And then articulates what companies need to do to actually benefit from AI. One point stood out to me: to paraphrase, the author noted it's important to design the right thing.
AI does not do this well right now. It'll help you build anything. Messily. But whether you should build that thing? To quote Claude: "That's a brilliant idea, Amy!" But, is it? Claude and ChatGPT will tell you anything is a great idea.
Before you get AI off and running on building an app or an agent, you should get clear on the customer or employee needs you're trying to address and what you hope to learn. Ask: what can we do today that we couldn't do a year ago? (An AI question). But also ask: who is it serving and how does it make their life better? (A design question.)
In reality, it's our job to know what is the right thing to build. Which is exactly why Light Actions are muscles we all need to be building.
So try this on Monday:
- Pick a problem you're facing in your organization
- Ask: what can we do today to address this that wasn't possible a year ago?
- Then ask: what's a small and safe way we can learn more?
Try the Light Actions Companion (beta alert!) to help you with the latter.
