a caller imaginative workplace called Both and Neither, to bring these frameworks to investigation package engineers and beyond to show that morals doesn’t person to beryllium 1 of nan difficult parts successful building technology.

Check retired nan afloat section to perceive much astir designing systems that are inclusive for group who aren’t successful nan room and why technologists already person morals (even if they don’t recognize it).

Group Created pinch Sketch.