Two entrepreneurs walk into a room. Both have access to Claude. Both have access to GPT. Both can generate copy, build apps, analyse data, and deploy agents that work around the clock. Their toolkits are functionally identical. Six months later, one has built something remarkable and the other has built something forgettable. The difference isn't technology. It's taste.
I've been thinking about this a lot recently, because the conversations I'm having with founders have shifted. A year ago, the question was "how do I use AI?" Now it's "I'm using AI for everything but the output feels… generic." They can't always articulate what's wrong. They just know something is off. The work is competent. It's also completely unremarkable.
That's the taste gap. And it's becoming the single most important competitive advantage in business.
What Taste Actually Is
Taste isn't aesthetics. It isn't having a nice font on your website or knowing which shade of blue converts best. Taste is the ability to look at something — a product, a strategy, a piece of writing, a customer experience — and know instantly whether it's right. Not whether it's correct. Whether it's right.
It's the founder who reads the AI-generated marketing copy and says "this is technically fine but it sounds like everyone else, rewrite it with the specific story from the client in Manchester." It's the product designer who looks at the AI-produced wireframe and says "this solves the brief but it doesn't surprise anyone, and we need surprise here." It's the CEO who reviews the AI-drafted strategy document and says "the analysis is solid but the conclusion is cowardly — we should be recommending the harder path."
Taste is editorial judgment at scale. It's knowing what good looks like before you see it. And it cannot be delegated to a machine.
The Commoditisation of Execution
Here's what's happened in the last eighteen months. AI has commoditised execution. Writing, coding, designing, analysing, summarising, planning — the doing of knowledge work is now available to everyone for essentially nothing. A solo founder can produce the output of a twenty-person team. We've all celebrated this. I've written about it myself.
But commoditisation cuts both ways. When everyone can execute at the same level, execution stops being a differentiator. The playing field doesn't just level — it flattens entirely. And when execution is flat, the only dimension left to compete on is the quality of the decisions that direct that execution.
That's taste. That's judgment. That's the thing between the prompt and the publish button.
AI didn't eliminate the need for taste. It made taste the only thing that matters.
The Taste Spectrum
I see three levels of AI usage right now, and they map directly to taste.
Level one: accept. The entrepreneur generates output and ships it. The blog post goes live as-is. The product feature launches with the default design. The email campaign sends with the first draft. This is fast and it's fine for internal operations where polish doesn't matter. But for anything customer-facing, anything that builds brand, anything where perception compounds — accepting the first output is leaving money and reputation on the table.
Level two: refine. The entrepreneur generates output, recognises what's wrong with it, and iterates. They push the AI through three, four, five rounds until the result matches their internal standard. This is where most good operators sit. It works. But it's slow, because each refinement requires the founder to articulate what "better" means — and most people struggle to verbalise taste. They know it when they see it. They can't always describe it before they see it.
Level three: direct. The entrepreneur has developed such clear taste that they can articulate what they want before they see it. Their prompts aren't "write me a blog post about X." They're "write me a blog post about X in the style of our best-performing piece, with the specific story about the client who nearly went under, opening with conflict not context, no more than 900 words, ending on a question not a statement." The output needs one pass, maybe none. The taste is encoded in the direction.
Level three is where the advantage lives. And getting there requires something AI can't give you: years of paying attention to what works and what doesn't, developing opinions, and being willing to reject competent work in pursuit of excellent work.
Why Most Entrepreneurs Have a Taste Problem
I want to be honest about something uncomfortable. Most entrepreneurs don't have great taste. Not because they're stupid — because they've never needed it at this resolution.
Before AI, you hired people with taste. You hired a copywriter who knew what good writing looked like. A designer who knew what good design looked like. A strategist who knew what good strategy looked like. Their taste was embedded in their craft. You didn't need to have taste yourself — you needed to have the judgment to hire people who did, and the wisdom to trust them.
AI broke that model. Now the entrepreneur is directing the work directly. There's no intermediary with taste acting as a quality filter. It's you and the machine. And if your taste isn't sharp enough to distinguish between competent and compelling, between functional and remarkable, between correct and right — every output lands at the same mediocre middle.
