Andrew Yip - Future of Food with AI Technologies
welcome welcome to another episode of the impact of ai supported by wilda women leaders in data and artificial intelligence every week we explore how ai impacts us personally and professionally and this week is no different today we're here with andrew yip he's a food technologist and also a chef and we're going to explore how ai could potentially impact food and we're not going to specifically focus on the food industry or the supply chain or the point of sale we're not going to talk about how ai is being used for drones to harvest you know the food in the farms but we're going to bring this forward to the behind the scenes the technology of food and how ai and ai technologies could be expanding this space differently and how it's going to impact us andrew welcome thank you we'd like to start with your professional journey as both a chef in the earlier career and how you've transformed that to your current role which is the future of food okay all right so i started as a coach during that time a junior role in a in a restaurant and progressively i woke up to the ladder in the kitchen hierarchy some point of time i know that that's really more about what i could do within about food so i diverted my career into a different path where i got myself a apprenticeship in one of the largest food company then from there i actually learned about that's really more than just about chicken beef soup and etc it's really about how to create a good product using the available ingredient and raw material that we are able to utilize them from the basic of food building up to a complete delicious product so from there actually i see a lot of interaction between human being and another human being try to find out the answer but as we progress on i do realize that there are possibility of using machine to do it but it's just not possible during 20 30 years time but somehow throughout the years especially the last three years i noticed that there are changes to this how machine can help us to work so i am i personally not able to use that facility or the machine yet but i see there is a great potential where we are going to hate on from now so as you've progressed from a chef to then working kind of behind the scenes your your role really moved into food technology which is an entire space in itself what is what is food technology okay food technologies uh using uh basic raw material of salt sugar uh certain amino acids and uh using of certain modified starches spices to create a basic taste environment for the food product then you build on to the food structure in terms of what kind of a food that will be going into for example if it's a liquid soup then you have more ingredient using vegetable protein as well as the starches into that manner okay that is if the product goes into the soup but then using the same principle this can apply to uh processed meat uh it can apply to noodles pastas ready meals etc etc so a food technologist is breaking down the components of food and looking at maybe new ways to experience it another way to put it food technologies uh enable the manufacturer to have a consistent uh method of preparation of uh food at the same time uh safeguard the integrity of the food itself so that it can be still as delicious and as tasty as should be but at the same time because we are talking about mass production so there's always uh course involved how to bring it down uh producing more cheaper and faster as usual the uh the business point of view of producing it are we going into processed food no i i believe that always uh how we see uh processed food as a lesser of evil in in that manner your current role is the current role you have is head of future food what is what does that mean i mean that's that's kind of a really i mean that could be a very broad broad role it is it is okay uh this one takes a bit more uh explanation a lot of time we are looking at this advancement in technology to produce household kitchen appliances we see the most recent one about about seven years ago we have this uh superheated airfryer instead of a deep frying a very uh same concept of heating but at a faster speed and the food can be brown at a faster way instead of using microwave which you can heat up the food but you can't really brown the food itself so air frying is a great technology to come by but at the same time we see the induction uh uh stove and things like that it give the end user the consumer the homemakers uh uh save a lot of time as well as in the industry itself we do not need uh the gas but the electric uh will just able to speed up the things so what i'm trying to say is that that's a that's a great improvement in terms of cooking appliances however what we are not seeing is how can the food be improved so that we can have it better more nutritional more sustainable at the same time good for the planet and good for us so in this aspect is about future food what we're going to consume in future but we're going to start with now how we're going to consume the food now that will impact us in regards to every day can we eat better but more sustainable can we eat better yet still maintain that deliciousness so future food is really about how we see food now how can we change it are we looking at the perfect apple or can we have it still apple but we can have it differently but still deliciously and enjoy the apple that actually brings me to a something that we wanted to talk about today is the utilization of ai technologies and when i say ai technologies i'm talking about um natural machine learning deep learning language processing being able to augment the the human condition faster more accurately and with you know oftentimes higher confidence and you mentioned breaking the food down so that it can