Asher's Blog

When anyone can generate a respectable article, image, presentation, product interface or piece of software, production itself becomes less valuable. The scarce ability shifts from making something to deciding what deserves to be made, what should be removed and what is good enough to represent you.

That ability is taste.

Taste is sometimes treated as a mysterious gift possessed by designers and artists. In business, it is more practical. It is the quality of judgment that allows a person to recognise coherence, usefulness and meaning among many technically acceptable possibilities. It is knowing why one sentence is honest and another is merely impressive, why one feature simplifies the customer’s life and another adds noise, and why a product that can do more may still be worse.

AI makes taste more important because it can produce almost unlimited options. When generation is cheap, selection becomes the work.

More output creates a filtering problem

Before generative tools, producing ten campaign concepts or interface alternatives required substantial time. Scarcity forced teams to commit early. Now a person can produce dozens of plausible versions in minutes.

This abundance feels like progress, but it can weaken decisions. Teams compare endless options, combine incompatible ideas and choose what looks polished rather than what serves the objective. The cost of creating another version approaches zero, while the cognitive cost of evaluating every version keeps rising.

Taste reduces the search space. A person with a clear understanding of the customer, brand and desired outcome can reject most possibilities quickly. They do not need to prove that an option is terrible; they can see that it is wrong for this problem.

Taste begins with attention

Good taste is not simply personal preference. It is built through sustained attention to excellent work and to the people being served.

A founder develops product taste by watching customers use the product, noticing where they hesitate and understanding which details create confidence. A writer develops taste by reading language that carries meaning without unnecessary performance. A leader develops organisational taste by observing the difference between a meeting that produces clarity and one that produces only activity.

The person must also study examples beyond their own company. Taste grows through comparison. When you examine excellent products, institutions and writing closely, you begin to understand the choices beneath the surface.

AI can show you examples, but it cannot perform the attentive living through which your standards become personal conviction.

The model averages; taste takes a position

Generative systems are exceptionally good at producing what is recognisable. They learn patterns and create outputs that resemble successful work. This makes them useful, but recognisability can drift towards sameness.

Taste allows a company to take a position. It decides that a product will be simpler even when competitors advertise more features, that a brand will speak plainly rather than imitate fashionable language, or that a financial service will sacrifice some speed to preserve trust.

These decisions cannot be produced solely by asking what is statistically common. They reflect values, context and the future the company wants to create.

A distinctive company will therefore use AI without surrendering authorship. The tool expands possibilities; the people decide which possibility fits the promise.

Taste is visible in subtraction

Average teams judge progress by what they add. More pages, more features, more campaigns and more explanations make the work appear substantial. People with taste often improve work by removing what does not belong.

Subtraction is difficult because every element usually has an advocate. The feature was requested by a customer, the sentence was written by a senior person and the process exists because of an old incident. Taste asks whether the whole remains coherent despite each local justification.

AI makes addition almost effortless, which means the discipline to remove becomes a competitive advantage. The best output may be the shortest draft produced after examining twenty longer ones. The best product may be the one that hides sophisticated complexity behind a few confident decisions.

Taste must be connected to consequences

There is a dangerous form of taste that is merely status. Leaders imitate the design, language or strategy of admired companies without asking whether those choices fit their own customers.

Real business taste is tested by outcomes. Does the product become easier to use? Does the message help the right customer understand the value? Does the decision strengthen trust and economic performance?

This does not mean every creative judgment can be reduced to a metric. Some choices shape long-term reputation and are difficult to attribute. However, taste should remain accountable to the purpose of the work rather than to the founder’s desire to appear sophisticated.

In African markets, this matters particularly because imported aesthetics can conceal local friction. A beautiful app that assumes uninterrupted internet or unfamiliar behaviour may win design praise while failing the user. Taste includes the sensitivity to design for reality.

Organisations can cultivate taste

Taste should not remain trapped inside the founder. If every piece of work requires the founder’s correction, the company has a dependency rather than a standard.

Leaders can cultivate taste by explaining decisions, not merely approving or rejecting work. Why is this option clearer? What customer truth does this message capture? Which unnecessary assumption does this flow remove? Over time, the team learns the principles beneath the preference.

Examples are useful. Maintain a collection of products, messages and experiences that represent the company’s standard, accompanied by notes explaining what makes them good. Review work together and discuss the trade-offs. Encourage people to defend choices with more than “I like it.”

AI can participate in this process by generating alternatives and critiques, but the organisation must decide what it wants to become.

Taste becomes a moral quality in important systems

In entertainment, poor taste may create forgettable work. In finance, healthcare, government or education, judgment affects dignity, access and trust.

Should a system exploit anxiety to increase engagement? Should an AI assistant sound certain when the evidence is weak? Should a financial product make borrowing effortless while hiding the consequences? These are design decisions, but they are also moral decisions.

Taste at this level includes restraint. It recognises that something can be technically possible and commercially attractive while still being unworthy of the company.

When everyone can make, choose better

AI will continue to improve, and the visible quality of average work will rise. More companies will have competent designs, polished language and functional software. Those things will remain necessary, but they will become less sufficient.

The moat will be the accumulated judgment behind the choices: the customer knowledge, standards, values and courage to reject what does not fit. Competitors can generate similar components, but reproducing the coherence of a company that knows what it believes is much harder.

Taste is not the last moat because machines will never produce beauty. It is the last moat because abundance makes preference unavoidable. Somebody must choose what enters the world under your name.

When anyone can generate anything, the advantage belongs to those who know what is worth keeping.