Somewhere in a laboratory that looks nothing like a perfumer's studio, no organ stacked with hundreds of brown bottles, no blotters fanning across a desk like a paper peacock, no stained leather apron hanging on a hook behind the door, a machine is composing a fragrance. The machine does not smell anything. It has no nose. It has no opinion about whether vetiver pairs well with grapefruit, no instinct about whether a composition needs more lift in the top or more warmth in the base. It has data. It has approximately four hundred thousand formulas from the past century, digitized and tagged with consumer panel scores, sales figures, regional preferences, and molecular descriptors. It has an algorithm trained to identify the statistical correlations between specific ingredient combinations and specific consumer outcomes, purchase intent, perceived quality, emotional association, likelihood of repurchase. And it has been asked to produce a formula that will be, by every measurable standard, optimal.
11 min read
It will succeed. The formula it produces will score well in consumer panels. It will test positively across multiple demographics. It will not offend anyone. It will not confuse anyone. It will occupy a comfortable, well-populated region of olfactory space, the kind of territory that the industry calls "commercial sweet spot" and that anyone with a functioning nose calls "familiar." It will smell, in the judgment of most people who encounter it, perfectly fine.
The question is whether "perfectly fine" is perfumery.
Machine learning at industrial scale in fragrance
The application of machine learning to fragrance development is not speculative. It is happening now, at industrial scale, in the research divisions of the world's largest fragrance and flavor companies. The technology varies in sophistication, some systems are relatively simple predictive models that suggest ingredient substitutions based on cost and availability; others are deep neural networks trained on decades of proprietary formula data, but the underlying logic is the same in all cases. Feed the machine a large corpus of existing formulas paired with consumer response data. Let the machine learn the statistical relationships between molecular composition and human preference. Then ask the machine to generate new formulas that maximize the probability of a desired consumer outcome.
This is, in essence, regression analysis applied to perfumery. It is not, in any meaningful sense, creation.
The distinction matters, and precision about why is necessary. Regression analysis, the mathematical technique at the heart of most machine learning, finds the line of best fit through a cloud of data points. It identifies the central tendency. It tells you where the average is. This is enormously useful for many applications. If you want to predict house prices, consumer behavior, disease trajectories, or election outcomes, knowing where the average is tells you a great deal. But perfumery is not a prediction problem. It is, or at least it has historically been, a creative problem. And creative problems are not solved by finding the center. They are solved by finding the edge.
Every fragrance that has genuinely changed the industry, every composition that, in retrospect, defined an era or opened a new category, did so by deviating from the consensus of its time. The first modern fougere was not what anyone expected a men's fragrance to smell like in 1882. The first great aldehyde floral was not what anyone expected a women's fragrance to smell like in the 1920s. The first dihydromyrcenol-and-hedione fresh masculine was not what anyone expected from a men's fragrance in the 1980s. The first clean-skin-musk molecular was not what anyone expected from any fragrance in the early 2000s. In each case, the composition succeeded not because it matched existing preferences but because it created new ones. It did not find the center. It moved the center.
An algorithm trained on historical data cannot, by construction, move the center. It can only find it. It can find it with surgical precision, and it can generate formulas that occupy the sweet spot with an efficiency that no human perfumer could match. But occupying the sweet spot is not innovation. It is optimization. And the history of perfumery suggests that optimization and innovation are not the same thing, and may in fact be opposed.
Human perfumers as biological neural networks
A counterargument, and it deserves serious consideration. The counterargument goes like this: human perfumers are also, in a sense, algorithms. They are biological neural networks trained on a corpus of olfactory data, everything they have smelled, every formula they have studied, every consumer response they have observed over the course of a career. Their creative process is not, as romantics like to imagine, a bolt of inspiration descending from the muse. It is pattern recognition, recombination, and iterative refinement. The perfumer sits at the organ, selects materials based on experience and intuition, blends a trial formula, evaluates it, adjusts, evaluates again. The process is empirical, not mystical. If a machine can perform the same operations faster and more systematically, what, exactly, is lost?
What is lost is the error.
