The tech world has recently rediscovered taste. Founders talk about it. Investors cite it as a differentiator. Paul Graham has argued that taste will only become more important as artificial intelligence takes over more of what once required technical skill.
The conversation has real energy, and some of it is genuinely useful.
But most of it stops too soon.
Taste gets named as an advantage without anyone seriously asking what it takes to develop it. And that question turns out to be considerably more difficult than the current conversation suggests. This series is an attempt to take it seriously.
Because taste, properly understood, is not just preference or style. It is a cultivated form of perception: the ability to recognize qualitative differences well. To notice what fits and what does not. To distinguish the merely competent from the genuinely excellent. To perceive coherence, proportion, integrity, vitality, depth, and meaning with increasing clarity.
That kind of perception does not emerge automatically. It has to be formed. It is not simply innate.
Over the coming installments, I want to examine what that formation actually requires and why many of the dominant conditions of digital life work against it. The AI context matters here not because AI makes taste irrelevant, but because it changes the conditions under which taste develops.
An earlier ARL piece argued that taste is not a skill. The distinction matters. Skills are usually legible, transferable, and testable. Taste is something else: a developed capacity for judgment, closer to what philosophers mean by discernment than what a hiring manager means by competency.
David Hume understood this clearly. Taste, for Hume, is not personal preference. Preferences are arbitrary and largely beyond criticism. Taste, by contrast, can be cultivated, refined, compared, and evaluated. Some judgments are better than others, and the difference depends on the quality of perception brought to bear.
Immanuel Kant pushed the analysis further. Aesthetic judgment, for Kant, is neither purely subjective nor purely intellectual. It begins in feeling but carries a claim that reaches beyond private sensation. When you say something is genuinely good, you are implicitly asking others to see what you see.
The figure that best captures this in practice is the connoisseur, not as a symbol of elitism or social status, but as a description of developed perception.
A wine connoisseur does not become one by reading about wine or having strong reactions to it. They become one through years of disciplined attention: exposure to many examples, repeated comparison, gradual refinement, and the experience of being wrong. Distinctions that were once invisible slowly become obvious. What originally required conscious effort becomes instinctive perception.
This is the basic model of taste development.
Now here is the problem.
Many of the conditions that historically formed taste are precisely the conditions AI tends to shortcut.
Productive friction develops judgment. When you sit with uncertainty, work through competing possibilities, make difficult evaluative decisions, and struggle toward clarity without immediate resolution, something develops. The struggle itself is formative.
When a fluent answer or polished output is always available, that development can weaken. Not necessarily because the output is bad, but because reaching for it too quickly closes down the evaluative process before it has done its work.
Wide exposure sharpens discrimination. Hume emphasized this repeatedly. Taste develops through encounter with a broad range of cases, especially cases that challenge your existing preferences and force revision.
But recommendation systems optimize for engagement, not formation. You are continually shown variations of what you already respond to. The narrowing often happens invisibly. The algorithm is functioning exactly as intended. The problem is that its goals and your development are not the same thing.
Sustained attention is equally essential. You cannot develop a perceptual capacity you rarely exercise. Taste requires concentration, patience, memory, and repeated engagement with difficult material. Yet many of the same systems that generate endless content also fragment the conditions necessary for deep attention.
This is not simply another complaint about distraction. It is a more specific observation: taste depends on forms of attention that current technological environments often erode. None of this means AI is the enemy of taste. The relationship is more complicated than that. These tools can assist perception, widen access, accelerate iteration, and expose people to work they might never have encountered otherwise.
But used unreflectively, they can also weaken the very capacities they are increasingly meant to amplify.
You do not develop taste by outsourcing judgment. You develop it by exercising judgment repeatedly under conditions that provide resistance, feedback, correction, and comparison.
Why does this matter beyond aesthetics?
Because taste, in the philosophical sense, is not primarily an aesthetic luxury. It is a form of perceptual competence.
People with developed judgment notice things others miss. They recognize when something is genuinely alive rather than merely optimized, genuinely excellent rather than merely acceptable, genuinely coherent rather than merely functional. They can hold standards in view even when pressure encourages compromise.
These are not soft capacities. They shape leadership, creativity, culture, design, education, communication, and organizational life more broadly.
Organizations often treat taste as decoration rather than a capacity that can be deliberately cultivated. When organizations neglect the conditions that develop it, discernment deteriorates quietly and without announcement.
And in a moment when AI increasingly produces work that is consistently competent, occasionally impressive, and frequently interchangeable, the ability to perceive meaningful qualitative differences becomes more valuable, not less.
The gap between competent and excellent is where taste operates. And it is precisely the kind of gap that automated systems tend to flatten.
Taste development, I want to argue, depends on a set of practices and conditions that can be deliberately cultivated.
Future installments will examine attention and what deep perception actually requires; exposure and how to widen it intentionally; comparison as a discipline rather than a casual habit; apprenticeship, imitation, and the role of exemplars; and the resistance to received opinion that marks genuinely independent judgment.
The question running through all of it is not whether to use AI tools. Most of us will. The real question is whether we are using them in ways that develop perception or quietly replace it.
That distinction becomes harder to maintain the more capable the tools become. Which is precisely why the problem matters now.
Because the future may belong less to those who can generate endlessly and more to those who have learned, patiently and rigorously, how to see well.
I work with organizations and individuals who want to think more carefully about judgment, meaning, and the conditions that shape everyday decisions.
If that sounds like your situation, you can reach me at michaelrspicher@gmail.com or learn more at michaelrspicher.com.
Eating and the Tasteful Subject
My latest for Fast Company: Are You Part of the ‘Distraction Economy’?
“The Aesthetic Edge” is a free ebook, co-written by myself and Hannah Spicher. Download your copy today!
I appeared on Jeffrey Shaw’s podcast to talk about beauty, luxury, and the aesthetic advantage.
Michael’s interview with BOLD Journey.
Digital Fashion: Theory, Practice, Implications, edited by Michael R. Spicher, Sara Emilia Bernat, and Doris Domoszlai-Lantner, is available for purchase! (Our book was featured in a BeautyMatter article.)
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.