Algorithmic Transparency
الگورتھمی شفافیت
10 lessons · 6 hours total
- 1
What transparency means in AI: black box vs glass box
35 min
- 2
Model cards: the one-page disclosure every procured AI must ship
35 min
- 3
Bias audits: measuring fairness in the Pakistani context
38 min
- 4
Right to Information for AI: how citizens can question algorithmic decisions
35 min
- 5
NAIP procurement rules: how a Pakistani government buyer evaluates AI vendors
38 min
- 6
Data lineage: knowing where your training data came from
36 min
- 7
Incident disclosure: what to publish when an AI fails publicly
35 min
- 8
Capstone: map a real Pakistani government AI deployment against the seven prior lessons
40 min
- 9
Explainability versus accuracy: choosing where to stand
20 min
- 10
Citizen redress: building appeal channels people actually use
20 min