- 2017, August 9, InnovationAus, “Keeping it real with CryptoPhoto“
- 2017, August 9, ZDNet, “True AI cannot be developed until the ‘brain code’ has been cracked: Starmind“
- 2017, August 8, MIT Technology Review, “Andrew Ng’s Next Trick: Training a Million AI Experts“
- 2017, August 8, Google, “Getting ready for Europe’s new data protection rules“
- 2017, August 8, Bird & Bird, “The UK Government publishes its Statement of Intent for Data Protection Bill and GDPR“
- 2017, August 8, The Conversation, “How artificial intelligence conquered democracy“
- 2017, August 8, LSE, “The Great Data Protection Rebranding Exercise“
- 2017, August 6, arXiv – 1708.01884, “Differential Privacy By Sampling“
- 2017, August 5, arXiv – 1708.01732, “Forensic Analysis of Epic Privacy Browser on Windows Operating Systems“
- 2017, July 29, arXiv – 1708.01870, “Challenges for Transparency” — some interesting points. I think the examples used to point out where transparency could lead to ‘bad’ outcomes are not cases against transparency per se, but highlight that sometimes access to the reasoning of the model cannot be widely accessible. In these cases the person affected by the output of a model should still have a means of access and remediation if the case is not strong. Toby Walsh makes a good point in his latest book (quote is near the end of Chapter 1 – I have no idea how to cite a Kindle ebook!) that there are instances where accessibility to the underlying processing of the model is more important than accuracy; I think he is saying that deep learning cannot be a panacea for predictive modelling due to its opacity.
- 2017, July 26, arXiv – 1708.01643, “Ensuring patients privacy in a cryptographic-based-electronic health records using bio-cryptography” — I like the idea presented in this paper – human biometrics as a cryptographic key. It highlights the importance of key management.
HBO Can’t Catch a Break!
I look forward to see how Andrew Ng has re-worked his Coursera Deep Learning offering.
Where have you been all my life? This is a great framework – appropriate mix of business and technology treatment. I just joined ISACA and look forward to working through the COBIT standard and its implementation.