Lots of people in the tech world think if we collect as much data as possible and run a lot of statistics, artificial intelligence will emerge organically. This author argues AI isn’t close to being intelligent and is woefully lacking common sense.

More sobriety about where things stand with AI in this fascinating conversation between MIT’s Lex Fridman and François Chollet, creator of several open source deep learning libraries.

Despite exponential growth in AI investment, Chollet says advances have been limited and linear. He also dismisses such narratives as the “intelligence explosion” and Singularity as unfounded belief systems.

I’ve long advised against using the term “immutable” in favor of “tamper evident” for blockchain protocols. This paper proposes consensus-based voting, parameterized by a policy that dictates the requirements and constraints to make on-chain redactions.

Sorting through some heavy math, “(I)f a redaction gathers enough votes the operation is performed on the chain”. No crypto required.

Last week, the FDA convened a special advisory panel to explore ways to improve patient notification and communication about medical device cyber-vulnerabilities. Not surprisingly, clinician education figured in the responses.

Here’s an excellent summary of last month’s ONC Interoperability Forum, specifically discussions about use of EHRs, from a leading researcher at The Pew Charitable Trust.

PowerPoint and infographic fatigue is real. Drawing from Edward Tufte’s seminal work, here are 10 tips about how to create effective data visualizations to promote understanding and engagement.