Jeff Bigham
Carnegie Mellon University
March 6, 2020
AI hype is deflating; what will remain is a rich harvest of human-centered technical work applying machine learning to important problems. This talk introduces current and past work in human-computer interaction and artificial intelligence to make the case that HCI will drive impact in intelligent systems over the near and medium term. In particular, we'll consider applications of AI in access technology, a domain that combines different utility curves for users, resulting early adoption, and yet presents use cases that have to work.
Learn more about Stanford's Human-Computer Interaction Group: [ Ссылка ]
Learn about Stanford's Graduate Certificate in HCI: [ Ссылка ]
View the full playlist: [ Ссылка ]
0:00 Introduction
2:51 Human-Computer Interaction
4:50 The Long Slog Against Ability Assumptions
8:41 What does a screen reader do?
15:14 Alt Text Today
16:06 Hiding Accessibility
19:29 First and Second Class Abilities
20:47 Designing What Users Don't Want
24:33 Discoverability
26:23 Help, my semantics became pixels
30:55 Audio CAPTCHAs and Screen Readers
33:17 People with Disabilities as Early Adopters
34:34 Machine Learning Arcs and Cycles
36:17 Designing for Ability Outliers
37:36 Dyslexia Detection
43:14 Making Memes Accessible
44:10 Real-World Devices
46:32 Real-World Touchscreens
46:39 Adding Risk-Free Exploration to Legacy Touchscreens
47:00 Blind Participant #5 Completing task with app guidance
#hci #accessibility #ai
![](https://s2.save4k.ru/pic/PJE_gnTreBo/maxresdefault.jpg)