Creating a new perfume with ChatGPT: drilling down

This message is a continuation of my previous message which can also be found at https://www.perfumeprojects.com/resources/ai-01.shtml

The perfume formula in my last message was generated by ChatGPT. You could create that perfume directly from the formula. If you were to do so, with the intention of selling it, you would have to deal with several issues that ChatGPT warns you of. First, all of the materials used are naturals. Secondly, some of these materials can be expensive.

The problem with the naturals is that they can vary greatly depending on the source and the season when they were harvested. So, if you were planning to sell your fragrance, you would want to lay in a considerable supply of each material so that a second batch of your fragrance would retain the odor characteristics of the first batch. (Just a warning!)

Now, going back to ChatGPT, suppose you wanted suggestions for some substitutes for raw materials in the formula. I asked ChatGPT what I might substitute for Patchouli. It gave me this response:

"If you're looking for a substitute for patchouli to maintain a strong woody aroma in your perfume formula, you might consider using one of the following materials:"

"1. Guaiacwood: Guaiacwood oil has a rich, smoky, and woody aroma that can provide depth and warmth to your perfume. It's often used as a base note in woody and oriental fragrances."

"2. Cypriol (Nagarmotha): Cypriol oil has a distinct earthy and woody scent with slightly smoky undertones. It can add complexity and depth to your perfume, similar to patchouli."

"3. Cedarwood Atlas or Virginia: Cedarwood essential oils have a strong woody aroma that can enhance the woody character of your perfume. Cedarwood Atlas has a slightly sweeter and balsamic scent, while Cedarwood Virginia has a drier and more pronounced woody aroma."

"4. Oud (Agarwood): Oud oil is derived from agarwood and offers a luxurious and intense woody scent with resinous and slightly smoky facets. It's often used in high-end fragrances to add a rich and exotic woodiness."

"5. Amyris: Amyris essential oil, also known as West Indian Sandalwood, has a soft and creamy woody aroma with similarities to sandalwood. It can be a good substitute if you're looking for a milder woody note."

"Remember that the characteristics of these oils can vary depending on their origin and quality. It's a good idea to test your chosen substitute in small quantities to ensure it fits well within your overall perfume composition and maintains the desired woody profile."

Then I asked if there was a synthetic substitute for Patchouli and ChatGPT answered:

"Yes, there are synthetic aroma chemicals that are used to mimic the scent of patchouli. Patchouli oil is known for its distinct and earthy aroma, and synthetic fragrance compounds have been developed to replicate this scent for use in perfumes, cosmetics, and other products. One commonly used synthetic aroma chemical for patchouli is Patchouli Alcohol. It is a key component in creating patchouli-like fragrances and is often used to add depth and richness to various scents. While it might not perfectly replicate the complexity of natural patchouli oil, it can come quite close and offers the advantage of consistency in fragrance compositions. Other aroma chemicals and compounds can also be combined to create synthetic patchouli-like scents, allowing perfumers and fragrance formulators to achieve a desired aroma profile without relying solely on natural patchouli oil. Keep in mind that the field of fragrance chemistry is constantly evolving, and new aroma chemicals and formulations may have been developed since my last knowledge update in September 2021."

Not satisfied with this answer, I asked:

"is Patchouli Alcohol a natural or synthetic odorant "

And ChatGPT replied:

"Patchouli Alcohol is a natural odorant found in patchouli oil. It's one of the major components responsible for the characteristic scent of patchouli. While it can also be synthesized in a lab, it is initially derived from natural sources, making it both a natural compound and an aroma chemical that can be produced synthetically for various applications, including perfumery and cosmetics. "

Now here's the problem. ChatGPT can't tell you everything. It can only generate answers from what has gone into it. In time its knowledge base will surely expand and it will be able to answer questions that it can't answer now. Meanwhile, if you want to used ChatGPT to develop a new perfume, it's an excellent starting point for the non-professional and it can teach you a lot. It's suggestions can get you started in interesting directions... but the really hard work -- trying and testing and adjusting -- is all up to you.

