When AI Makes Everything Good, Taste Becomes Everything
Yesterday, your intern replaced your $15k agency. Today, AI wrote a better strategy deck than your MBA consultant.
Yesterday your intern replaced your $15K agency.
Today, AI wrote a better strategy deck than your MBA consultant.
Scary? It gets worse.
This isn’t a tech problem. It’s an everyone problem.
Your competitor’s AI spat out a logo in ten seconds and yeah, it looks fine. But design isn’t supposed to just look good. It’s supposed to feel good.That’s the part no machine can fake.
That “unique” brand tone you’re proud of?
AI wrote it during a coffee break.
If it hasn’t hit you yet, it will.
Execution isn’t the edge anymore.When everyone can make things perfect, taste is all that’s left.
The death of difficulty
In 2018, I remember late nights with my design team arguing over gradients that no one but us would notice, fighting over budgets, and learning (the hard way) how to keep a project on track. It was messy, exhausting, and deeply human. But that’s the thing. It’s the mess that made the work feel real.
Back then, good design wasn't just expensive it was challenging. The process itself was a filter. Not everyone could translate an idea into a polished product. The complexity demanded real skill, real learning.
Now, everything feels different.
For $20 a month, AI can do what took us months to learn. And let’s be honest, some founders think that makes them the next Jony Ive.
But real craft takes more than just hitting 'generate.'
Just because something looks professional doesn't mean it's good.
We're drowning in competent looking but soulless work. Those perfect websites and sleek interfaces? They're starting to blur together.
Take Apple's early days.
Steve Jobs didn't just want a computer that worked. He wanted a computer that felt magical. That's the difference between executing a task and creating something memorable. AI can generate code, but it can't generate that spark of human insight.
This is about understanding that the most interesting innovations come from deep understanding, not just surface-level execution.
The new creative crisis
Let's look at your competition's tech stack:
Website? Framer AI barfed out another clean, modern interface
Brand? Midjourney's interpretation of trustworthy but disruptive
Code? Replit greatest hits, now with 30% more technical debt
Copy? I’ve seen founders use Claude to pull the same AI-generated recycled marketing lines and somehow think it’s original.
When AI makes everything polished, the definition of 'good' shifts.
I’ve seen founders obsess over AI generated pitch decks that look stunning but feel hollow. The problem is the lack of a soul. A slick deck might get you a meeting, but it’s the raw honesty of your story that closes the deal.
The startup cover band crisis
I've watched founders get seduced by AI generated perfection.
Silicon Valley feels like a karaoke night where every startup is trying to sing the same song. These days, every tech product looks the same. It’s like Airbnb and Notion decided to have a kid.
More and more brands are starting to look the same, like they’ve all been stamped out by the same AI machine.
It's happening in writing, in product development, in marketing. AI can generate a pitch deck that looks like it was crafted by a Stanford MBA, but it can't capture the raw passion that actually sells an idea.
Remember the original iPod vs Zune drama? Same tech, same time. One screamed "I've got more features than you can count," the other just said "1000 songs in your pocket."
Apple’s genius was in making people feel like they’d joined a club.
The hidden value of craft
Now anyone can hit the professional baseline with AI. This makes actually understanding craft more valuable than ever.
Writers still need to study why Hemingway hits different, even if AI can help with grammar. Sure, AI can frame a shot, but what about making people feel something? That's why we study Kubrick and Hitchcock.
Want to build great apps? AI will help you ship clean code and slick interfaces. But you still need to know why great systems scale and how classic design principles actually work.
Even if you're using AI prompts, you need to understand why great design works, not just how to copy it. You're not just learning how. You're learning why.
The economics of looking like everyone else
Pre-2024:
You needed enough cash to hire that pretentious agency that wouldn't return your calls or emails.
Post-2024:
Everyone's a designer.
Instagram turned everyone into a photographer. Except, most of us were just messing around with filters and hoping for the best. AI’s version of this doesn’t even have the charm of getting it wrong. It’s just competent to a fault.
What can't AI fake? The stuff that actually matters:
Knowing when minimalism means "premium" vs "we ran out of ideas"
Knowing why your AI writing sounds like every other LinkedIn guru having a breakdown
Recognising why clean code matters before your entire engineering team rage-quits in Slack
Building your taste engine
Study Epic Fails
When Coca-Cola launched New Coke in 1985, they thought better taste was enough. But they missed what their audience really wanted – not a better cola, but the emotional connection to their classic brand. The product wasn't bad, it just ignored why people actually chose Coke.
Get Opinionated
Every design decision sends a message. Using gradients could say 'modern startup' or scream 'trying too hard.' The difference? Having a clear point of view about why you chose it, not just mimicking what's trending.
Stop Following Your Competition:
They're all copying each other anyway. Study how Nike sells stories, not shoes. How Wes Anderson makes symmetry feel rebellious. How Apple knows when to shut up about features nobody asked for.
Learn from disasters. Get strong opinions. And please, stop copying everyone else.
Taste in action
Without taste: "Our AI-powered platform leverages cutting-edge machine learning to optimise your workflow"
With taste: "We fixed the part that made you want to quit your job"
Without taste: "Another productivity app with infinite folders and dark mode"
With taste: "An app that knows most people hate organising folders in the first place"
North star for taste
The best products will win because someone had the guts to let their own voice shine through. So, yeah, learn and use the tools. But don’t let them do all the talking. That’s your job.
The middle ground is vanishing faster than Web3 promises. AI’s leveling the playing field, and it’s up to you to use it to stand out, or risk blending in with the rest.
Taste isn't just an acquired skill. It's what separates you from the thousands of wannabes letting AI think for them.
In the end, taste might be the last thing that truly sets you apart.
Enjoy the holidays and until next week,
Martin
P.S. This week's soundtrack is "Stressed Out" by Twenty One Pilots. Because while everyone else is letting AI do their work, some of us remember when creating something meant actually being stressed out about it. And maybe that stress wasn't such a bad thing after all.
Tool of the week - getimg.ai
Another AI image generator, but this one actually does what you tell it to. No more "abstract humanoid figure in a corporate setting" when you asked for "founder having existential crisis." The free tier gives you 100 images/month. Perfect for when your designer quits after seeing your 4am product roadmap.
Recommended Book
Lovemarks - The Future Beyond Brands - by Kevin Roberts
While Silicon Valley chases AI hallucinations, this 2004 book about brand romance is weirdly relevant. Great products don't win because of features anymore. They win on taste, storytelling, and that impossible-to-fake feeling of "they just get it.
This piece is part of my Innovation pillar in the G.R.I.T. framework - focused on helping you scale intelligently without burning out or burning cash.
Human taste and creative judgment 💯