Talk & transcript
Olfactory RGB
Sean Raspet
Transcript
Download textAutomatically generated transcript. May contain errors. Select a timestamp to watch that moment on YouTube.
[Music]
[Music] Good morning.
What should I do? Well, it's like it's right. So, if it gets worse,
[Music] [Music]
[Music] Hello. Bailey.
[Music]
Do you do you guys want to help pass out um the people who can help train you? Okay. Anyone? Anyone? Please call
[Music] So basically [Music] [Music] [Music]
[Music]
[Music] [Music] 10. Okay.
What are we doing?
All
right. [Music] [Music]
[Music] right now. Yeah. Yeah. [Music] Yeah. [Music]
Okay. [Music]
Yeah, I'm ready. I I'll talk like [Music] [Music]
was [Music]
What [Music] are you talking about?
[Music] [Music] This is here. [Music] like [Music]
Wi-Fi. [Music] Oh, one last Oh,
there he is.
All
right, we're ready to start. Three, two, one. Yay. Congratulations, everyone, for being together here tonight. It's such a wonderful night. There's no rain, which is incredible. It's so cool. Okay, so we have a packed house tonight. It smells okay for given the size, but hopefully we have some samples coming um very shortly which are going to improve the smell of the uh of the thing. And also it's a small room, so uh if you see your neighbor chatting throughout the thing, feel free to tap them on the shoulder and try and get them to quiet
down. Okay, so this is Dside NYC. It is the biggest and best DI meetup in New York. And what is decentralized science? If you want to find out more, you can learn more on our website which is here or which is on our Twitter which is here. So you can follow us. If you take if you have any pictures, tag us here. Uh Instagram, tag us here.
All right, step two. So this is one of the events we've done. We've had over 10,000 people um come to each of our events. We've had a lot of you are PhD students, investors, crypto people, uh undergrads, um perfume enthusiasts. I was looking at the uh the invite and the Luma and it literally like 10 to 15% of people here are like specialists in the perfume industry and I had no idea that there is that many people in New York uh who knew about that. So, that's super cool.
Um, we've had a lot of speakers here. Okay, we have some upcoming events. Um, the ne we've there's been a lot of DSI events recently. There's been like three in a row for the past 3 weeks. So, we're kind of a little exhausted, but um the next one it'll be in a couple weeks and it's on um nutrient transport and cell metabolism. This is a really interesting and pertinent topic. Um and a lot of people are talking about the metabolism cell metabolism as being one of the most important things for longevity. Um Casey Means who is now the um in charge of health in the government talks a lot about this. So this was going to be a very interesting talk. If you're interested you can RSVP here.
Couple more people have their phone out. I'll give it a second. So that will be very very cool. Oh, also while I take this opportunity while people are taking a picture is if you have ideas um on who would be a good speaker for DINYC in the future um then you should let me know because there's only so many people that we can think of and at some point we have to crowdsource it and get more information from other people. So, if you have ideas on who would be a good speaker, um, ping me, ping ping, uh, this here, send us a DM on Twitter on Twitter and, uh, we'll we'll investigate.
Um, after that, this one's really really cool. So, we had a talk on superb babies um by Jonathan Anomaly and that was where you could screen children for different scores to see how smart they were going to be, the color of their eye, predispositions to different types of genetic diseases. But this is even taking it a step further where you can instead of just screening for uh different things, you can edit um edit what edit this the genome and so then you can actually modify it which is going to be very very interesting.
Um, I'm just going to jump in right quickly so that you can do this while we're doing it. We're conducting one of the largest sleep studies. Um, probably we have like 5,000 people a night sleeping. If you guys want to try and contribute to the sleep study, we will give you this beautiful sleep mask later on. Um, they actually really help quite a bit um in sleeping. So, if you come up to me after, um, you can download the sleep app, which I'm going to talk about in a little bit here or here, depending on if you have Android or iOS, and I will give you this wonderful sleep mask, and you can be one of 5,000 people that are currently sleeping each night.
It's actually really cool. So, I'm going to talk about it in a second. And then also, um, I just want to put it out here. We're growing and we're having a lot of users. So, if you're looking for a job as an engineer or a designer or a product person or a social media person, um, scan this QR code on the far left. Um, we're we're hiring and we're looking for like talented people that are really passionate about science or passionate about design or passionate about uh, computer science and uh, we have a lot of cool thing job opportunities there.
I'll I'll show that again in a second, but uh, I'll just we have some people who are still taking pictures, so uh, I'll leave it there. Any questions? Okay, so a lot of people say what is distributed science like how does that differ from normal science and the way it works is like distributed science is where anybody can participate cuz we're seeing a lot of places where part where science is done in academia and in the university but there's a lot of things changing right so uh universities are getting defunded universities are having a lot of PhD students leave um the types of research Arch that is being done in a university setting is also changing a
lot. Although people think it's not as interesting. So uh what we're trying to do is uh use the crowdsource research crowdsource data crowdsource contributions to data and then give people equity in that. So if someone does a lot to make a research project successful they earn crypto for that. That's sort of at the high level of how it works. So what is one of the things that a lot of people could contribute to each night is sleep. A lot of people want everyone sleeps. Everyone has a lot of sleep data and right now a lot of it's siloed within Aura Ring, Whoop, Garmin, and that type of thing. So, we're we're doing a study where people are contributing their sleep data and earning crypto in exchange for it.
And we're not sort of the first ones to do this. um the open a uh during sleep 8 during when Sam Alman was fired uh did this sleep study where they said that they checked and there was a 20% increase in people getting under 5 hours of sleep during the open AI drama. So this is just sort of like the first step in looking at ways in which you can use these distributed devices um to conduct new types of of of research.
Alan in the back we can hear you up here Alain. Alain sorry. Okay. Um and then there's other things that you can do with exercise and sleep. So, uh you people are Whoop is doing all this research into uh understanding uh how how exercise impacts sleep. And they found that the closer to exercise, the more the less the more exercise someone does when they're closer to getting to bed, the less well they sleep. But this really is a new type of research where it's using wearables, distributed um lots of people outside of a lab stud setting to to do these new types of research findings.
Um so the way our study works is it works with any wearable, any phone, any device. So if you have a Whoop, Garmin, or a ring, it works. But if you also if you just have your normal phone, we can derive a sleep score just from that by checking to see when the person last uses the phone and first uses the phone in the morning. So it really is uh accessible to a lot of people.
Um and then as I said for each each data contribution you earn equity in the research project. And this is like a very different way of thinking about research. It's kind of like getting your name on the paper. Um but instead of that you're earn you're getting you're earning crypto for this specific crypto token that's related to this research project for based on your contribution. So it's kind of like a game. As you contribute to the science project, you're earning points and you're climbing the leaderboard.
So, you can track your points, you can climb the leaderboard, you can share with other people. And then this really interesting thing is we're we're starting to do these studies. So, we're looking at these minor interventions like sleep masks, like melatonin, like red light glasses and seeing how they affect different types of people in different ways. So, one of the interesting things that we're finding so far is everyone is very different. So there is no oneizefits-all for for sleep or I think in generally for medicine. A lot of it should actually n equal one sort of stuff. So um we're starting to do these different studies where in the morning you fill out when was the last time you had caffeine? Did you have a sleep mask? Did you um did you have
alcohol last night? And we're we're building these different data sets that are that are that is that's very interesting. I hope to share more once we compile the data into interest in in compile the data. Um one of the things that we found already is that um the sleep quality improves a lot depending on when the last time you ate is. So the further away you can you can um put your the last meal you eat um in relation to when you go to bed, the better you'll generally sleep.
So again, um if you'd like to participate in the study and be sort of the 5,000th in one, um here's the different things. We'll happily we're give you this sleep mask so that you can participate in trying out these different things as well. Um and also if you would like to uh work if this is really interesting to you, we've definitely hired a couple people from this audience in the past. So people do come. We there's a lot of smart people here. So if you're one of those people, um consider applying here. And uh thank you very much. If there's any questions, let me know.
