"Taste is a new core skill"
Actual curation is choosing what you refuse to consume
As of March 10th 2025, 71% of all images on social media are AI generated.
This is forecasted to increase to 90% throughout this year.
23 million videos are uploaded to TikTok daily.
300 to 500 hours of video is uploaded to YouTube every minute.
The unending and colossal wall of content that we are exposed to the moment we unlock our phones cannot be understated. Human made, AI generated, or a subliminal combination of the two.
OpenAI president Greg Brockman posted on X last month, “Taste is a new core skill,”
In this succinct statement he implicitly acknowledges two truths, both of which we’ve felt to be true for some time:
That many of our established core competencies are soon to be outdated
In order to continue to be valuable, we need to become subjectively unreplicable.
Whoever sees fire through the smoke, whoever wades through all the muck to find relevant and timely signals, will be…layoff proof I guess?
Historically, taste has been class selective. The upper classes cultivated preferences that required a lengthy education or expensive craft work. But during the 21st century wave of “pop-timism”, taste was democratically placed back in the hands of the working class.
The message of pop-timism was simple: that if something was mass consumed it didn’t necessarily mean that it was bad. Gone were the days that you had to indulge in pop culture as a guilty pleasure. This embrace of the mainstream was adopted by early tech giants like Mark Zuckerberg and his young silicon valley associates. They didn’t need to wear suits to work…a hoodie and jeans won them billions in investments.
Fast forward to 2026. With the rise of algorithmic content platforms oversaturated with AI, and the mainstreaming of fringe right wing ideology, it’s impossible for anyone left of liberal to embrace the mainstream in its current form.
So being sophisticated is cool again. And who knows this better than anyone? The companies who own the trend prediction data.
Last week The New Yorker released an article on the origins of “taste-washing”, citing Anthropic’s pop up in New York selling stylish ball caps featuring the phrase “Keep Thinking”, or OpenAI’s use of handheld cinematography in their SuperBowl ad.
Following his traditional masculinity makeover, Mark Zuckerberg’s front row attendance at Prada ruffled more than a few feathers. Following the Meta Glasses & Ray Ban collaboration, Meta is seeking to target a higher-brow luxury audience with a new partnership.
Online audiences were disappointed to see Law Roach, long time stylist of Zendaya, dressing Lauren Sanchez for Paris Fashion week.
OpenAI has recently announced a collaboration with legendary former Apple product designer, Jony Ive, on an AI-integrated hardware project.
Millions and millions of dollars are being poured into the “taste-washing” of tech companies and their founders. The fact is, AI and tech companies are inherently uncool. Their product is diametrically opposed to human creative intuition, and poses an active threat to the livelihoods of millions of Americans. It makes sense that they would need a complete image overhaul.
But here’s the question: is the elevated, stylish aesthetic that they’re working so hard to convey…actually “taste”? And if having “taste” is as important as they say it is…what does that actually mean?
“How to have taste” itself is a content genre. It’s as hard to describe the term “taste” as it is to nail down the meaning of “camp”.
A viral essay on X by investor Will Manidis details the ways that the definition of “taste” has changed alongside our relationship to art and consumption. He argues that “taste becoming a new core skill” isn’t elevating our human skills of judgement, it’s a demotion to consumer while the machines get to do all the creating.
He writes: “The taste thesis, at its deepest and most simple structure, reverses this order. It places man at the end of the chain of creation, evaluating what has already been generated, rather than at the beginning, participating in the generation itself.”
“Taste” as it’s been defined across the ages, was the realm of the creator and the critic. “Taste” was developed over time and was sustained by a deep relationship to the source material. It cannot be synonymous with “aesthetic preferences”, because visual subjectivity doesn’t provide enough information to understand the context from which an artistic work derives.
The most respected creators and muses of any medium are always the most hyper individualistic. As Eleni Di Luzio writes in her essay, “Taste is almost autobiographical, it’s your life, expressed through your eye.”
This is where the 2026 definition of “taste” and “taste-makers” falls down. Scouring dozens of product recommendations for the next hyper trendy pre-biotic soda or coffee table book may get you points in certain circles, but this data can and will be replicated eventually by prediction algorithms. “Taste” is not about searching for what is the most palatable to others. This is just the algorithm in disguise.
And the algorithm will never show you something you dislike…twice. It will show you things that will make you angry, or make you sad, but it will never purposefully challenge you if you don’t linger on the video for more than a few seconds. How can one expect to “develop personal taste” within an echo chamber of all the things you already like?
This algorithmic form of “taste” is exactly what is being co-opted by these tech companies. It’s just a more cosmopolitan form of consumerism.
To conclude, how do we actually develop a sense of “taste” within the echo chamber?
A lot of the advice that I see online is “know yourself”. Understand your own motivations, engage thoughtfully with subject matter, and take an active role in your own cultural education. All of these I agree with, but I want to suggest that another productive way to engage with “tasteful” content is to take even more notice of what you’re not engaging with, compared to what you already are.
We can easily imagine what we will do, but knowing what we won’t do requires having been tempted and refused. To explore the negative space requires engaging your value system around it. “Taste” is the weight of full life lived, discomfort included.
The algorithm can only be optimized. The most human thing left may simply be choosing what you refuse to consume.
Very interesting read. Made me wonder - we’re talking about “taste” as if it’s something you can develop despite the system, when historically it was something that emerged because of constraints.
You didn’t just choose what to engage with, you were limited by language, geography, access, people around you. Taste was partly an accident of exposure, and then a negotiation with it.
Now, the system removes those constraints but replaces them with invisible ones. So your exposure feels expansive, but is actually narrower in a very specific, optimised way.
Which makes me think - is the problem really that we’re becoming “just consumers”?
Or that we’re losing the conditions under which taste becomes interesting in the first place?
Because without constraint, refusal doesn’t carry weight. And without weight, “taste” starts collapsing into aesthetic signalling again, just with better data behind it.
Taste is naturally reductive. (Refined or distilled might be better terms.) As you point out, it's helpful if you know yourself well but that requires time not actively consuming (i.e., off social media) to reflect.
Ideally, in that down time you're creating—as humans are meant to, and creation is an expression of one's identity (which feeds back into taste).
I don't think that can be rushed (nor should it) but exposure goes a long way to help recognize what you opt for. Constraints help, too.