Ever thought that a little ‘please’ and ‘thank you’ might do more than just brighten someone’s day? Murray Shanahan, a senior researcher at Google DeepMind, suggests that these small courtesies could actually influence the way language models respond. He points out that when models perceive a polite request, they tend to offer more considered replies.
Imagine an AI that sees itself in the role of a clever intern—friendly, responsive, and eager to help. With polite wording, the interaction feels less mechanical and more like a genuine conversation with a helpful colleague. While the benefits aren’t uniform across all scenarios and models, embracing a courteous tone often leads to more engaging responses.
If you’ve ever struggled with getting the best out of your AI tools, it might be worth trying a softer approach. After all, a touch of humanity goes a long way—even in the digital world.