Does anyone have any question? Yes. Go ahead. Are you using the data for this study or are you planning on selling them? Just for this study. It's just for science, right? For just for science. Yeah. Uh, anyone else? Three. Oh, yeah. Go ahead. Is there an age or uh I think just as long as you're over 18 it's fine. Is there within GCP within GCP practice right there's a rule there's a whole set of rules of good practice. Uh, I think that's if you're doing like a So, what's interesting is this isn't like an official clinical
trial, right? This is like more like something that's more similar to a Reddit thread where people are contributing data as they would to a Reddit system. So, on any Reddit thread, whether it be peptides, whether it be um sleep, whether it be, you know, whatever disease you have, r/ that has a lot of people doing interesting research to contribute data on that. And this is what we're trying to do is contrib.
Um, and so it's not a clinical trial. It's more like an internet forum. And and we also don't tell people what to do really. We let people contribute and tell us what they did. So we don't we don't want to be uh tea telling people how to improve it but more uh collecting data for when people experiment at the at the edges and seeing how that changes the and seeing how that changes types of data that produces.
Cool. Okay. So without further ado, we will go on to the main speaker tonight um Sean. And Shawn is an Olafactory expert um and does some really interesting work on the intersection of AI and smell. And I'll let you take it from here. Thank you very much. Very great turnout. Thank everyone for coming. Uh oh wait, let's let's get a let's get a picture. Okay, everyone uh smile.
Three, two, one. [Music] Three, two, one. And then one this way just cool. Um so yeah, so I um have been studying kind of basically like chemical structure and the kind of oldactory impression it has. to what they call structured odor relationships for about 12 years. Um, and I actually got to it from I don't know if I can speak any louder than this. Um, closer to your mouth. Okay. Got it. Okay. Okay. Thanks.
Um, and so yeah, I've been uh studying that for about 12 years. And I got into it actually through art. So I first I studied art and then I at some point got a little bit like bored I guess with it uh with visual art and you know it kind of seemed like everything if you look through art history like everything's been done already. Um it's not like there's new colors that you can invent.
Uh and so gradually that led me to you know become very interested in flavor and fragrance chemistry uh and kind of structure of relationships uh as a kind of different space a new space for sensory impression and a new kind of medium for art. Um and in particular I'll go into a little bit of like some of my I guess journey to to get there but uh in particular what I find really interesting is designing new molecules designing new smells. Um it is an area where you can create a new color as it were. Um and and that's you know pretty fun and exciting and in fact we'll have a lot of we'll have about eight different molecules for you all to smell
uh pretty soon. So um yeah go to the next one. Okay. Yeah. So just briefly um you know since I I've got into the area I've also done a lot of work in uh the food industry and to some extent in fragrance. Uh this is a and I saw this as part of my artwork actually um to to basically like have an artwork that circulates in the economy at large uh rather than something that's just on a on a wall. Uh and I don't know maybe we have some some art people here maybe not. Hopefully I'm not boring you. Um you know if you're if that's not really of interest but for me it's it's something where I thought okay you can have an expanded definition of art. It doesn't have to be a painting. It could
be a product. Could be something that you drink that gives you nutrition. Uh this is a product that I did when I was at Soilent. I was doing flavor um you know work for them and um it's called nectar. It's actually a um two molecules. The flavor of it is two molecules. It's citral and geranial and those are in a certain ratio uh 2 to 2:1 and it's basically that is a synthetic version of what they call the nasanog pheromone. So the nasanog pheromone is released by honeybees when they mark food. Uh so when they're kind of foraging and looking for, you know, food sources, looking for um you know, flowers that have a lot of nectar, they'll release this the this uh pheromone. And so what I did was
basically take that pheromone and use it as a flavor. So it's kind of like in a way marketing food for humans um you know through this soy product. So that was kind of one um project that I did. And then yeah, more recently I've been at Simulate. So simulate basically simulates chicken. Uh we're also working on some uh beef products, but uh yeah, this is a fiber spun chicken breast uh that I worked on and did the flavor as well as led the team that did the texture. Um and then this is the the kind of fibers close up. Um those are about 100 micron fibers, so they very closely mimic the kind of muscle texture of chicken meat. Um so those are some
other kind of things I've done in in different uh companies. But what I'm most interested in talking about is my um I guess long-standing hobby of making new molecules and also more recently uh kind of startup called patina that I'm working on to design new molecules and kind of decode the sense of smell. Um so yeah, so we'll just go ahead and kind of get right into the molecule sampling. So what I'd love to do is just use you all as test subjects if you don't mind. Um, and you'll get to sample the molecules, but I'd love to get your feedback. So, if you scan this QR code, it'll take you to a thing called Slido. Um, and if you have any problems with the QR code, you
can go to slido.com and then just put in uh this hashtag new molecules and you should be able to get to it from there. And then, so yeah, if the we have some very uh generous volunteers that are going to help hand out uh the scent on scent strips. And so, we'll start this. So this is going to be molecule one. And if you can't hear, come to this side. I know that the noise comes up from there.
So that you can actually hear very well over here. So um if you do have the QR code live, I'm just going to go ahead and um make that this part of the survey live. So basically the first part is just describe this the scent of this molecule in a few words. So you can add, you know, as many words as you want, but just kind of keep it keep it brief. Just like what are the key words of this sense?
And I guess for anyone that has the app open, let me know if you have any trouble with the app. It should be live, but just just let me know in case there's any issues. Should be gone. Do you guys smell anything or anything? It could be subtle. It could be subtle. Sometime so there's a thing called specific anosmia. So sometimes people don't smell certain smells and this one may be subtle. So and that's fine. If you don't smell anything then just say I don't smell anything. But you might have to smell it like a few times. Um,
no. This one's not a This one's not a throw. And I guess um so for I guess to try to try to um to try to make it efficient, we'll just if you get the sensit just like hand them out really quickly, like hand them to the next person so that we can um put in the information and then we'll go to the next one. And if you if you don't have a send strip, raise your hand and we'll try to send one your way. Who doesn't have one?
Okay. you try to get something. It's a it's a challenging layout for distributing these. So, yeah, thank you for your patience.
[Music]
All right. And uh so now I'm going to do the second part of this this kind of quiz which is so there's two things. So you'll give a rating of 1 to nine. Nine's the highest obviously um for pleasantness. So one is extremely unpleasant, five is neutral and nine is extremely pleasant. Um, and that the one through nine is just a kind a common kind of sensory score. Um, so I'm just using that. It's easy. Um, does that make sense? So, I'm going to make that part of the quiz live right now. And, uh, and then we'll go back to talk more about that molecule.
What's that? Yeah, you could. If you don't smell anything, it's up to you. You could say I didn't smell anything and that's an unpleasant thing because I wanted to smell something. Or you could say neutral. Neutral is fine. Um so like if it's neither pleasant or unpleasant then you would put a fine. And so yeah, first it asks you for pleasantness and then the next then it'll go to the next question which is the same thing. It's a rating of one to nine but it's for familiarity. So, if the molecule is kind of recognizable, if it reminds you of something, um, if it's
if it's very familiar and completely recognizable, you would give it a nine. Like, that's the, you know, I I know exactly what that is. If it's completely unfamiliar, you have no idea what it is, you've never smelled that kind of thing before, then it's a one. Um, so you'll see both of those questions when you go into the the quiz. Yeah. Yeah. And it's it's kind of developed from um you know some standard practice like the one to nine scale is standard practice. Um we use it a lot for flavor and like food product development. But uh for me the the question of pleasantness and familiarity. If I have to ask if I wanted to ask two questions and I can only get those two points those are the two most useful things. Um
yeah. Yeah. Usually it's like kind of like this. It'll be, you know, up to 80 people. Usually, sometimes 30. Um, yeah. The more the better, of course. Yeah. Yeah. Cool. All right. So, um, this molecule, I'll talk a little bit about it. With some of them, I'm not going to talk about them until the end, but with this one, I'll talk about it real quick. So, this is called isoe super. Um, it's one of the most widely used fragrance molecules uh by quantity. Uh, part of that is because it doesn't smell super strong. Uh, and it's it's not too expensive, so people can use a lot of it in a fragrance and um get a kind of softness or or kind of enhancing effect
to their fragrance. So, it's kind of easy it's easier to dose into the fragrance. Um, some molecules if they smell super strong, it's hard to find the right dose, you know, because if you go a little bit too high, it just, you know, it's too it's too much. So, um, anyway, this is, yeah, a very common molecule in the flavor and fragrance space. So this is kind of like a good uh almost like background um sample. So this will help kind of calibrate.
