Algorithms aren't neutral Making UX inclusive
summary
Shows concrete failures and fixes: voice assistants don't flag "I am being abused," Apple HealthKit omits reproductive tracking for a year, Amazon Rekognition misidentifies Black members of Congress, and Nextdoor reduces racist reports by making race descriptions mandatory which cuts completion and raises post quality. Recommends practical steps: run misuse cases, stress and regret tests, hire and trust representatives of affected groups, slow down launches, and measure whether defaults actually help the intended users.
AI-written summary · about
chapters (11)
- 0:00Introduction and speaker intro
- 1:02Defining inclusive UX
- 3:09Algorithms and interfaces aren't neutral
- 6:02Examples of exclusion in tech
- 9:40Why we build exclusionary software
- 11:06Techniques for inclusive design
- 14:39Building the right teams
- 17:05Move slower and break less
- 19:58Prioritizing who you build for
- 21:00Examples of inclusive success
- 22:29Conclusion and call to action
related talks
23mYour Algorithm isn't neutral...
27mCrouching Harm, Hidden Bias—Addressing issues of fairness....
30mFireside chat: Breaking the wall of unethical AI
42mLet's Talk about AI, ML and Bias
55mFrom observer to responsible: stopping the impact of racist biases in UX
32mDebugging the tech industry - a talk about you and me
23mMachine Learning Ethics
35mDesigning Inclusive and Accessible Experiences in the Era of AI