The End of Taste

Or, when beautiful things stopped proving anything

Lately I’ve been thinking about taste. This started, strangely enough, with a lamp. I get a lot of catalogs and one of them is from Design Within Reach. Every few months I sit with a cup of tea and flip through it like a devotional text. Recently I found myself obsessing over the Louis Poulsen lamps, especially the new 150th anniversary color wave. They’re beautiful. Thoughtful and tasteful in the exact way tasteful things are supposed to be. Soft colors, perfect proportions, a sense of restraint that quietly signals someone serious designed them.

And yet the lamp I actually got (and love) is a ceramic lamp I found abandoned on the side of the road. It’s tall and painted bronze and curves upward like a genie bottle. I found it in a pile of what looked like discarded Palm Springs motel furniture from the 70s. It’s very me to deeply admire two completely different things at once. To either have very unique taste or perhaps no coherent taste at all. But it got me thinking about what taste even is now, and whether it still functions the way it used to. Because for a long time, taste was one of the ways humans identified quality, but AI is rapidly changing that.

Taste was shorthand
Beautiful things, thoughtful things, elegant things helped us know that someone serious was behind them. A tasteful object implied care, investment, and artistry/craft. And humans need taste because it takes knowledge, experience and time to determine if an object has these qualities. We cannot deeply inspect everything from first principles. So civilization developed shortcuts. Taste became one of them. And for a while, the shortcut worked surprisingly well.

The internet rewarded the signaling layer
For twenty years, a tasteful website implied a thoughtful company. A restrained brand implied seriousness, good design implied product quality and a cinematic black and white campaign implied depth. Even in startups, taste bias absolutely exists. Historically, those signals were directionally useful. Maybe not perfect but useful. But since AI hit the scene, the internet has increasingly detached the signals of taste from the underlying discernment that once produced them.

To be fair, this always happened. Fashion runs in cycles one year the hottest sneaker is in the following years TJ Maxx bin (RIP <3). But, it took time. You’d get a year or so. But in the knowledge economy, these cycles are super short and AI is accelerating taste cycles to an absurd degree.

AI makes polished outputs abundant
And now that these outputs are abundant what happens to taste? At first, it seems like taste simply shifts. Historically, this is how taste evolved anyway. Fashion, art, architecture, music. The signal changes when the old one becomes too accessible. But AI compresses these cycles toward zero. The moment culture identifies a new aesthetic signal (maximalist, ironic, eccentric) AI replicates it everywhere almost instantly. Not just on beautiful products, but on mediocre ones too. This is the important distinction. AI did not destroy quality it destroyed the reliability of quality signals. Taste no longer means anything.

We are running out of likes
This is the strange emotional feeling many people have online right now. Everything suddenly looks good. Soon every startup will have perfect gradients, cinematic photography, elegant typography, and emotionally calibrated copy. Every product will feel designed. Our feeds are saturated with sophisticated-looking things, and our brains still respond to them because humans evolved to associate coherence, elegance, polish, and restraint with competence, intelligence, and care. But increasingly, much of this feels like simulacrum. The Las Vegas version of Paris becoming emotionally interchangeable with Paris itself. So we hit like, like, like. But eventually the emotional economy breaks.Everything cannot be beautiful simultaneously. Everything cannot signal thoughtfulness simultaneously. The signal stops differentiating. Beautiful things no longer prove anything. This is the end of taste.

Not the end of beauty, quality, or craft. The end of taste as proof of them.

The terrifying possibility is that maybe it doesn’t matter
There is a darker possibility underneath all of this. Maybe humans are perfectly happy consuming polished mediocrity. Maybe simulated sophistication is emotionally satisfying enough. Maybe most people do not actually want originality, depth, or craft. Maybe they simply want recognizable signals of them. That thought honestly terrifies me a little. Because it suggests we may have spent years rewarding performative intelligence, aesthetic coherence and strategic theater. We’re consuming polished sameness while confusing it with greatness itself and the underlying thoughtfulness becomes optional.

But if taste collapses, reality matters again
And yet I think there is a very hopeful inversion hiding inside all this. If everybody can generate beautiful stuff then the jig is up. Substance has room to win. For years the signaling layer could deceive us. People could perform sophistication without enduring the struggle underneath it. They could ship the appearance of craft without the craft itself. But when sophistication becomes infinitely reproducible, sophistication stops being scarce. Substance becomes visible again.

The future belongs to the irrationally obsessed
The builders who matter in the next decade may not look the smartest or sound these smartest. Instead they will understand systems more deeply, survive rejection longer, evolve faster and obsess over tiny details again. The Romantics emerged during industrialization because they feared society was overvaluing polish, rational systems and mechanized production and undervaluing obsession, genius, individuality and emotional truth. In some ways AI may force a new Romanticism. Not because humans reject technology, but because once sophistication becomes automated, people start searching for signs of actual consciousness again.

Vibes lose economic power
This is especially important in startups. We’ve lived through an era where storytelling, design taste and strategic fluency could shape fundraising, hiring and customer trust. There was prestige in aesthetic coherence in our attention economy. And these things will still matter. Humans are emotional creatures. But if everyone can manufacture those signals instantly, their advantage compresses. And eventually product depth, execution and endurance matter more because the old shortcuts stopped differentiating.

We stop trusting aesthetics and start trusting people
So how do we find great things in a post-taste world? I think we go deeper. We stop outsourcing judgment entirely to aesthetics. We try the product. We spend time with the builders. We listen to long-form conversations instead of polished launch videos. We ask difficult questions. We look for coherence over time. We sit for the strange technical tangent instead of the steak dinner.

I don’t want a perfectly tasteful website with pithy copy. I want to understand why your product works differently. I want to hear the weird obsessive insight that made you build it in the first place. I want to meet the team and watch what happens when they go off-script. Because increasingly, trust will move away from aesthetics and back toward direct experience.

The best recommendations become private again
I also suspect word of mouth becomes much more important. Not social proof or likes, not even virality. The real recommendations that come when you and your friend are having coffee. When no one is looking and you can talk about what really works and who is actually just a poser. When you can talk about the product you cannot stop thinking about. That’s a very different kind of signal. It’s more relational and experiential. It’s less scalable but much more real.

This is not the end of greatness
I don’t think this is the end of quality, beauty, or discernment. It may actually be the opposite. When beautiful things no longer prove anything, we are forced to look deeper. And maybe that’s healthy. Maybe this is not the death of discernment. Maybe it’s the death of outsourced discernment. And maybe, after years of rewarding the people best at performing sophistication, we finally return to rewarding the people who actually understand how things work.