If you are like me, you are probably still trying to navigate how you feel about AI. What human spaces should it occupy? Where can it be truly helpful and increase our productivity, and where should we be fighting to protect things that feel uniquely human?
I think one of the biggest questions of the next couple of years will be what things AI can not, or should not do for us. What skills do humans hone in an age of AI, and what skills are worth developing?
If you are interested in this conversation, you’ve probably heard the AI labs talk about the importance of “taste.” What they mean by this is discernment. A human judgment of what is good, interesting, thoughtful, or resonant. Their basic argument is that AI lowers the cost of execution, and so humans will differentiate through judgment, point of view, and an eye for quality that only we can provide.
To the extent that I buy this argument, I realized I haven’t spent much time actively thinking about my own “taste” as someone who is not in a traditionally creative field. I know when something feels thoughtful or differentiated, but I don’t always know how to articulate the pattern behind it.
Earlier this year, I built something as a bit of a joke in my household. It’s a gift guide, an AI project to help my husband buy gifts I’m actually going to love. I trained it on things I already loved, then played around asking it to recommend gifts, rank options, and explain why certain things resonated with me. It was a fun project and worked surprisingly well.
At the time, it felt a bit superficial, which is why I never suggested it as a challenge. But this week, after thinking more about taste, judgment, and discernment, I started wondering whether something similar could help me better understand my own point of view and my “taste”.
Before we start: if you are getting this email forwarded to you and are not yet part of the lab - you can join the tiny AI lab here.
The Challenge:
Build yourself a personal taste curator.
The goal of this project is to create an AI system that helps you better understand your taste, put words to what you intuitively identify as “good,” and uncover patterns in your own preferences.
You will give your AI examples of things you like. You can keep this work-related and share peers in your industry you admire, websites you like, or marketing launches you thought were incredible. Or you can make it more personal and share music, design, fashion, travel, or anything else you are drawn to (or both)
Then use AI to identify patterns behind your taste and recommend other things you may enjoy based on those patterns.
The prompt to get you started
Drop this into your AI tool of choice:
I want you to act as my personal taste curator.
I am trying to better understand my taste and identify the deeper patterns behind the things I consistently admire across work, products, experiences, aesthetics, and people.
I am going to share examples of things I like, admire, or find compelling. Your job is to identify the patterns, values, emotional qualities, and standards that connect these examples.
As you analyze my examples:
explain what I consistently associate with quality
help articulate my point of view more clearly
identify contradictions or tensions in my preferences
recommend adjacent ideas, products, creators, experiences, and companies I may genuinely love
As our conversation continues, continuously refine your understanding of my taste and improve your recommendations over time.
Too Easy?
Try turning your taste curator into an active recommendation engine.
Connect it to places where your taste already exists digitally, then have it send you a small curated email digest every week or every day with recommendations it thinks you would genuinely love.
You can use this prompt to get started:
I want you to act as my personal curator and recommendation engine.
Your goal is to understand my taste well enough to recommend things that feel genuinely aligned with my interests, values, and curiosity.
I would like you to regularly generate a curated recommendation digest for me via email. The recommendations should feel thoughtful and personal, not generic or trend-driven.
I am going to share examples of things I like over time. As you analyze them, identify deeper patterns in my preferences and use those patterns to improve your recommendations.
For each recommendation, explain why you think it matches my taste.
Why this matters?
There’s a great Brian Eno interview that I love where he talks about art and how powerful art is really about making choices. Taste is the ability to decide what matters, what feels right, what should exist, and what should not.
I think about that a lot in the context of AI.
The AI labs might be right that execution is becoming cheaper. It is easier than ever to generate writing, images, code, products, strategies, music, and ideas. But a lot of it is not good.
The quality will continue to improve, but that does not mean everything we can generate should exist. If we are going to navigate the next couple of years well, refining our own sense of what we believe is worth building will matter.
I think of AI as a partner in that process, not something we should defer our judgment to.
That’s it for this week.
Hit reply and tell me what you thought of the challenge this week.
I read every response, and your experience helps me shape these challenges. If you get stuck or have questions, reach me at [email protected]