intersectiona11y

accessibility, ai and the future of work

intersectionality:

how identities / associations overlap and interact

intersection(a11y)

accessibility
artificial intelligence
remote work
diversity, equity, inclusion and belonging

holistic tech

rhymes with but legally distinct from

responsive design
universal design
human-centered design

empathy

applies up and down the "stack"

employees / managers
users
monetization strategy
communication (internal and external)
product design

how it applies to

accessibility
ai
remote work
deib

a11y

Accessibility, broadly, is making sure everyone can access your content and product

a11y

WCAG aims to ensure content is:

Perceivable
Operable
Understandable
Robust

perceivable multi-sensory content and interfaces. Essentially, you should still able to access the fundamental value of the content even if you cannot engage with its primary medium; the common examples here are alt text for images or captions for videos.
operable: Users must be able to operate user interface controls in multiple modalities. The most common example of this is keyboard navigability
understandable: Text needs to readable and understandable, and user interface elements should behave in predictable ways. Headers should always act like headers.
robustness, which amounts to future-proofing. Make sure you adhere to the specs so that future products that are trying to parse your content know they can do si in a coherent manner.

"it’s insignificant”

everyone is only temporarily “abled”

vision impairment
fine motor skills
finger control/touch
cognitive impairment
processing differences
fatigue/pain

artificial intelligence

not skynet
(ab)used to replace human decisions
orgs try to skip the work

generative ai is a series of next best guesses

ai is good at

pattern recognition
summarization

ai is bad at

novelty

A spotless brown giraffe against a fence

ai is bad at

novelty
being wrong
comprehension

An AI-generated picture of Spongebob Squarepants flying a plane toward the Twin Towers in New York

failure of ai and humans

the ai cannot reason, and thus does not understand what's being asked
you can't think of everything to safeguard against

other issues

training set bias
ai can only as good as the best of its data or model
short-sighted cost/benefit calculus

A Wall Street Journal headline reading 'Can AI do empathy even better than humans?'

i don't hate ai!

success with ai

make sure value consideration goes beyond monetary
ensure your data and intent are aligned
let ai advise, not decide

brief: video sharing and browsing app
market: 5.19 billion people

basic a11y features

transcript / captions
audio description track / transcript
keyboard navigable

A pie chart describing the addressable audience. 15% of people having a hearing impairment, 4% have an RSI, and 11% have a visual impairment.

w/o a11y, addressable market:

3.63 billion

beyond "need":
40% of netflix viewers use captions
seo benefits
remote interface = keyboard navigable

how can we jam in some ai?

search ✔︎
captioning ✔︎
recommendations 😬
audio description ⛔
ux ⛔

remote / hybrid work

issues:

prioritizing in-person
incentives
how work gets done

human-first approach

remote-focused meetings
invest in tools
everyone needs to communicate

diversity

deib is a culture, not an initiative
"belonging" sums up the purpose
quotas didn't create diversity,
so their removal shouldn't end it

takeaways

have empathy
take a holistic tech approach
more accessibility = more options
use ai for advice, not decisions
remote work takes work
real belonging is a culture, not a committee

thank you

questions?
hello@kait.dev

written version
https://kait.dev/intersectiona11y