This is why so much AI-generated content looks the same. It's not that the AI can't produce distinctive work. It's that the people directing it can't tell the difference between distinctive and generic. The bottleneck moved from execution to editorial, and most people didn't notice.
Developing Taste
The good news is that taste isn't innate. It's trained. Like any skill, it develops through deliberate exposure and practice. Here's how I think about it.
Consume widely and critically. Read things outside your industry. Look at brands that make you feel something. Study marketing that stopped you scrolling. Pay attention to your own reactions — when something impresses you, dissect why. Most people consume passively. Taste develops through active consumption.
Develop strong opinions. Taste requires a point of view. You need to believe that some approaches are better than others, not just different. This means being willing to say "that's bad" even when it's popular, and "that's brilliant" even when it's obscure. Opinions are the scaffolding of taste.
Study the delta. Every time you refine AI output — every edit you make, every thing you reject, every iteration you push through — write down what you changed and why. Over time, you'll see patterns. Those patterns are your taste, made visible. Once they're visible, you can encode them into your direction and skip the refinement loop entirely.
Get feedback from people with better taste than you. This is the one most entrepreneurs skip. Find someone whose work you genuinely admire and ask them to critique yours. Not your strategy. Your output. Your website copy, your product design, your investor deck. Their reactions will recalibrate your internal standard faster than anything else.
You can outsource execution to AI. You cannot outsource the ability to recognise what good looks like. That's yours to build.
The Business Case for Taste
Let me make this concrete, because "develop better taste" can sound like lifestyle advice rather than business strategy.
Two competing consultancies use AI to produce client deliverables. Consultancy A ships what the AI generates with minor formatting tweaks. Consultancy B has a principal with sharp taste who rewrites the executive summary, restructures the recommendations to lead with the counterintuitive finding, and adds a single chart that tells the whole story at a glance. The underlying analysis is identical. The client experience is worlds apart. Consultancy B charges 40% more and has a 90% renewal rate. The difference is taste.
Two e-commerce brands sell similar products. Both use AI for product descriptions, email campaigns, and social content. Brand A sounds like every other brand in the category — clean, professional, forgettable. Brand B has a founder who insists on a specific voice, rejects anything that feels safe, and would rather publish nothing than publish something mediocre. Brand B builds a cult following. Brand A competes on price. The difference is taste.
This isn't marginal. In a world where execution is commoditised, taste is the primary driver of pricing power, retention, and brand equity. The businesses with the sharpest taste will charge more, keep clients longer, and attract better talent — because quality is magnetic, and quality starts with someone who can see it.
The Uncomfortable Truth for Solo Founders
I've written a lot about the one-person business powered by AI. I believe in that model deeply. But here's the tension: when you're the only human in the loop, your taste is the ceiling. There's no colleague to push back. No creative director to say "this isn't good enough." No team culture that enforces a standard.
If you're building a company of one, your single most important investment isn't a better AI tool or another automation. It's sharpening your own taste. Read more. Look at more. Develop opinions about what great looks like in your specific domain. Build a mental library of excellence so deep that mediocrity becomes physically uncomfortable to you.
And if you genuinely can't develop taste in a particular area — design, writing, strategy, whatever it is — then that's the one thing you should still hire a human for. Not for execution. For judgment. A fractional creative director. A brand advisor you talk to monthly. Someone whose taste you trust to catch what yours misses.
The Gap That's Coming
Here's my prediction. Over the next two years, we're going to see a dramatic split in the market. On one side, a sea of competent, AI-generated, indistinguishable businesses — all using the same tools, producing the same quality, competing on price in a race to the bottom. On the other, a smaller group of businesses that feel different. That have a point of view. That make you feel something. That charge a premium and earn it.
The difference won't be technology. Everyone will have the technology. The difference will be the human at the centre — their taste, their judgment, their willingness to reject the good in pursuit of the great.
AI gave everyone a voice. Taste determines who's worth listening to.