Okay. Sorry. Yeah. So this this will help kind of calibrate. This is kind of a good background sample. It'll help calibrate um you know the scores of the other molecules. All right. But this is not a molecule that I came up with. All right. So now I think we're ready for number two. See? And so I'll go ahead and make that live. So yeah, same same structure as the last time. It's um you know, describe the molecule in a few words, kind of the key words that come to mind and then we'll do the pleasantness and then the um
familiarity on the one scale of 1 to nine. Uh and then we have after molecule two, we have two more molecules. So, four molecules total for this round. Then I'll just do some more talking. Um, and then we'll at the very end there'll be another round of molecules and I'll also talk about the molecule structures along the way. And by the way, if anyone has questions just like feel free to raise your hand as long as there's not too many questions, you know, then we'll I'll try to get through all of them.
There you go.
[Music]
Okay. So, I'm going to assume everybody like had a chance to put in the keywords and then I'll move to the next part of the quiz. Yeah, definitely a lot of good words and give it a couple more seconds. The audio is not as good over there because the sound, but if you go over here, it's actually pretty good. All right, cool. So, let's do the pleasantness and familiarity. So, I'm gonna start that one.
By the way, can you see the results when you when you put your results in? No. Okay. Um, so then I'll read out some of the results, too. Yeah. Yeah. Yeah. If you buy like a a fragrance, you know, cologne, perfume. Yeah. Usually it's, you know, some of them use extracts from plants, but I think a majority, most people would say a majority of the um breakers composition is going to be these these single molecules that have been developed.
Is there like Yeah, definitely. Um yeah, similar to essential oil extracts um you know where rosa oil is much more expensive than um even jasmine or uh definitely a lot more expensive than like lavender in a similar way. Some of the molecules are more expensive. um it has to do with kind of like how hard they are to synthesize and also to some extent if they're captive molecules. So if they're under patent then you know only one manufacturer can or one company can legally manufacture them. So that's another factor. What about
Oh the the price you know I I think I mean fragrance it's kind of it's a luxury good in a lot of cases. So, the brand is, you know, probably more of a factor in what people will pay than the scent itself. And I I think they've probably done some some blind studies of like, you know, what people prefer if they don't know what the brand is. Um, so there is that aspect. But, you know, I think there are some like I guess brands that are known for like more interesting or adventurous fragrances and and you do see that, you know, sometimes where where others are going to be more kind of like run-of-the-mill.
Um, so there's there's differences in that way and and the ones that are run the mill might be a little bit lower price. Yeah. I'm just like wondering like Yeah. Yeah. Oh, I see. Like as I think as a as a perfumer it could be useful like if you wanted to have a unique note. So that's where actually captives come into the the kind of flavor and fragrance market a lot is you know generally it's like a large company that wants to have their uh formulas not be easy to copy because you know um right now you know you could generally you can break down like what a
formula is you can use GCMS and then a combination of that and just like smelling and trial by error to like you know get something that pretty much you know is is mostly indistinguishable if not 100%. So, um, the idea of captive molecules is it makes it much harder to do that because if this company that made this fragrance is the only one that can use this particular molecule that's in it, then it becomes harder to replicate.
Cool. Um, so we're on molecule two. Um, so I'm going to read out some of the descriptions now that we're we're pretty much done. Um, so here's what I have. And I unfortunately I can't show this right now cuz it's like I'm on two different computers, but um basically a lot a lot of people said floral, honey, rose, sweet, jasmine, pleasant, strong, uh popery, uh rose grass, helotrope, elmer's blue, smell, smell, flowers, fresh, rosy, floral, and faintly earthy, merry gold, kind of citrusy, kind of flowery. um and that oh and pollen and
gardinia. Um so yeah, that's that's the kind of um keywords that everybody came up with. So thank you for that. That's that's useful for me. Um cool. And then let's go to molecule 3 and then we just have one more after that and then you know we'll just chat some more. How would you describe it? You have like a set description. Oh okay. Yeah actually I should probably let me show you all the structure too.
Um, and yeah, so this this molecule, actually, that's a good segue. So, I'll talk about this molecule. So, basically, um, I call it hyperchlor, and it was a molecule that I developed where I was trying to make something almost like what I would call a neon floral or like something that's like so, um, so floral, so kind of planty that it almost seems like it's it's not natural or just kind of like pushing it into this this very extreme kind of space. Um, and that molecule is quite strong smelling. Um but you can see from this uh kind of image here it starts from phenino acid aldahhide which is something that is found in rose and it's known to be very like green um very somewhat floral but
just like very kind of green in a certain way not like green grass but a different kind of green like just like uh growing plants and uh so what I did is basically through a couple different reactions uh got to this structure. So first reacted the phenyl acid alahhide with a glycerol to form a glycerol acetal and then oxidized that to get to this uh hyperore. And the reason for oxidizing it is that um so when you have a glycerol acetal then you have an alcohol group and those tend to not be as strong um in in the resulting molecule than if you have either a ketone or an aldahhide. Um, so generally speaking, if you want to design a molecule that's going to be stronger,
uh, you go with a, um, ketone or aldahhide. And let's see, I guess, is everyone still uh, doing the survey? I'll come back to this in a second. Why did you make that decision? Oh, yeah. modification. Um, so from the glycerol acetal to the oxidation or from the just general concept of of like what what I was trying to make. Yeah. Like that particular modification. Yeah. So it kind of you know so it was kind of from a theory of like okay so phenol acid aldahhide is good but it's I
mean it's in the right direction generally but it's kind of like simple. It's kind of like one note and if you smell it in, you know, if you smell it kind of by itself, it can be a little bit like chemically. Um, and what I wanted to do is uh make it more complex. And so in in a sense like making the molecule more complex helps to make the smell more complex in a in a very kind of simplistic way. But also um having extra oxygen atoms can uh so you have the the kind of two oxo ether kind of uh things there and then you have the ketones. So having extra oxygen helps to helps with hydrogen bonding in the receptor. So if you if they're in close proximity to the ketone, they're not too
far away because then that will change the kind of um you know electron gradient of the molecule. But so if it's still in the same like directional basis, but then you're adding some more kind of uh oxygen groups, then you can make it stronger. making it stronger also tends to make it more complex because um it just is activating more receptors. Uh so so and basically when we talk about like um you know molecules and receptors it's usually like a given molecule is probably activating like multiple receptors and then that gives you like the smell and then any smell that you encounter in the world is probably like you know hundreds of types of molecules that are all working
together and then it forms this kind of average. Um, and an interesting thing about smell is like they say that you can smell about a trillion different scents. Uh, where then and you could distinguish that many different scents where with uh color vision you could distinguish probably a million or less. Um, so it's a several orders of magnitude, you know, more. So yeah. So that's interesting. Um, okay. So did did everybody get the molecule three yet? Oh, okay. Okay.
Yeah, let's do let's do molecule tree. Thanks. Yeah, I guess while we're passing that out, I'll talk a little bit about this because this comes from a kind of an art project actually. So yeah, as I mentioned, it's like my work with molecules came out of my initially out of my art practice and then eventually became like a really, you know, kind of full-fledged research and then now it's a also a startup. But basically this is in the kind of earlier days. This is one of the first new molecules that I made was hyperf flooror and uh it shows it at the Okyama Art Summit in in Okyama,
Japan, which was a kind of international art exhibition. Um and then we uh basically diffused it throughout the city of Okyama in with these different diffusers that you can see. Um so that's that's what that is in that in that photo.
first. Thank you very much. Yeah.
