How to Improve UX with Machine Learning: Insights and learnings until now
summary
Machine learning fits problems that are easier to show with examples than to write rules for, and is useful for recommendations, clustering and automating tests but brings imprecision and a measurable margin of error (autonomous car failures are a concrete risk). Practical notes: supervised, unsupervised, reinforcement map to labeled prediction, clustering/association, and reward-driven learning; adding an intelligence layer can auto-optimize design system components and run A/B cycles; guard against privacy by design failures and hyper-personalization filter bubbles.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker background
- 5:29Why ML design differs from traditional UX
- 13:59Types of machine learning approaches
- 20:23When to use or avoid machine learning
- 23:40Hyper-personalization in UX design
- 31:14Anticipatory design and its challenges
- 36:30Privacy by design in ML systems
- 41:15Future of ML in UX and responsibilities
related talks
43mUX and the Changing Relationship Between Mind and Machine
26mHow Machine Learning Can Help
49mUsing Machine Learning and Not Dying Trying
38mDesigning Machine Learning for Humans
23mUser Experience Now Includes Machine Experience
21mDemystifying Machine Learning for Programmatic UA & Retargeting
49mCXO Panel: Cybersecurity: Is Machine Learning The Answer?
25mDesign for Context with Machine Learning