~/talk

Operationalizing Responsible AI

on YouTube 31m
Talkat TWIMLcon: AI Platforms 2019 · 2019-10-01..10-02· [machine-learning]

speakers

name:Rachel Thomas
links:none on file
name:Guillaume Saint-Jacques
links:none on file
name:Parinaz Sobhani
links:none on file
name:Khari Johnson
links:none on file

resources

video:https://youtu.be/A4J1UGHFaUk
slides:none on file

summary

Teams describe concrete tooling and outcomes: use TensorFlow Privacy, differential privacy and federated learning for data protection, and interpretability tools like LIME and the What If Tool for debugging. Fairness testing with FairTest uncovers higher false positives against non‑native speakers in a plagiarism detector and the bias is mitigated. LinkedIn builds a representative ranker and an instant job notification feature that surprisingly equalizes application rates across groups. Teams insist on guardrails, error plans and system‑level outcome monitoring.

AI-written summary · about

chapters (12)

  1. 0:00Introduction and panel setup
  2. 1:27Icebreaker and panelist introductions
  3. 4:22First steps in operationalizing AI ethics
  4. 7:30Frameworks and tools for responsible AI
  5. 10:38Biggest concerns in responsible AI today
  6. 14:51Scaling responsible AI in large organizations
  7. 17:17Responsible AI in startups
  8. 21:05Power dynamics in AI ethics conversations
  9. 23:13Success stories in AI ethics
  10. 26:46Audience Q&A begins
  11. 27:07Principled approaches in deep learning
  12. 30:41Closing remarks and summary

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.