Post 3: Data Doesn't Equal Understanding
- James W.
- 3 days ago
- 1 min read

You can have a dataset with perfect demographic representation and still build AI systems that fundamentally fail to understand what it's like to be the people in that dataset.
Data is a snapshot. Lived experience is continuous navigation of barriers, workarounds, context.
An image dataset of wheelchair users navigating buildings doesn't capture:
The difference between "technically accessible" and "actually usable"
The cumulative effect of multiple barriers
The energy cost of workarounds
Unwritten social rules that exclude despite technical access
Genuine inclusion requires centering disabled people and marginalized communities directly in design governance—not as consultants, but as decision-makers.
Data representation is not lived experience inclusion.

Comments