Oh my god.
Yeah, that's a that's a good question. Someone asked about dilution. So, um, with different molecules, they can often have a very different strength and they'll often smell very different in different dilutions. Um, and a lot of times you want to get if you want to evaluate them, then you want to get them to like a more average range of strength. Um, and so these range from 20% to 10% for the most part. Uh it would in one case hyperl that was 5%.
You know about good sense like a website. Yeah. Yeah. Yeah. That's uh he mentioned the the good is it Tim? Right. Yeah. Tim mentioned the the good sense company which is a a good resource if you want to kind of look up molecules and and what they smell like. They have some information. There's some other um databases out there but um yeah the good sense is is pretty good for that.
website looks like and they're not they're not updating it. I the guy that made it I think he made it in the like 90s. He was just like really into it. Yeah. Yeah. So he and he was a web designer I think first and then he was like why don't I make a website about smell and he just like started putting it all together and I think a lot of you know uh perfumers still use it especially when they're starting out. Yeah. Myself very much included. It was like Oh right. Yeah.
Yeah. Yeah, cool. Um, all right. So, I think we're pretty good on descriptions for molecule 3. So, I'm going to switch over to pleasantness and familiarity and then we'll have molecule 4 and then we'll be, you know, done with the molecules for a little bit. Thanks everyone. Yeah.
Yeah, that's a very good question. Okay. Well, let me we'll just do the last molecule of this group and then because we're going to talk a little bit about the the structures and stuff. Um, cool. And then really quick, um, so I think you all have rated it mostly. Great. And then let's do molecule four and then we'll go into the talk. I'm trying different formats, you know, like sometimes I'll do the the molecules all at the end, sometimes at the beginning, but I think it's like it's kind of abstract if you're not smell, you know, if I'm just like talking about them and then you're smelling them later. So So yeah. So we're trying it out.
molecules you except for the first molecules you invented all the molecules. Yeah, that's correct for this group. Um, let me look and make sure. Yeah. Yeah. Right. Yeah. Is the first one. So that's not me. That's not me. Yeah. Yeah. And I I'll get into the details of those in a little bit. Yeah. Um, cool. All right. I'll kind of like do some of the talk though. Um, so this is a kind of early artwork that I did
that first got me into flavor and fragrance chemistry. And basically the idea was to come up with the chemical difference between Coke and Pepsi. Um, as a Guys, let's quiet down a little bit. Yeah. We're going to we're going to try to multitask. So I'll I'll try to like talk while you're evaluating a little bit. So this is a artwork that I did kind of early one that got me into flavor and fragrance chemistry. And it is um basically like the the concept is it's the difference between Coke and Pepsi, but it's the actual like chemical difference. So like if you added up all the molecules in Coke, all the molecules in Pepsi, and you kind of figure out like what's the vend diagram there?
What's the difference that you know Coke has that Pepsi doesn't that Pepsi has that Coke doesn't? uh that calculated chemical difference as a substance. So this was like a just weird art project. But in the process of doing this, I did a lot of GCMS analysis of Coke Pepsi, of the flavor, and um you know started seeing this relationship between structures and odors and and doing a lot of research. And so that's what like first me first led me into this uh area and then um yeah, this is an art exhibition where it's basically different molecules. Each of those containers is a single molecule and they're arranged by kind of structural features into a grid. So you have methylmethanoate which is the the
lightest molecule. It's like a carbon it's like a methanol escertified with formic acid and um so it's the lightest and then in the bottom uh right you have budanomate which is a a fourcarbon you know molecule with also a four carbon and so it goes in this kind of grid. So, this idea of like structural relationships and like being able to smell those differences and sometimes like putting them all together, kind of averaging it out in order to get to some sort of um almost like general odor arising from the specifics of the different molecules.
That was a kind of um a way that I was working in in these these projects. Um but then yeah the first molecule that I new molecule that I made was uh around the time of that other um exhibition which was 20 2014 and um it was kind of a modified version of hydroxy citronanel. So if you're a perfumer you might know it. It's kind of like a a floral they call it mugay. It's like a kind of white flower lily of the valley.
And um the interesting thing about it is it's one of the very first completely synthetic molecules. So they were able to synthesize venelin and some other molecules that are found in nature before then. This was in the I believe late 1800s early 19 either that or the early 1900s. And um but this was the first fully synthetic molecule. This never existed in nature. And um the only thing is it's it's pretty interesting and they say that it captured uh kind of lily of the valley better than the extract which is you know basically you can't really make an extract from it. Um so that's interesting. But what I wanted to do with this modification which is which is down here um is by adding a
silicon atom make the molecule more flexible so that this kind of aldahhide group and then this alcohol group can get in closer proximity which then um theoretically is going to activate the receptor better. Uh and so you know I tried with some different chemical companies and they were not able to make this molecule. Um but you know I kind of kept it around in the back of my mind and eventually it came to this other molecule uh that you see here on the right. So even though this one doesn't really have that feature of the alcohol group at the other end uh it does have an extra oxygen atom and it does have the silicon atom that gives it kind of a lot of like bulky um bulky interactions.
So just kind of like mass that can interact with the receptor uh in a hydrophobic way and um yeah so that this molecule does actually smell kind of like a supercharged lily of the valley. Um yeah and we don't have that molecule here unfortunately. Sorry. Uh yes. Yeah. So the silicone it's not a heavy metal per se. Um but it's it's a great question. Um we we do a toxicity screening um it's basically incilico actually several incilico toxicity screenings uh for any you know structure that we come up with and what we you know anything that that I have
synthesized is going to be like safer than the average fragrance molecule that's in use today. Uh and silicone is in a lot of cosmetics like in dimethicone. Um it's by itself it's not toxic. It just has like a a low toxicity in comparison to other you know molecules that are used in fragrance. But yeah it's a very good question. Yeah. Um cool. And I think um yeah again multitasking here but I think we are ready for the really quick. We'll just do the that last molecule's um pleasantness and familiarity and we'll get that out of the way and then we'll move on.
Um you know they It was a long time ago, so I don't remember exactly what they said, but yeah, I talked with a few companies that were specialized in uh silicon uh organo silicon chemistry and the one that you know got the farthest like said that they couldn't do it again. I think it was just either unstable or it just didn't produce the reaction that they thought really.
But does that change? Yeah. So if you if you try a version of hydroxy citronan oil that doesn't have the alcohol like then it will be pretty different. Um it's more just like green grass at that point and a little bit more like like if you just took that alcohol off it's like a little bit um citrusy and grassy and it's not very um white flower like coral. Um so so it it probably does play a big role. But here for whatever reason this molecule it's just kind of almost like a rearrangement of that other molecule. It does have the um the kind of value.
There's a lot of accidents. So like you you think that you're making this molecule that's going to smell this way and you're always surprised. But the longer you do it, the more often you do it, you eventually find um certain your intuition gets better. you're able to do a better job. Cool. Um, so the last molecule uh is is actually so when you were asking the question about if there's other molecules that that I didn't make.
So the last one that you just smelled, I didn't make that one. That's al that's called uh or also known as methyl dihydrojasminate. That's also another very popular uh fragrance molecule. So I wanted that as another benchmark and then you know we can kind of see like where uh where the other one will step up. But uh some of the words uh someone said transparent powdery heavy someone got heavy. Who got heavy? Who who got it? Who got it? Okay. Hey. Okay. Yeah.