This sounds paradoxical, so let me be specific. Human perfumers make mistakes. They over-dose an ingredient and discover that the over-dose creates an effect they did not intend and could not have predicted. They accidentally contaminate a trial batch and find that the contaminant adds something interesting. The history of synthetic breakthroughs in perfumery is littered with such happy collisions. They misread their own notes and combine materials they did not mean to combine, and the result is better than what they planned. The history of perfumery is littered with these accidents, compositions that owe their character not to deliberate design but to some unplanned collision of materials that a more careful process would have prevented.
An algorithm does not make these mistakes. An algorithm does exactly what it is told to do. It optimizes the objective function. It follows the gradient. It does not wander into uncharted territory by accident, because it does not wander at all. It moves, with mathematical precision, toward the optimum. And the optimum, as defined by consumer panel data, is always the center. The average. The consensus.
The creative potential of error is not a romantic conceit. It is a well-documented phenomenon in every creative field. The biologist who discovers penicillin because of a contaminated petri dish. The physicist who discovers cosmic microwave background radiation because of unexplained noise in an antenna. The musician who discovers a new harmonic language because a string broke mid-performance and forced an improvisation. These are not apocryphal stories told to comfort the accident-prone. They are documented instances of a general principle: creative breakthroughs often originate in deviations from the plan, and systems designed to eliminate deviation will, by construction, also eliminate the possibility of breakthrough.
Consumer panel data and the nature of preference
A second, more philosophical objection to computational perfumery, and it has to do with the nature of preference itself.
Consumer panel data, the data on which these algorithms are trained, measures stated preference. It records what people say they like when asked. But stated preference and actual preference are not the same thing. Stated preference is conservative. When asked to choose between the familiar and the unfamiliar, most people, in most contexts, choose the familiar. This is not stupidity. It is a well-documented cognitive bias, the mere exposure effect, first described by the psychologist Robert Zajonc in a landmark 1968 paper in the Journal of Personality and Social Psychology, and it operates powerfully in olfactory evaluation, where the absence of a shared vocabulary makes it exceptionally difficult for consumers to articulate why they like or dislike something. Faced with a fragrance that is genuinely novel, that does not map onto any existing category, that confuses and intrigues in equal measure, a consumer panel will, more often than not, give it a low score. Not because the fragrance is bad, but because the panel lacks the framework to evaluate it.
An algorithm trained on consumer panel data inherits this conservatism. It learns that novelty is risky and familiarity is safe. It learns that the fragrances people rate highest are the ones that most closely resemble fragrances they have rated highly before. It learns, in short, the most fundamental lesson of consumer research: people like what they already like. And it optimizes accordingly.
The result is a machine that is supremely good at producing what the industry calls "safe bets", fragrances that will not fail, that will achieve a minimum level of commercial viability, that will not surprise, disturb, or challenge anyone who smells them. These fragrances will sell. Some of them will sell very well. But they will not change the industry, because changing the industry requires producing something that consumer panels do not know how to score. The compositions that changed perfumery were all, at the moment of their creation, surprises. They were things that no one had asked for, things that did not score well in preliminary testing, things that succeeded not because the data said they would but because a single person, a perfumer, a creative director, an entrepreneur, believed in them despite the data.
An algorithm cannot believe in anything despite the data. Believing despite the data is the one thing an algorithm is constitutionally incapable of doing. An algorithm follows the data. That is its virtue and its limitation. And in a field where the most important decisions are the ones that contradict the data, where the entire history of creative advancement is a history of people ignoring the consensus and being proven right, this limitation is not minor. It is fundamental.
Where AI belongs and where it does not
Let me be clear about what I am not arguing. I am not arguing that artificial intelligence has no role in perfumery. It has obvious and valuable applications. It can accelerate the reformulation process when a regulatory change forces the removal of a restricted ingredient. It can suggest cost-effective substitutions that maintain a composition's character while reducing its price. It can analyze large datasets of consumer feedback and identify trends that a human analyst might miss. It can map the vast, multi-dimensional space of possible ingredient combinations and highlight regions that human perfumers have not yet explored, much as gas chromatography once decoded formulas that were previously locked in trade secrecy. These are useful functions. They save time, reduce cost, and expand the perfumer's toolkit. No serious person objects to them.
