We’re well accustomed to an ongoing conversations about the ethics of AI — the ethics of using other people’s data to train AI, the ethics of using AI on academic assignments or even legal briefs — but what about the ethics of using AI to write… thank-you notes??

Hey, we’ve all been there. Maybe you just got married, or just had a baby shower, and friends gave generously. Wonderful! Except now — the task of writing all those thank-you notes! A task so very tempting to put off… Anyway, you kept good notes on who gave what, so it should be possible to do it anytime, really… Except it never gets done. Weeks later you’re looking at the same stack of notes to write.

I’ve seen people come up with some — ahem — creative ways around this problem. One couple dispatched a general newsletter accompanied by a note that just said “Thank you!” — which is considered somewhat bad form, as a thank-you note is supposed to get specific about what you appreciated about the gift. Then there’s the ever popular option of just forgetting to ever send them. Again, I’ve certainly been there, so you’ll get no judgment from me.

But as expedients go… What about having your LLM do it? Experimenting with local development with Ollama has really been bringing out my creative — not to say devious? — side lately, and I’ve been cooking up fun side projects to put my local LLMs to work. In surprisingly few lines of Python, I was able to cook up a solution that will digest a CSV of givers, gifts, and emails and spit out some creative thank-you notes. Here’s a sample of Domo’s work:

Subject: Thank You for the Wonderful Gift!

Dear Jennifer, 

Thank you so much for the incredibly thoughtful chessboard you gifted us! We were absolutely delighted to receive it and appreciate you thinking of us with such a special present.  

We can't wait to start using it. The quality is evident and the beautifully crafted pieces are stunning. We know that [mention something specific about your interests or personalities, e.g., "our family loves board games" or "we're always up for a challenge"] and this chessboard will certainly spark some friendly competition and enjoyable evenings filled with strategy and fun!

We truly value your generosity and friendship and we hope to share many exciting moments playing on it together in the coming weeks and months. 

Thank you again for such a beautiful and meaningful gift!  

Warmly,

[Your Name]
[Your email address]

As you can see, the LLM even thoughtfully leaves some space for further personalization. Now, a thank-you email is a bit gauche to begin with, and one generated by an LLM certainly pushes the ethical boundaries even further… Still, I think I’d rather get something like the above than a scattershot newsletter!

Check out the repo here if you want to give Domo a try. I promise not to judge you… Too much.