Yeah. You're a perfumer so Yeah. Yeah. Um but whoever said transparent probably knows what you know probably is familiar with because it's kind of famously transparent. Um, someone said coffee, which is interesting. Um, Pepsi. I I think that was probably cuz I was talking about Coke and Pepsi. Um, cuz that that is the power of suggestion or you know, um, if you see something at the same time that you're thinking of it, there's there's a lot of like crossmodal kind of talk in the brain. So, you probably did smell Pepsi, but it was maybe from another for another reason. Um and then yeah that someone says velvet soap friendly vanilla crisp. Uh someone also said
nothing floral super detergent shampoo wax lemon pine sal um yeah so that's what we got there. Cool. Thank you. All right. So I'm going to just continue on. Oh yeah and this is um so yeah just continuing from this project. Uh this is a um installation that is also in Okana. concludes that scent. And then also there's some mutated morning glories that I grew in this this pool area. Um but I'm going to skip through a couple of these scents actually.
Yeah. So here talking more about structure. Um so this is a a project where I put together kind of analoges. So molecules that are have kind of similar structures but not quite the same. And I basically made like a a fragrance out of those analoges. So, you know, each molecule in the fragrance is uh similar to the next one. And the idea is to try to get to that like averaging effect and try to get to like a general sense of a particular type of molecule but without all the specifics of like this molecule, that molecule, that molecule. It's kind of like a like a blending. Um, and so yeah, this was in this um exhibition. There's these diffusers on the wall and these are some
of the structures and I kind of laid out the structures and showed you know kind of where they overlap. So these are the components in this particular mixture. Um here they are superimposed and then this is like the common structure uh down here and then a similar thing with this with this group. So each of those scents, each of those diffusers uh was showing like one of these like so this one uh this one and uh yeah they're kind of you know I'm kind of like playing around with them in different ways to um align them and then make a fragrance that is yeah kind of abstract fragrance arising from those those different chemicals. Um and then actually here is something kind
of similar to hyperfloor. um you know this structure has the penal structure but it's different in that um like so for example this one is the alcohol where hydrochlor is a key so it has a stronger and kind of like more full smell uh and then another thing is that uh I I put predicted olfactory activity so this was in 2018 so I think the predicted alactory activity was you know kind of iffy but there was a a uh research group in Shanghai that that had this thing called odoractor here. So I was using that to predict the alactory uh receptor activity for for all of these. You can see a little bit of that there.
Um and yeah, so this is kind of uh like one series of molecules that I've done. And you smell hyperf flooror, which is this one. Um you'll smell another one of these uh towards the end. But um there's this kind of it's basically a motif this kind of structural motif where you can take an aldahhide and then turn it into a glycerol acetal and then oxidize it and then you get this and it generally makes the smell more like multiaceted and stronger. Um but it and sometimes it changes it specifically or significantly from the parent molecule. Other times it just kind of enhances it and and makes it stronger.
And uh yeah, so lately I've been doing a lot more with like receptor modeling um using alphaold and some other tools. And you know the thing with alactory receptors is they're not the structure is not very well characterized. There is a there have been some uh receptors that were characterized with cryoMem but um but generally we have you know 400 different types of receptors and you know we don't really know that much about the actual um kind of folding and function uh but with yeah with this you know some some different tools I've been working with a kind of simulated insilicone version of the receptor and then uh kind of docking different
molecules onto it And um and that's that's pretty fun because it's almost like this almost like a collaboration with the the receptor where you're kind of uh trying to get a better score, get a better binding affinity by adding an extra carbon or like moving this or that around and then you try it again. Um so that's been another kind of direction that I've been doing uh in terms of molecular design. So so there's the kind of more structural based approach and then there's the receptor-based approach.
Um, and this is a molecule called E12. Um, just to show more of the, you know, kind of electric field and and different aspects of it in 3D. Cool. And here's a bunch of other molecules. Um, these have a common motif of like a dimethyl group or a turbutal group. So, these little these little branched structures. Um, they smell pretty different, but there's some similarities. And I would say, you know, thinking about molecule design here, it's the the kind of dimethyl group is more of like an accent. So, um, instead of instead of the dimethyl group being like the word, it's more of like an accent that you put on the word. And and so compared to the molecule that doesn't
have the dimethyl group, it it generally makes it more full uh and maybe more complex. Uh and then yeah so I'll just talk a little bit about patina uh which is a startup that I'm working on uh that is basically our goal is to decode scent and in particular make a lot of interesting uh cool scent molecules and we're using these kind of incilico techniques and some machine learning as well. uh basically kind of trying to trying to come up with like what are the primary colors of scent uh and then trying to you know longer term trying to understand the entire factory system from that basis but starting with a kind
of more receptor- based approach initially um and that's the team like co-founder Laura Sisson and uh these graphs from our deck but the other thing that's pretty interesting is ectopical factory receptors so once you kind of are modeling uh the receptors and you're kind of finding very good alactory molecules that bind to them. You're essentially doing you know drug development and uh what's interesting is that there there are drug targets in the body that um are factory receptors.
There's what we call ectopically expressed um receptors which basically you know you have these all these receptors in your nose. you have about 400 different types but those are al also expressed in different tissues and um they have specialized functions but for the most part those are unknown. Uh it's a relatively recent discovery and uh so yeah these are generally like olfactory receptors are not well understood or explored by you know the pharmaceutical industry. So that's another reason that you know they're pretty interesting. And so basically if you're developing you know some molecules there's a good chance that it might also have another function if you find the right place to deliver it to in
the body. And that's kind of what we're thinking of long term with patina. But in the short term we're focused on you know just better understanding these receptors. Um, one one example though is a certain there's a certain receptor found in skin that when it's activated uh it is involved in wound healing. So it speeds up uh wound healing and kind of skin regeneration and potentially rejuvenation. So that's something that we're quite interested in like further exploring. And we have uh so based on our insilicose studies, we have actually a a molecule that we've developed that has a binding affinity of -10 calories per mole, which is quite good. Um and
compared to like a a kind of uh known odorant that's an um an agonist for this uh for this receptor, it's about 190 times stronger. Um now again these are just insilicone numbers so like we do have to validate that but uh that's kind of where we're at right now. Cool. And then so yeah here's just a little bit more info about ectopical factor receptors and then this is um kind of like a a chart of like where they're found throughout the body. So you can see quite a lot of different OSAC receptors and you know here they are in the atapost tissue in the lung you know this one this one O2W3 is found
in the lungs um these were all coming from uh just sequencing you know just sequencing DNA in these different areas um so you can see it's like they're quite widespread even though we know very little bit very little about them and I guess if you think about it it kind of makes sense because okay so the The nose is like a chemical sensing organ. So it has all these tools, these little switches that you know respond to different molecules and you know your other cells also need to understand their chemical environment. So if they're coming from the same genome, you if they're working from the same kind of toolkit then they'll express the you know same oldactory receptors but just use it for a different purpose. So it
kind of makes intuitive sense but um yeah again like right now not much is known except for in a few isolated cases. Um but yeah so I think you know basically you know discovering the primary colors of scent is is very fundamental actually for human health for the future of human health and that's something that we're you know very excited about um at Tina. And so with that so that's kind of the the last slide. Do you all want to try four more molecules or do you want to just like not try four more four more molecules?
Um I see some yeses and I see some Yeah. Okay. What I'll What I'll say is like um if you're feeling up to it, then let's do it. And then if you're not, then yeah, you can just like go get a get a drink or whatever. Um that's fine, too. I don't want to hold you against your will. But um yeah, otherwise Yeah. And I guess real quick, any any questions? Uh yeah. Can you tell us what was in three and four? Yeah, three and four. So three. Yes. Let's go to one of the image slides. Um so three is this guy. Uh yeah, that one right here.
And so this has a dimethyl group. I showed it in that slide with the other ones with the a lot of dimethyl groups. Um now four is heavy. Uh so that's not a molecule that I invented. It's a it's a very popular molecule similar to isoe super in in the level of use and it's just on the more floral side where isoe super is more woody. Um yeah. Um are you see any kind of head space as a starting point like where you're capturing existing sets?
So we we don't do too much of that. I mean I I've done a fair amount of like GCMS headsp space in the past but you know what we're working from is usually pretty much just like synthetic. It's just like, okay, we'll have an idea for a molecule or we'll be going over like the kind of chemical space and like and and realize, okay, this part of the chemical space is kind of like under represented and and let's let's take a look at that. Um, so we're mainly starting from that kind of starting point.