What I am arguing is that these are all optimization functions. They make existing processes more efficient. They do not create. The distinction between optimizing and creating is not semantic. It is the distinction between finding the best route through a known landscape and discovering that the landscape extends beyond its known borders. Machine learning excels at the former. It is structurally incapable of the latter, because the latter requires, by definition, going beyond the data, and machine learning is, by definition, a method for extracting patterns from data.
The fragrance industry's enthusiasm for computational tools is understandable. The economics of modern perfumery are brutal. The average development timeline for a commercial fragrance has been compressed from years to months. Briefs are tighter. Budgets are smaller. The cost of failure is higher. In this environment, a tool that can reduce the number of iterations needed to reach an acceptable formula is enormously valuable. But "acceptable" is doing a lot of work in that sentence. An acceptable formula is one that meets the brief, scores adequately in testing, and does not exceed the cost ceiling. An acceptable formula is not a masterpiece. It is not even, in most cases, particularly interesting. It is adequate. And adequacy, at industrial scale, is the enemy of art.
Convergence toward a statistical mean
There is one final consideration, and it is perhaps the most troubling. The more that fragrance development relies on algorithmic tools trained on consumer data, the more the industry's output will converge toward a statistical mean. Each new AI-optimized formula will, by design, occupy the center of the preference distribution. Over time, the center itself will shift, but it will shift slowly, because the algorithm's output reinforces the very preferences it was trained on. Consumers who are repeatedly exposed to AI-optimized fragrances will develop preferences shaped by those fragrances, and those preferences will in turn become the training data for the next generation of algorithms. The result is a feedback loop: the machine produces what people like, people learn to like what the machine produces, and the machine produces more of it.
This is not a hypothetical scenario. It is a precise description of what has already happened in other creative industries that have adopted algorithmic recommendation and generation systems. Music streaming platforms, whose algorithms optimize for engagement, have produced a measurable convergence in the sonic characteristics of popular music, louder, shorter, more repetitive, with the hook arriving earlier and the dynamic range narrowing. Social media platforms, whose algorithms optimize for attention, have produced a convergence in the visual characteristics of popular content, more saturated, more cropped, more emotionally extreme. The algorithm does not flatten the landscape deliberately. It flattens it as a side effect of optimizing for the average.
Perfumery is not immune to this dynamic. If the industry's development pipeline becomes increasingly dependent on AI tools that optimize for consensus, the inevitable result is a narrowing of the olfactory field. Not a narrowing to a single scent, the market is too large and too segmented for that, but a narrowing within each segment. The fresh masculines will converge. The sweet feminines will converge. The oud orientals will converge. Each category will become more internally homogeneous, because the algorithm that designs each new entry is being trained on the same data that produced the existing entries. The field will not shrink to a point. It will shrink to a cluster.
Whether this matters depends on what you think perfumery is for. If it is an industry, a business that produces consumer goods designed to meet market demand, then optimization is the correct strategy, and convergence is an acceptable cost. Consumers get what they want. Companies make money. No one complains.
But if perfumery is also an art, a creative discipline whose purpose extends beyond satisfying existing preferences but to reveal new possibilities of olfactory experience, then convergence is not a cost. It is a catastrophe. Because art, by any definition worth defending, requires the possibility of surprise. It requires the possibility that the next composition will be something that no one has smelled before, something that no data set predicted, something that a consumer panel would have rejected because it did not fit any existing category.
An algorithm cannot produce that. A perfumer can. Not reliably, not consistently, not on schedule, not within budget. But occasionally, unpredictably, against all commercial logic, a human being sitting at an organ surrounded by hundreds of brown bottles will combine materials in a way that no machine would have suggested, and the result will be something genuinely new. Something that moves the center rather than occupying it. Something that the data said should not work.
Those moments are rare. They are getting rarer. And if the industry is not careful, they will stop happening altogether, not because the technology forbids them, but because the economics no longer leave room for them. The machine will compose. The machine will optimize. The machine will produce perfectly fine fragrances that score well in every panel and offend no one.
Whether that counts as perfumery is a question the machine is not equipped to answer. It will have to be answered by a nose.
Seven extraits at 20%, one collection. The Discovery Set carries all seven in 2 ml.
