Why data privacy is critical for robust data management?
summary
Argues that data is radioactive waste: collect less, expire what you store, and avoid copying raw stores by running algorithms where data lives; demonstrates practical habits — drop unnecessary columns immediately, declare needed fields at the top of notebooks, compartmentalize access, and treat temporary files as attack vectors; evaluates federated learning and anonymization as helpful but imperfect; frames tradeoffs between model quality and authentication work and urges engineering discipline rather than vague platitudes.
AI-written summary · about
chapters (4)
related talks
24mStop Making Your Data Team the 'Data Police'
48mDealing with Data Pack Rats: Reduce Data Footprints to Control Risks
43mPanel: Privacy Design Patterns for AI Systems: Threats and Protections
19mKeeping people away from data: A generative AI use case
39mPrivacy-preserving Data Sharing
39mData Privacy Engineering to power Responsible Enterprise AI
01h 59mDeep Dive Workshop – Data Security and Data Protection
16mData Ownership – Basic building block of privacy