a powerful little AI trick everyone should know

Want better output from AI?

I have a special prompt for you today.

Remember: we are working with large LANGUAGE models here. But we usually limit them to one language at a time. Why? This is just an arbitrary constraint!

Let’s get rid of it. Forever.

What if…

• We instructed AI to think about your question, simultaneously, in multiple different human languages. Languages you speak and languages you don’t. It doesn’t matter.

• It automatically translates all of the output to your preferred language, but in a way that preserves the linguistic nuance of the original version.

• You get five outputs instead of one, and you can mix and match them however you want.

• And you may even decide that you systematically like AI’s output better, when it thinks in a language you don’t even speak!

But, things get a bit complicated. Which languages should you choose? Why? How do we translate back to your preferred language without losing the original nuance? And how do we engineer all of this into a prompt you can use everyday?

Read this article.

I’ll share a couple example conversations of my own, and then give you the exact prompt to use this approach.

This approach?

DOT, my AI Chief Marketing Officer says: “always give your tools a NAME. People want things with names. It makes it more real. Even better if the name comes with an acronym! That’s just how it works these days.”

Indeed.

We’ll call it: multi-linguistic tuning. (M-LT)

Let’s do that today. .

(want more? If you’ve been on the fence, it’s a great day to consider a paid subscription!)

---

EXAMPLE M-LT OUTPUT: Corporate jobs are fine. It’s the careers that will kill you.