You Had No Taste Before AI

There’s been an influx of people telling others to develop taste to use AI. Designers. Marketers. Developers. All of them touting the same message. It’s ironic, though. These are the same people who never questioned why their designs all look identical, never iterated beyond the first draft, and never asked if their work actually solved the problem at hand.

They’re not alone. The loudest voices preaching about taste and AI are often the ones who never demonstrated taste before AI.

What is Taste? #

The technology industry has a tendency to use words that mean multiple things without describing which definition they are referring to. When I read about taste and AI I usually see people referring to the following definition.

Critical judgment, discernment, or appreciation of aesthetic quality.

In the context of AI, this definition manifests itself in several ways.

Contextual Appropriateness: Knowing when AI-generated content fits the situation and when it doesn’t. Put another way, knowing when a human touch is needed (e.g., a message to a loved one).

Quality Recognition: Being able to distinguish between useful AI-generated content and slop. This requires domain knowledge to truly discern aesthetic quality rather than just functional quality.

Iterative Refinement: Understanding that AI is a starting point that requires further iteration. This point is most similar to how culinary taste is applied to refine a dish by iterating on the recipe and presentation.

Ethical Boundaries: Recognizing when AI crosses the lines of authenticity, legality, and respect. Basically, don’t use AI to do bad things.

None of these skills are new. These are the same skills we should have been applying to our work all along. Why are we asking about taste and AI now when we should have been applying taste the whole time? Perhaps people advocating for taste are telling on themselves.

Being Tasteless #

Some people have no taste. In the best case that may be due to lack of experience but in the worst case it may be due to ignorance. I’m noticing that many people worried about tasteless AI-generated content are often guilty of producing tasteless content themselves, usually manifesting as the following.

• Copying and pasting code without understanding it.

• Sending resumes and emails that aren’t proofread and edited.

• Asking others to review code without giving it a self review.

• Noticing a quality issue and failing to document or fix it.

• Designing websites that look exactly like every other company’s website.

• Regurgitating content from the trending influencer of the week.

Where’s the taste here? Where’s the critical judgment, discernment, or appreciation of aesthetic quality that separates mediocrity from excellence?

It’s not there because most people haven’t developed their taste yet. AI didn’t create this tasteless problem. People did. Now that everyone can generate content at the speed of thought we’re noticing that not all content is actually good. To play on a popular quote from Ratatouille, anyone can cook, but not everyone is a chef. Don’t complain about mediocre work when you’re producing mediocre work yourself.

Spectrum of Taste #

What about the nature of taste itself? Should people focus on developing depth of taste in specific domains or breadth of taste across many domains? My short answer is a bit of both, if possible.

Depth of taste means becoming an expert within a particular domain. We’ve all met such experts and even asked them for help on tricky, bespoke topics within their domain. A person with depth of taste can recognize when AI-generated content is refined and of high quality versus merely functional. This kind of taste comes from years of experience in a specific role coupled with deep domain knowledge.

Breadth of taste means becoming knowledgeable across multiple domains and understanding how those domains interface with one another. A person with breadth of taste can recognize when AI-generated content is contextually appropriate, authentic, and of enough quality to use for their needs. This kind of taste comes from years of experience across multiple roles coupled with moderate domain knowledge.

Breadth of taste is more valuable with AI. When using AI, you’re constantly switching between domains: a software engineer writing documentation, a marketer creating designs. Breadth lets you maintain quality across these contexts while recognizing when you need domain expertise. You iterate faster because you have opinions about what “good enough” looks like across multiple domains.

The people I see being most effective with AI developed a breadth of taste that they use to determine what good AI-generated content looks like, regardless of domain. They can recognize when something feels off, even if they can’t articulate exactly why. They understand their own limitations and know when to seek expertise in a specific domain. That’s not to say those with depth of taste can’t be successful with AI, but I see those people reluctant to use AI because they are more knowledgeable than AI in a particular domain.

It Tastes Bitter #

If you’re reading this thinking you have to spend time developing your taste, good! Perhaps I’ve left a bitter taste in your mouth. The good news is you’re not alone. There are many people that need to hear this to better their taste, myself included. The challenge here is recognizing that it’s not about developing taste for AI but rather about developing taste, period. If you’ve had poor taste before AI you’ll have poor taste with AI. If you’ve had good taste before AI, you’ll be able to apply that taste with AI.

Instead of treating AI taste as some mystical new skill, focus on the fundamentals that were always important. Here are some actionable ways to develop your taste.

Tomorrow: Pick one piece of work you’re proud of and one you’re not. Write down specifically what makes them different. That’s taste in action.

This week: Find three examples of excellence in a domain you work in. Study them. What patterns emerge? What choices did the creators make?

This month: Take something you’ve created with or without AI and iterate on it a few times. Each iteration should have a specific improvement based on a specific critique.

Always: When someone preaches about AI taste, ask them to show you their work from before AI. If they can’t demonstrate taste in their pre-AI work, they’re not qualified to lecture you about it now.

The people succeeding with AI aren’t the ones who suddenly discovered taste. They’re the ones who already had it and simply adapted their standards to a new tool. Develop your taste with or without AI. The medium doesn’t matter, the fundamentals do.

Stop waiting for AI to force you to develop taste. Start now.