Creating perfume with artificial intelligence (AI)

Let's suppose you have an idea for a perfume. You can describe what you're looking for in words but you don't know what raw materials would be needed to create this scent. You're (not yet!) a skilled perfumer. Artificial intelligence (AI) can help you create perfumes and teach you a good deal about aroma materials. The only prerequisite is your nose. You must be able to guide the AI engine and, once a raw formula is written, be able to make final adjustments yourself, following what your nose tells you.

The "perfume making tool" I have been using is ChatGPT. At the moment (August 2023) it can be used without any payment, its use if FREE, although that may change in the future. To use it you go to the website and register. Then you can begin to put it to work.

My first query failed because I asked it to write a formula for a perfume using materials from a particular supplier. ChatGPT replied "I'm sorry, but I don't have access to specific databases..." but it offered to guide me through "the general process of creating a perfume" with the scent I wanted "using commonly available perfume materials." So I rephrased my question, eliminating a requirement that the materials were sourced from a particular vendor and instead asked:

"write a formula for a new womans perfume with a strong woody note"

ChatGPT now presented me with a formula I could "consider":

Top Notes:

• Bergamot: 10%
• Pink Pepper: 5%
• Mandarin: 5%

Heart Notes:

• Cedarwood: 20%
Sandalwood: 15%
• Jasmine: 10%

Base Notes:

• Patchouli: 15%
• Vetiver: 10%
• Vanilla: 10%

Then ChatGPT noted:

"This formula is just a starting point and can be adjusted based on your preferences and the materials you have available. The percentages represent the relative strength of each ingredient in the final blend. You may need to dilute essential oils or fragrance oils to achieve the desired concentration.
Remember that perfumery is an art, and experimenting with different proportions and combinations will help you create a unique and captivating scent. It's a good idea to start with small test batches to refine your formula before producing larger quantities."

Of course this is just a starting point but, using ChatGPT, you can drill down. For example, you can ask what substitutes might be appropriate for each or any of the materials in the formula. I'll give you an example in my next post.

Make a perfume for yourself

If you are developing a fragrance you intend to sell, it is helpful to be very sensitive to the tastes of your intended market. You want to develop a fragrance that will be pleasing to them and, hopefully, more pleasing than the alternatives that are available to them. What they might, and what you might, want aren't necessarily the same.

For example, I'm enchanted by a fragrance now called "Xotic." It's a strong fragrance with a wake-up quality and I like the smell of it in the morning. But I know, from experience, that this is not a fragrance women would rush to buy for their men and a large number of their men, thinking of their women, would likely shun it too. Thus is doesn't have strong commercial appeal.

If you want suggestions for a commercial creation, the logical starting inspiration is the perfume counter of a department store or specialty retailer. Smelling the scents that are being mass marketed will help define the boundaries of your own creation. Professional perfumers developing commercial fragrances themselves take this approach.

But suppose that, for once in your life at least, you want to design a fragrance for a market of one: yourself. This can be an exciting project.

Recently I had a fix-it project that called for the use of a small amount of a solvent. I'm away from home for the summer, away from all those bottles of alcohol on the shelf, but I had on my desk in front of me a bottle of one of my fragrances, one that I had decided I didn't like very much and had already reformulated, considering the original composition junk.

So, being junk and about 90 percent alcohol, I decided to use it for the solvent I needed. I put some paper down on the desk and poured a bit of the fragrance, wetting the paper slightly. The room was immediately filled with the fragrance. I enjoyed the smell.

A piece of the paper remained on my desk for several weeks, scenting the room. What I had discarded as junk I was now finding quite pleasing. But it wasn't a fragrance I would try to sell. It was just a scent that, for me, was uplifting and pleasant.

If you decide to make a perfume just for yourself, without any commercial intentions, there are two obstacles, two hurdles to be overcome. First, you have to decide what you want your fragrance to be, what aroma you want. You have to be able to fix this is your mind -- which can be very difficult as you are dealing with an intangible. Some people have better scent imagination than others, just as some people have more visual imagination than others. It's something you can work on.

The second obstacle is technology. Once you have that scent fixed in your mind, how are you going to create it? You need the raw materials. Currently the professional perfumer has more than 3,000 potential materials to work with but a familiarity with as few as fifty can give you many possibilities. As a practical matter, you may want to adjust your perfect personal scent, the perfume you are going to make for yourself, to a mental image that will fit within the bounds of the materials you have on hand and understand.

