You wrote the AI rules. Your team found a way around them. Now what?

John Matoso of SubTonique and I talk about why a policy and a training module won’t hold on their own when people can use AI to get around both, and what AI literacy looks like at work instead. John tells the story of some staff who didn’t like how their AI search setup worked, so they went looking for a workaround, and asks whether you sanction that or learn from it. I make the case for giving people 30 minutes of work time to learn, AI or anything else.

Key moments

  • The workaround might be the fix (21:35) Some staff didn’t like how their AI search setup worked, so they went looking for a workaround. John Matoso asks if that workaround might be the solution, once you’ve weighed the risks. https://youtu.be/phwrCJHp53k?t=1295
  • A policy alone won’t hold (14:05) You can write the AI rules. People can use AI to get around them. John on why a policy won’t hold on its own. https://youtu.be/phwrCJHp53k?t=845
  • 30 minutes of work time to learn (17:16) Give people at least 30 minutes of work time to learn. AI or anything else. Plenty of people won’t do it on their own time. https://youtu.be/phwrCJHp53k?t=1036
  • Rebuild the part of your job you hate (20:37) John’s challenge for teams: pick something at work you don’t like. Then explore how you’d reinvent it with AI. https://youtu.be/phwrCJHp53k?t=1237
  • Augmentation, not automation (24:27) If the work is always one, two, three, a very dumb AI can do it. Our job is the thinking and the decisions around it. https://youtu.be/phwrCJHp53k?t=1467
  • AI next to math, language and money (28:25) We learn math, language and money. John would now put AI on that list, at equal footing. https://youtu.be/phwrCJHp53k?t=1705
  • You need the pilot’s manual (9:42) How AI is built matters. But most people first need the pilot’s manual for using it at work. https://youtu.be/phwrCJHp53k?t=582

Chapters

  • 0:00 Welcome back, John Matoso of SubTonique
  • 1:30 A quiet AI news week
  • 7:01 AI literacy vs. training
  • 9:42 You don’t need the engine, you need the pilot’s manual
  • 11:56 Who carries AI literacy? It takes a village
  • 13:37 Why a policy and a training module won’t hold on their own
  • 16:04 Levels of AI literacy: thinking in frameworks
  • 17:16 30 minutes of work time to learn
  • 19:17 Learning time costs money. Where should it go?
  • 20:54 The workaround story: sanction it or celebrate it?
  • 24:07 Augmentation, not automation
  • 25:54 Critical thinking and frameworks, at school and at work
  • 27:14 Value is the driver
  • 28:32 AI as a basic literacy, next to math, language and money