I'm sorry. Uh, validation of like the alactory receptor So um yeah, so if it's if it's a so we have kind of like two workflows. One is like the general just kind of structure based workflow. That one you know it's just you just synthesize it and you smell it. So it's like quite quick. Um for the other workflow where you know we're we're looking at uh receptors and we're starting in silicone the validation you know we have yet to validate that in the lab. So we're we're in the very beginning. But once we once we validate that in the lab um and see that there's a correlation with what our kind of like you know predictions are
then we can actually uh do the rest of the development much more quickly. Uh and that's because you know we'll know that okay if we're if we're getting to this kind of um direction uh from a sensory point of view then that means that it does correspond to a certain receptor activation pattern.
Yeah. Yeah, that's right. Yeah, this is a great question. So about the pleasantness and then the familiarity. Um, so actually the reason that I'm interested in those two things, pleasantness is is obvious like you want generally you want pleasant molecules. They're especially if you're designing them like for for a business. Um, but the familiarity is where I think it gets interesting. Um and and I think you you pointed out that there they tend to be correlated. Um we tend to like uh sense we tend to think that scents are more pleasant if they're more familiar, but there are times when that kind of diverges and that's what I'm you know most interested in. I think that's where okay while that means that we found like a new smell that also has maybe some
commercial potential that has a kind of interesting um yeah sensory kind of profile. Yeah. eating that drug by factory drug. I think it's like or something. I think I've heard of this and it may be um yeah it's it's it's going towards the bomeroasal receptor I believe and it's kind of um based on a hormone I think maybe yeah something to like trigger your Yeah. Yeah. They they say that there's like a direct pathway to the lyic system through this one alactory receptor.
Yeah, I think that's another like interesting area and and that's pretty much Yeah, along with the ectopic alactory receptors, there's the connection to the lyic system. We know that scent is connected but it's it's a question of like is it can it be connected at like a single receptor level or is it more just like this kind of um is it more complex like can you reduce it down to like just activating one receptor causes one effect you know in the in the gilla or just in the brain so yeah I think that's very interesting area I don't know if you're familiar there is like some theory that it's that that the structure of the
molecule is not enough. Yeah. And that you need to know also how the molecule vibrates and there are like like one of the one of the latest of this theory. Yeah. The determinant. Yeah. I'm curious if you're taking into consideration this theory. Yeah. Or like like which is your opinion? Yeah. Yeah. So, you know, I think so Luca Turin um is you know that he has like a lot of very good ideas. I I think the jury's out on on that on the vibrational theory. I think the majority of people or majority of scientists
would say no, it's more of just like you know traditional pharmarmacology almost. um you know we have the their receptors after all and so you know they they respond to lians in a similar way. Um I would tend to believe that personally but I think at the same time there is something still a little mysterious about it. So maybe Lucurin is on to something you know what in in exploring that vibrational theory and this kind of electron tunneling. Um I just from what I've read I haven't heard like very strong validation of that. But if if that's true, then you could I don't know like excite you could you could activate receptors with certain photons that are like at the right wavelength which is
pretty interesting. Um but so far in terms of the kind of development of our molecules it's all just in silicone using traditional like docking methods traditional like pharmaceutical. like how do you measure your simulation as a molecule? Yeah. So it's it's a little so I mentioned about the binding affinity. So we have a a method for predicting binding affinity but it's a little bit more complicated than that because um with you know with kind of alactory molecules there's usually some sort of key amino acid interactions that involve
hydrogen bonding. So you might have a a molecule that binds really really well, really strongly uh and it's all hydrophobic interactions and then you'll find that it's actually more of like an antagonist uh and it's blocking the receptor. So you do have to keep that in mind. So it's so we do have a prediction for binding affinity and that's like like if we're iterating in silicone then we're we're aiming to just increase that and increase that but then we also look at the we also have a prediction for the um amino acids that are being interacted with and we're looking for some certain key hydrogen bonds in certain areas.
I was going to ask a similar question about you know some uh there's been research that suggest whether or not Yeah. Yeah. Right. I was going to ask you do you believe in like I think I chat with professor when I was still in school. Yeah. And he said the reason you can taste heavy water sweet because the autoionization cost is lower. Yeah. there. What do you I Yeah.
No, I think that there is some definitely I think there's something to that. And I think also like Yeah. Like I don't want to reduce like what Luca Turn is talking about to just one thing, you know, because I think there's a lot of other like there's a lot of unexplained uh phenomena in in alactory perception and he's pointed them out and come up with good explanations for them. Um, like one one thing is like that you I think I read this in one of his his books, but it's like if you um are under a pressurized atmosphere like like similar to like if you're scuba diving, then you can actually smell methane and it smells canorious. Um, but it has to do with like the pressure and and so under certain pressures then
molecules will you will be able to smell molecules that you wouldn't otherwise be able to at atmospheric pressure. And that's maybe a little bit like the mechanism is maybe somewhat similar to this kind of duterrated um molecules as well. Yeah. This is a question for people in this room that are like not brilliant chemistry experts like some of you are. Some of us are here as just like hobbyists or like fans. So I pausing on like the molecular thing for a second. I'm just curious on like your company and your process. So like I'm I'm following like the
molecular but I'm curious about your process and what's the end as we're talking about receptors like what is it that you're trying to get your manipulation or like who is giving you direct design? Yeah. Yeah. That's a great question. Um so I mean we're a startup so it's like it's very early. Um, but I think what we're what we're imagining is is a few different things. One is I mean we're going to design our own molecules. We're already doing that and we're going to release those. But the other thing that we think is very interesting, we haven't done quite yet, but we're starting to do pretty soon, we think, is work with
perfumers to develop a custom captive. So, you know, give them their own captive molecule. So rather than it being just the purview of like very very large companies that have these unique molecules that only they can have um you know working with a perfumer to identify like what's a note that they find very interesting um and would be would be useful to them that they would use a lot and then make that like an exclusive capture to them and that requires that the molecule be under patent or at least you know pending patent you know turning into a patent. Um but yeah that's that's an idea that we really want to explore.
Um and and we are talking to some perfumers about that. But then the main thing is right now is just molecules. Um in fact, we're going to start selling some on Perfumers Apprentice. Uh I don't know if you're familiar with that, but um yeah, we're going to start selling a small kit of some of our molecules. Some of the ones that you smelled today on Perfumers Apprentice fairly soon. And um longer term, you know, this ectopic olfactory receptor is is something that we see as like an an outgrowth of what we're doing with the the other, you know, molecule stuff and and that has other applications outside of fragrance potentially. Um so we're looking at like a cosmetic application, but that is going to be further down the road. But yeah, a skin cream that contains a a
molecule that activates the receptor in skin that causes it to regrow faster and kind of rejuvenate. That that requires some validation before we get to that stage. But, you know, that's one possible application that we're looking at. [Music] Then you also talk about you know changing molecule making. Yeah. So I'm curious how do you kind of translate more like very more like abstract? Yeah. I think it's always a struggle to
um describe how a molecule smells and um you know I'm kind of interested in like trying to develop other language that's not so referential um you know so like like in the English language we can only say that something smells like something we can't smell we can't say that something is something with smell so like for example you can say the sky is blue and blue is like a specific color that has a specific receptor in the eye that it corresponds to but you generally don't say that like a scent is something you say that it's like something. So it's it's like a rose or it's like a or it's nine, you know, which is another way of saying that it's like so um so I'm going off on a little bit of a tangent, but I think in terms of the
yeah like like how to describe the molecule and the nomenclature, yeah, I do I kind of use whatever words kind of kind of work, but then you know that I know that they're they're never quite right. And that's I think what keeps scent very interesting is that it's like it it's elusive to language. You know, you can't quite pin it down. And as much as I get all these like words and and kind of numbers from all of you, like it's not going to it's just a a kind of lowresolution reflection of like what the molecule experience actually is. Um but in terms of like the structures like the visual structures and and the kind of which relate to the 3D structures of the molecules um you know there's kind of like little like I was trying to talk
about it being a little bit like a grammar and uh yeah like here like so here it's like these are different molecules and they smell each one smells pretty different but they do have these methyl these these kind of either um turbutal which is like a three methyl groups uh or have two methyl groups and the methyl groups are just these little carbons that are coming off of the main chain.