Of course once you've created one perfume that is "just for yourself," it's likely you'll want to expand your knowledge of raw materials, the building blocks of perfumery, so that you can expand you imagined scents ... and create them.

Perfume marketing opportunity: Offer what others refuse to create

Two days ago I began to explore possibilities for a new perfume. Although I could be spending more time promoting fragrances I've already created, since the summer I've been sidetracked by a project having little to do with perfume. But suddenly this week perfume was again very much on my mind. I'll tell you why.

Recently I received some unexpected compliments on a perfume I had created some time ago and then largely ignored. Positive feedback provides inspiration and I was inspired to do some new smelling. I now had a rough idea of what I wanted but I had to do some mixing and sniffing to see if my vague idea might come alive. My first try was a strike out but my second offered possibilities so I'm letting it sit for a few days to see how it evolves and whether, with some work, it might become something. As I was working and sniffing a thought came to me which I would like to share.

There is a whole world out there of glorious smells, of potential perfumes, that you will never find at any perfume counter. This near infinite world of possibilities offers you unlimited opportunities to latch onto one of these totally unclaimed smells and make it your own. Let me explain this.

There are currently available not just hundreds but thousands of odorants sold for use in perfumery. Some don't smell like much; some might even be said to stink... until they are blended with other odorants. Then magic occurs and you have a beautiful scent. Professional perfumers understand what scents can be produced by blending two or more materials. They don't have to experiment. They know from their training. The work they do produces very subtle scents such as those found in bottles at perfume counters. But there are hundreds of aroma materials out there that smell good, all by themselves, without any blending. This is the key to possibilities for you.

Explore the raw materials of perfumery: the naturals, the synthetics, the specialty bases. Single out materials that smell good by themselves. Then play with them, mixing them with other nice smelling or not so nice smelling odorants. This is not going to give you a finished perfume but it can get you started on a scent theme that will be quite different than anything currently found at perfume counters. Remember, what you find in stores, on perfume counters, is but a tiny sampling of possibilities. In fact, each of these professionally created fragrances has been assembled with multiple raw materials and simply by changing the balance of these materials a new perfume is created.

For the most part, the professionally composed fragrances you find in the market were created in response to and in competition with existing professionally created fragrances already on the market. For the most part the investors in perfume will try to find spots in the market not too far different in scent than existing fragrances, especially the successful ones.

When creating a fragrance of your own, you have two paths. You can try to imitate what has already been done, particularly that which has already been done successfully, or you can follow your own nose to beauty without reference to others.

For the less skilled and I include myself, this second path is rewarding creatively and it can be more rewarding in profit for you are now selling an original story. You are selling something that you created, that you believe in, and that you want to share with others.

Think about it.

Do you wear perfume to the beach?

Smells surround us. We live in a world of smells. We are exposed to the strongest smells in the kitchen -- and in taking out the garbage. But can we ever walk down the street without experiencing a cornucopia of smells? The smell of a big truck; the smell of a motorbike, the smell of perfume, the smell of a dog -- and all the various random smells we can or cannot identify. Do you wear perfume to the beach?

Think about beach smells. The smells on a lonely, ocean-washed beach are quite different than the smells of a beach crowded with vacationers. The smells of a virgin lake are quite different than the smells of the ocean -- and far different than the smells of a lake favored by powerboats or freighters. Would these differences inspire you to wear different perfumes?

Do you think about environment when you select a perfume?

Think about a near-deserted beach along an ocean and its smells. Here would you wear a perfume that harmonizes with the ocean smells or would you deliberately or not wear a perfume that stands out boldly from the environment? The same question could be asked when planning for a social evening with fine dining and dancing. Do you select a perfume that harmonizes with the scent of the food you will be eating or one that rises above the meal and the crowd and identifies you out of all the happy mass gathered?

Perfumers of course have these thoughts. It's a little like selecting the right wine for the right course but far more delicate and important. The perfumer knows that most will not be paying attention to their environments but still feels an obligation to guide them so that their decisions are not too outré. Fragrances for the masses are engineered to go anywhere.

