Sponsored Session: The Search for Transparency and Accountability in the Age of AI: RAI & XAI as Essential Tools
Talkat Open Source Summit Europe 2023 · on Thursday, 2023-09-21· at 11:00· Room 0D-2-0D-3 (Floor 0), Abandoibarra Etorb., 4, 48011 Bilbao, Bizkaia, Spain·
summary
Summarizes responsible AI as four concrete requirements — fairness, transparency, accountability, privacy — and shows actual tooling: uses IBM AI Fairness 360 and Google What‑If for bias checks, highlights privacy techniques like PPML and OpenFL, and demos Intel Explainable AI toolkit that wraps SHAP/LIME to explain TensorFlow, scikit‑learn and BERT models (heart‑disease classifier and text sentiment). Presents a user study on a Merlin bird app comparing heatmaps, example‑based, concept‑based and prototype explanations, finds a large gap between high‑AI and low‑AI users, and argues XAI must answer why, not just what.
AI-written summary · about
chapters (8)
- 0:00Responsible explainable AI
- 5:43Responsible AI development tools
- 12:31Explainability perspectives
- 16:23Model explainability challenges
- 22:23Centralized explainable AI tools
- 26:41Human-centered explainability
- 32:07Principles to responsible AI
- 36:05Q&A
related talks
39mThe Search for Transparency and Accountability in the Age of...
37mTrusted and Responsible AI (Explainability, Adversarial, Bias, Fairness)
37mBuilding a case for Fair and bias-free AI
39mResponsible AI
19mA developer's guide to responsible AI review processes
34mResponsible AI: The Key to Ethical, Safe and Inclusive Software Development
36mKeynote: Ethics in AI: Building Fair and Just Intelligence Systems
15mResponsible AI 2.0 : NO AI without RAI