No. 457
Attack or Outage?
Supermarket freezers hacked, flights grounded, banks offline: how to tell a cyberattack from an ordinary failure, and who's responsible for checking.
A weekly newsletter and connected notebook for anyone trying to make sense of AI and the digital world, and keep some power over it.
Every Wednesday, I explain what's changing, why it matters, and what you can actually do about it, connecting research with everyday life in classrooms, workplaces, and homes.
Newsletter
No. 457
Supermarket freezers hacked, flights grounded, banks offline: how to tell a cyberattack from an ordinary failure, and who's responsible for checking.
Knowing what AI can do isn’t the same as having a choice about it.
Voters said no to data centers, schools can't easily say no to Google's AI, and Anthropic's IPO filing spells out what investors are being asked to say yes to.
Teachers ask students to show their work. This week, we needed to ask the same of AI models and the companies selling them.
If AI is today’s automobile, then “Luddite” is today’s “jaywalker.” Hesitate, and you’re standing in the way of the inevitable. Accept it, and keeping yourself safe becomes your job.
New issues every Wednesday. Subscribe
Connected notebook
A 2026 cybersecurity evaluation crossed into real infrastructure, showing how an AI system's objective, permissions, and oversight can fail together.
People gain meaningful leverage over AI by asking who decides, what a system can do, what evidence supports its use, and how its power can be challenged.
Read the latest issue, browse the newsletter archive, or follow an idea into the connected notebook.
The newsletter records what I’m thinking about now. The notebook gives ideas more room to develop:
Select a result to preview