And um yeah, I I see that as like by itself that wouldn't even have a smell. So if you just smelled like a turbutal um kind of like like carbon arrangement with no oxygen like it probably wouldn't have much of a smell if any. But when you when you attach it to another molecule, then it becomes almost like an accent and it tends to make that molecule tends to like push it either push it in a certain direction or just like make it a little bit stronger. Um, and so for me that's just been like one of the tricks that I kind of use or or something that I tend to fall back on when I'm designing molecules. I have like some of my like kind of favorite things to to do like when I'm kind of designing a structure. Yeah. Um, sorry.
Sorry.
Yeah. Yeah. I think that's that's very true. Um I I think I have like a a very mild synthesia with like smell and other things like music. It's not like there's I think there's like involuntary senesthesia where like you can't really help it. You're just going to like if you uh see the color yellow, you smell toast or something. That that's one type of very, you know, um specific sesthesia. But I think on a lower level or or on a more common level, yeah, there's a lot of people that have that in some form. And I I tend to like picture things or have a kind of like feeling like when I'm like smelling a molecule, I have a kind of like feeling of it. I think a lot of people in their language to describe the molecules used
certain terms that were not exactly referential but like you know like velvety or transparent like I think those are almost like sesthesia terms you know they're not talking about you're not saying it's it smells like a window you know that's transparent you're saying that it it provides a transparency um and yeah I think that a lot of there there have been some studies that like crossodal language can be a good way to describe smell um so borrowing from other uh senses basically. But yeah, it's a very good question. Yeah.
Oh yeah, that's a very good question. It's just for me it's just for this one it's just limited to humans. There have been um some studies where they created what they call consensus receptors from combining the DNA sequences of multiple species and they were able to better express those in um in some other organism. But yeah, for us we're just interested in the human um receptor. Does uh does the weather like affect like it's a high pressure day or high or something like that? Yeah, definitely high humidity affects it. Um, heat generally is going to make molecules come out and kind of, you know, um, come
into the air. Pressure probably affects it. I'm not Yeah, I assume it probably if it's high pressure, maybe they're coming out less, but on the other hand, it tends to be humid when it's high pressure. So, probably comes out on balance, maybe more. I think it's high pressure. And when it's hot, it's low pressure, I think. Oh, okay. So, I don't know if that made a difference. Yeah. Yeah.
There's definitely like um effects from the atmosphere like for sure. Yeah. Cuz you're talking about scoop. Oh, yes. Yeah. True. True. Yeah. Theoretically, if you were Yeah. under a super high pressure environment, then yeah, you would smell things differently, at least according to this study about the about methane. That makes sense. Yeah. Yeah. Cool. Um any other questions? Um okay. Yeah. Have you AIS? Yeah. So um so we're definitely you know using AI. I guess you can define AI in a few different ways. And you know in one sense we're using just traditional like
machine learning. I guess in the other one in the other sense we're using like alcohold and the kind of like 3D you know um physics based kind of simulations. And I guess both of those are a kind of AI. So we basically use both of those. Uh for the structure based approach, you know, there's a bit of like intuition and then we're kind of using um you know, some machine learning to boost that. Uh so so it's kind of like the stuff that I was doing already with designing new molecules, but then trying to cover more chemical space and and amplify it using machine learning.
Can you help that smell like something? Yeah, there's some like there's some open source um things for for that. And it's kind of interesting, but like I I always find that like the the language it's like I I'm always just mostly interested in pleasantness and familiarity because I think the language can be elusive and there's lots of different ways to to kind of label it. But I think theoretically it's like interesting to know at least in what direction, you know, if it's like a green note or a kind of sulfurous note and things like that. But um yeah, definitely I think it can help it can
help with the prediction. So far, I think we're in a point where it's more like um it's more helping to search the chemical space for interesting things and then we'll screen those things and then refind the model based on that. Yeah. Sweet. All right. I think I think that's good. Okay. Yeah, maybe that's the last question. Okay. We'll talk after if anyone has I'll be around if you have more questions. A round of applause.
Thank you. Thank you. So, um I guess like for people that really I would say for people that really want to smell the other molecules and that's that's great. Then just just come over to this section and then for anyone else, you know, feel free to mingle. Um but uh yeah, we have some more molecules if you want to smell it. Yeah. And if you want mask um come over here and get the sleep mask as well. Thank you guys. All right. Thanks everyone.
Hey, actually I want to mention one more thing um before people break. So Apatina, we're looking for some excellent chemists to join us. Um so if you know a synthetic chemist that actually does, you know, synthesis of molecules, um then yeah, we'd love to to meet them. Thanks. I think it's possible that I haven't I haven't produced anything that does that. Um but yeah, theoretical I think we could um you know we go toxicity screen [Music]
if it's within a certain
I feel like I know some Yeah. Yeah. I mean I think like I mean I'm always interested I know that you know it might be cool. Yeah. Yeah.
So, what does this map
Universalist working on I don't know but I read not because 198 was on like biology. technology. So I talk about like how you can It's only biology. Sorry.
[Music] Yeah. [Music] Yeah. Yeah.
[Music] Okay. So, Oh, thank you.
[Music] You can talk to him a lot about [Music]
So you couldn't [Music] How much are you doing? It's more like
Stay inside
here and show another 15. [Music]
[Music] Okay. [Music] Yeah. [Music] [Music] Nice little
But no, I just [Music]
[Music] Yeah. [Music]
for me is [Music]
[Music]
[Music] My name [Music] Did you get the [Music] No. [Music] I'm always asking. [Music] [Music] Oh, perfect.
I can't believe [Music]
[Music]
I actually [Music] [Music] I down. [Music]
Yeah. [Music]
[Music] Is that all day? [Applause] [Music]
We got the we got the video. [Music]
I still have a question.
All right. [Music] [Music] We're starting to [Music]
Yeah, we wanted [Music] for you. [Music]
[Music] Yeah.
[Music] No, I don't know how [Music] [Music] But like afraid [Music] that [Music] you could I was talking to like
[Music] [Music] [Music] Do you always know? [Music] like he would like it in like
[Music] You know what? [Music]
[Music] [Music] Yeah. [Music] [Music] Okay, I'll look.
[Music] like [Music] we all Is it related to the look of a question?
totally fine
like [Music] First phase is kind of basically producing like the best fragrance molecules and then selling those to there's a few companies that control most of the market and so basically like [Music] I tell them to larger so that's kind of like the town's market and then I Yeah. Yeah. So we have like right now we have like five pending patterns.
Yes.
[Music] I wonder about that too. Like I wonder like if you had
like spacele [Music] there are atoms in the universe the space is huge. That being said, I think the factory is small. So you have like
here and there. So finding the pockets. I think there is a bit of race to that.