be healthier one of the concepts that we've talked about earlier was utilizing ai to create new recipes either improve recipes create recipes or change the recipes so that the foods that we're eating today is healthier i'd like to understand your perspectives of would ai in this scenario actually provide more benefits to us okay so uh from the uh food technology's point of view all the uh food starts with a basic formulation of understanding the raw material use now all these oral material used can be measured and can be quantified in terms of the quality as well as the quantity they can be transformed into numbers so if that can be done i see no reason why it cannot be fit into the algorithm of an ai and transform them into a data to be processed by the machine that is my very simple thoughts of that to begin with yeah because you mentioned um because really the food technology space is i think where we're seeing more innovation and you mentioned earlier just the innovation and compliance and how the consumers you know how the consumer takes the food but also i think how you know from the behind the scenes the research and the development i could absolutely see where an ai technology could take a food that we currently use today whatever it is mashed potatoes something that's consumed quite broadly on a global scale and very quickly run scenarios and simulations to say hey if we just make these little bit of tweaks we could actually serve the same type of mashed potatoes but with these extra ingredients or with this extra twist it could be healthier so now individuals consuming that product hey i love mashed potatoes but it doesn't really do anything for me now can say i love mashed potatoes and now i'm getting majority of my health yes i could provide another example uh in fact this is what i currently doing so uh in the space of meat alternative that i'm doing now we are utilizing soy the soy protein that we are sourcing uh all over the world but there are so many that we could source there are so many data that we could gather there are so many grades of soy protein in regards to different grades different protein content and most importantly where they are coming from within certain radius of of in terms of the the source of where this coming from now in order to to have a very sustainable way to to produce we shouldn't import and invest too much on the carbon footprint but in order for us to to narrow down the search and choosing the right balance between the price the supply as well as the functionality of the soy protein if we take human to do it i'm sure we could do it but it takes a long long time but if we could feed all this information and turn it into a data that can be processed the combination and also the permutation that comes about will be much faster and the analysis that done by the machine helps the human to put things together i see that is a possibility i want to i want to stay a little bit on the topic of the alternative meat products or alternative alternative meat products so these are you know for everyone listening these are your plant-based meat industry uh companies like amy's kitchen beyond meat uh boca foods and i think probably here in the state something a little bit more known or more popular is the impossible foods for example they now have the impossible burger at burger king and here in the states with starbucks have the impossible breakfast sandwich now what i've i've noticed with some of these companies is you know the the perception of the the alternative meats and the hasn't really been you know exciting you know a lot of times the the vegetarians that i speak with or the foods that i try myself they're very dry nobody really likes it and then suddenly technology has burst in the past couple years where the flavor profiles have been raised within these alternative meats where it almost tastes exactly like and it looks like and it it burns and cooks like real meat so with your experience in alternative meats have have you been able to would you have any thoughts about where that technology is is coming from or how we've been able to expand that that that industry sure so let's take a piece of burger to do as our uh discussion point yes now a piece of burger we talk about the structure of the burger we also talk about the uh the colored appearance of the burger then after that we also need to talk about how can the burger smell and especially before cooking during cooking and after cooking the aroma generated by the burger itself so all these three aspects and not to forget the nutritional aspect of the burger itself so it's really about uh three-dimensional but then again within that three-dimensional that there's always a space how can we make the more affordable more sustainable at the same time make the burger more as delicious as it is if it's not coming from the animal okay so let's uh let me jump into the the quickest one in terms of the color the color can be uh natural colors that we can get from derived from vegetable process so that one is not difficult to achieve but to achieve the correct tone of the color we need a combination of two or three different colors to make it a much more vibrant in terms of slightly more closer to red meat or closer to white meat or dark meat so the tone of color can be an interesting combination to begin with in fact it's not as easy as mixing color a and color b it's about the different shades of color and there are some colors can be encapsulated and come in much later when it's heat activated and change the color when the cooking process begins so this uh something that the technology can help on the food integrity itself class one yes i i i'm seeing this now oh go ahead yep yeah so next we move on to the texture now this is a big topic the texture to put in