Thinking about the daily smells you smell, recognizing the odd, unexpected smell helps you think about perfume compositions. Building a new fragrance? How do you want it to fit into a man or woman's life? Will it be for special, controlled occasions or will it be for "all round" use? Think about how much recognition your audience has of the smells that surround them. How aware are they? What new smell -- in the form of your fragrance -- could make their lives more joyous? A tiny whiff of fragrance can have a very large influence on our happi9ness... and even our well being.

There's a formula behind every perfume you'll ever make. Too bad if you didn't record it.

How do YOU go about making a new perfume? Do you sketch out a plan, do you imagine a theme, or do you just plunge into it, mixing materials that seem to go together nicely? Do you find yourself creating many perfumes in small batches but never making the same perfume twice, even when everyone is asking you for it? To make more you need the formula.

There is a formula for every perfume you've ever made, even if you didn't record it. Perhaps the word "formula" is scary. Perhaps there's a fear that writing a few notes won't qualify as a formula and won't really be enough to bring the favorites you've made back to life again. But all it takes to bring a perfume of yours back to life in a new batch is incredibly simple, if you took a few important notes.

For example, say you created your fragrance using materials A, B, C, D, and E. Counting our drops, your initial formula might look something like this:

Then, if we were to assume that all the drops were of the same size (volume), we could go a step farther and produce a formula by percentages. It would look like this:

Now, to make more, you just mix the percentages. For example, to make 500 ml of your perfume oil:

When using your formula to produce more of your fragrance, you can use any unit of liquid measurement you want: milliliters, fluid ounces, liters, or even gallons. It's up to you.

If you have just made the original batch of your fragrance, nothing more need be done. But if some time has gone by, there is an issue you will have to deal with, being able to get more of the materials you used in your original batch. This brings up a second issue: keeping a record of the materials you used in your original fragrance.

If you are serious about your creative work in perfumery you will keep a very precise record of those materials including where you got them and even stock numbers. I've written a bit more about this in my book How to create an international production formula for your homemade perfume that will help you develop even more accurate formulas. The more accurate your formula is, the closer your new batch will be to your original. But it's never more complicated than just taking a few simple notes. You just want to be sure you're taking the right notes!

The detail you can't leave out when you're writing your formula

Anyone who makes perfume, even the most skilled and successful professional, has those times when he or she wants to experiment, just like a pure hobbyist. It generally starts with an idea. Then some aroma materials are chosen and the mixing and experimenting begins. Sometimes nothing comes of it but sometimes the result is pure magic. Can the magic be repeated? Can what has been done be done again? It all comes down to the formula.

Having the formula should allow you to produce more of that "bingo" fragrance you just created. For the professional it would rarely be a problem. But what about you? To produce more of the same, not only do you need the formula, you need the exact same aroma materials you used in your original batch. Can you get them? This is where the hobby perfumer so often gets stuck.

The issues:

To get more of the same materials you have to know exactly what they were; then you have to be able to obtain more. The first issue is the precise identity of what you used. Did you keep a full and complete record including vendor and vendor's stock number? Bergamot essential oil? Pure Bergamot or a Bergamot blend (the label should show). And whose Bergamot did you use? It makes a difference.

Can your source supply more of exactly what you used the first time? Does your source still sell it or have they switched to a substitute? Have they simply dropped it from their line leaving you to search for a substitute from another vendor?

Then there's the scenario where you were really careless in recording your materials and find something which seems to carry the same name but when you open the bottle and sniff, it turns out to be nothing like what you expected. (I have had this experience and it's a real wake-up call to keep better records.)

If you can't obtain more of something that went into your fragrance, that's the end of it. Finished. There will never be more. You can hope to come up with sometime inspired by the memory of that magical fragrance but it will never be the same.

Consider your sources before you buy your materials. Seek out vendors who are likely to offer continuity, vendors who understand the needs of serious perfume makers. And remember, if you are experimenting, developing a new perfume, you may run out of one or more materials you are using before you get it right. I have. If, at that point, you can't get more, your whole project is in the toilet.