Did you get download? Did you get the app? I'll do it in a second. [Music]
I wonder. Let me let me download it right now. Okay. Yeah. Yeah. Thanks, Mikey. Yeah. [Music]
totally totally I think also it's like interesting gets a little bit with perfumemers because they they actually want something that can't be copied. It's like they need like so it's like so like my general tendency is like cool but then it's like easy for like a large company to copy like
[Music] That'd be great. I'd love to chat. Yeah. All right. Nice to meet you, Robert. Oh, yeah. I might. Yeah. Yeah. Um, I guess different backgrounds. I think I started out really interested in molecules and then I learned them kind of like how to design them for fragrance. And I think they started out being like really good at machine learning and then they kind of me interested. So like we kind of ended up in the same place like where they put a lot of machine learning into it to like maybe a guy that's like you know been
studying it for a long time. Um, but I think uh for me I'm kind of trying to kind of boost what it is that I do with I already know how to do more like [Music] a certain range and a lot of times I'm surprised but so we have model that predicts the smells was like, you know, but then also if I look at something like I can have a general sense of how it smells, but then I'm also surprised
like you know 30%. Yeah, there's always there's so many surprises. Thank you for your appreciate it. Nice meeting you about the operating. Yeah. Yeah. [Music] Usually I'll be doing one of two. What is going [Music]
to wear hat? Oh, okay. [Music] That's awesome. That's so exciting. That's great. Like I think uh you know pretty have someone that's like you know I can start coming. Yeah. One of the things that we're trying to fix now is having like actual real people put in some money just like pretty
and then And so when so
[Music]
yeah I didn't mention that it's super competitors but I was like kind of thinking should I like but uh yeah [Music] right now we had we had someone their service like we were working back and for now and they're like it and so now we have to like so I know it's fine but it's just We're doing like a lot of just like little tough things. [Music]
I'm here. I've been here a full month. Are you a job or Yeah, I work at Oh. Oh, that's This is the connection that you do. the lab that I worked at I was research I was going to ask I mean really this is like a after work I'm busy this weekend but like you know maybe next [Music] more
rushing. Yeah. I always [Music]
[Music] me. I got to get I was thinking about hosting events. I went to the AI government. Oh, cool. What's up with the number? Okay.
[Music]
Yeah. Yeah.
[Music] You guys are important. He comes to fun. Thank you. It's just me, but it's almost you want to help volunteer or I mean kind of sure [Music] that was
Are you all from the cosmetics industry, packaging, the chemical engineering should have done this? I went the other I went process engineering and I used to make bounty. So, and then I got into water. I did 10 years. used to do um assembly line. So I'm like a weird Pokemon that nobody's ever heard of. Yeah. And I'm I'm getting a refresher tomorrow. I want to like crack open my old books. I've forgotten a lot. I see you. Oh, nice.
I might need to take an edible caroxal. That's that's what they know. I think that's where it revamped chemistry right trying to get out fragrances in that world also is right a lot of time lemon [Music] or like one of those kind of places if you wanted to connect some because that was my universe for a little That's pretty. I mean, there's a big commercial, right? But but they got pretty
trying to do, you know, a lot of really amazing like more sustainable. [Music] Right. There is. Yeah. Yeah. We screen everything we try to make [Music] my hats off. We were just sitting around trying to figure out what is computer. Yeah. Well, imagine. Do you have an understanding of Instagram model [Music] like you're able to like predict super cool and this lady is amazing to know as
well. I come from another world. I work in the space. The blockchain space I could Where do you live in the city? I stuff in the chain. I got a nice I just moved to Williamsburg. I hate it was terrible. It's too far away. You going to come here? I was like they like automated. It lives in just talking with them about the house, but um tribute. Yeah. Well, you Yeah. You have your I I have a friend
that works there, so Yeah. Yeah. It's not automated automated. Yeah. Yeah. Your nemesis. Okay. True. Yeah, [Music]
they just invest. Yeah, like on their uh website they have like a range, you know, like I guess maybe by chance you can call upon some good stuff but then like also have
so that was a great question. You know, let's let's be tangible here. what's the, you know, what's the product kind of thing. So that was very helpful. No, because my friend of the American sitting and going, she said, "This is very bad. Why are you coming? I want to be here." So excited. Yeah. What is the thing is And so it's like kind of just holding out process. So it's like we're
kind of like have different possible but the first one is [Music] really Oh yeah. What's the last? Yeah. Okay.
Right. Right. Yeah. Yeah. Yeah, I mean I think our main thing is we want to make the molecules to make two um but then for some of them you know some of the molecules could have a skinny effect. So that would be like the second like so first we make the molecules that have very good but then we can take some of the learning about like use that to design that activate.
But I understand if I do this kind of like yes I understand what you're saying maybe right now we're raising money the recent one develop that and then validate oh yeah so the validation of the of that particular receptor first we need to there's one of there's a couple different ways and we're kind of still researching like what's the best way but we could definitely have like a CRO So it's like a test.
Yeah. Correct. So in this case we're focusing on a cosmetic where we [Music] support [Music] you can have that become and you can have that kind of so that's much less regul and then you can publish the scientific research that um you know we're trying to like find a different channel you know that's not as difficult as like
Yeah. Yeah. Exactly. Yeah. We're trying to start small trying to get the product on the market right away. Um that way we can also start to Yeah. Not that we could arises after we're like [Music]
cool. Yeah, including like Yeah, that's great. Canadian, but my my effort is uh Yes, I have WhatsApp. [Music] So, so I I posted an [Music] I have like a [Music] partner. Oh, yes. Okay, I'll do that. That's kind of like the only group that I knew about a year. I mean, I've been moleu Sorry. No, it's fine. Um, no. Um, the reason the reason
[Music] Yeah. Yeah. Yeah. I have one. How about this? I don't know if you have any too much information in my radar. So I have to but uh but there's also Thank you so much for maybe that [Music] and there's not enough seating but I kind of like it. All right, it's good. Yeah, but then we do need better. Yeah, I just need to figure it out. Um,
I I can help you figure that out. I did the same during your tech week. I I still very nice for me because of texting. But if I go again, you know, see or something, let's see, maybe I can cover it. So maybe we can figure out ways. Uh there's also [Music] really my friend.
No, we can figure it out. I can I can I can I can I can try to put it together. if um if you agree. So, so let's just figure out the I mean so do we exchange the number I know yeah there's like there's one there's like the official right [Music] [Music] nothing luxury something simple around town like You don't have to leave my butt when I have
to say and then I would like to be here. whatever, but I [Music] congratulations to see you. Good night. All right.
[Music]
So my question
um so basically not yet [Music] and it's put on by a lot of performers. [Music] There's a lot of people interested but I one is that you know that just recently but then the other thing is like two of the molecules um it's hard for us to have like a little bit but anyway yeah starting to like send out samples to them and yeah I've talked Yeah. Yeah. Yeah. So, so basically that.
Yeah. So, they would tell us like that they think is missing, but I don't know if you can be. So, at this point, I don't know if they're going to be so specific as like to to answer that and then they can pick the best and So that way and then out of that then they would get like a hard you know
me So I think this is [Music]
I thought [Music] I My [Music] second [Music] down, [Music] you know, like so I think the first stage like the easiest way that are part of the ingredients and then we do want to work up to where we're licensing [Music] ingredient. Yeah. Yeah. So like we would just give them these ingredients. are developing our own brand all that but like the brand is not really it's more like you know
yeah so that would be the first stage and then we do want to work up towards you know licing what we think the best way to do them
Yeah. Yeah. Yeah. I've heard Yeah, sure. That's fine. Yeah. Yeah. Actually, there was a guy from Osmo. I know some of the people from there.
He's like tall guy. Yeah. [Music] I guess like my most like sure but but basically like I think so like The way I there's like two different terms and I think similar but yeah I mean I think recently they pivoted to just formulation and then like fulfilling small orders or like or like designing uh fragrances with low. Um so I know that that's like their main So that's a little bit different. At
some point they were looking more on talking with people there. I've been looking at you know but we'll see. We'll see. Maybe they'll do the back. I I guess basically the difference I think is maybe the approach to AI learn you know how to um and then I'm just using that and accelerate that and kind of like expand it. Um I think they started where they they
program didn't really know about fragrance maybe were interested in fragrance but didn't have that same background. So they have to build up a whole system to get them to the point of like maybe whatever and then I think it's like [Music] trying to build system which is want to do that through I think they're focused on to the extent that they're they're kind of thinking about the model Yeah.
Yeah. Yeah. [Music] Right.
big ide
is it just like I think what to me is like I mean I think that there's definitely because we do have those receptors body and that's how a lot of drugs work today but they haven't used them on the perfect so I got a lot of you know it's something where if you fully understood the receptors you could introduce different scents um you could have like for example like a byebye nice to meet you could have like a
almost like virtual virtual reality like thing You would need to know how to like that compress the entire system to like