In this episode of the Cyberlaw Podcast, I take advantage of Scott Shapiro's participation to interview him about his book, Fancy Bear Goes Phishing – The Dark History of the Information Age, in Five Extraordinary Hacks. It's a remarkable tutorial on cybersecurity, told through stories that you may think you've already heard until you see what Scott has turned up by digging into historical and legal records. We cover the Morris worm, the Paris Hilton hack, and the earliest Bulgarian virus writer's nemesis. Along the way, we share views about the refreshing emergence of a well-paid profession largely free of the credentialism that infects so much of the American economy. In keeping with the rest of the episode, I ask Bing Image Creator to generate alternative artwork for the book. 
In the news roundup, Michael Ellis walks us through the "sweeping" 
White House executive order on artificial intelligence. The tl;dr: the order may or may not actually have real impact on the field. The same can probably be said of the advice now being dispensed by AI's "godfathers."—the keepers of the flame for AI existential risk who have urged that AI companies devote a third of their R&D budgets to AI safety and security and accept liability for serious harm. Scott and I puzzle over how dangerous AI can be when even the most advanced engines can only do multiplication successfully 85% of the time. Along the way, we evaluate methods for poisoning training data and their utility for helping starving artists get paid when their work is repurposed by AI.
Speaking of AI regulation, Nick Weaver offers a real-life example: the California DMV's immediate suspension of Cruise's robotaxi permit after a serious accident that the company handled poorly.