ingredients to look like a meat is not as simple as what we used to have however with the the technology that able to to come up with certain uh binders and so that emulsifier and certain stabilizers to combine the water the protein and the fats which normally all this street doesn't add together but the emulsifying effect through the uh the advancement of technology the water the oil the the fats as well as the protein able to come in together very nicely and firmly and that's where you get the good texture as close as what animal protein offers now that's on the texture in a very simplicity way the third one is about the aroma now which is the most difficult of all because now animal protein do inheritantly they do have certain smell and aroma and the smell changes as we coat the piece of burger or chicken however if it is a plant-based can we do that again the technology will enable the flavor to activate at certain temperature point to release certain aroma and as we continue to cook and it will continue to activate based on the the temperature itself or that what we call a time release activation of encapsulation so the longer we cope to until a certain point of time the piece of burger will release the full flavor when it's being consumed by the consumer so here it goes technology again play an important role in putting up the flavor together because you do not want a piece of burger to smell the same before during cooking and after you would like it to change like a real animal protein how can we do that it doesn't happen easily but it can be done so i think this is a probably a really good use case for ai technology you've mentioned you've mentioned breaking food down almost to its chemical component and specifically with alternative meats where we're looking at plant-based well anything you know that's non let me uh let me rephrase that specifically looking at alternative products so whether it's spirulina or some protein that's attaching itself to a plant under the water all of these things are not known to us we haven't experienced cooking with them before so if i'm looking at this on the other side the research side i'm taking all of these products from from around the world that we've never experienced i'm having to break it down into its components almost down to its chemical compounds and then i need to be able to run simulations after simulations after simulations with just that product but then i need to also understand how it's going to combine with other products and you've mentioned just three and i'm sure there's more three areas that the the flavor itself the texture and how it sits and and contains itself and then also the smell and how it cooks over time yes i i absolutely think this is probably the best use case i could i could think of of where ai could come in and augment the amount of time that it would take to research all of this into a shorter period of time right so when i i'm leading a team currently i'm leading a team of food technologies so one of the reminder i always tell them do not have too many variables keep your constant and have a not more than two variable as you do on the trials now there's a lot of permutation but with the experienced food technologies they will know how many trials they could they will do to achieve certain results but again if we talk about a new product if we talk about a ingredient that we never used before you talk about uh evaluation of certain uh replacement product that's a lot of trial and error and after that we will have to really taste them in order to give a correct organoleptic sensory evaluation so this takes a lot of time okay so whereby if we have certain uh all these are being captured in in numbers when we taste certain product we are giving them scores we are giving them a very meaningful evaluation based on sensory data and that is a huge topic itself sensory data science that can be also be used by the machine interpret and analyze by the machine the algorithm will have a good uh how to say to give the the meaningful interpretation of what the numbers are so in this case it will help us from the food technologies point of view to come up with certain permutation faster without going through all the series of work that i mentioned earlier you mentioned earlier the the sense of of smell and i and i know that taste and smell often play a very have a very close relationship with each other i i was reading online and i i'm gonna paraphrase because i don't uh have the exact quote but online there was a chef that mentioned that the sense of taste is best in the morning when it's been unencumbered by other things and as we go throughout the day that sense of taste waivers and it becomes less and every time we eat something else that sense of taste gets less and less throughout the end of the day another use case that i i think that we've we could probably discuss is how we could use ai to understand the flavors of the food and then be able to impact how we consume it throughout the day yes that is true as what you said you currently pointed out it's best to taste uh our senses are more sensitive more acute uh in the morning so there's a sensory fatigue as with days goes by because it's the same tongue that we have and we can't change another time so we just have to use back the same one and that relies on the the the memory that we have to taste so there's always a sensory fatigue cake skin so best that is taste in the morning as we speak now they are already companies especially dealing with flavors and chemicals already using aai to process the data to come up with a better flavor a better combination of chemical molecules combination of them to come up with a more more price sensitive flavor more stable flavor that can uh withstand certain uh heat resistant and also the flavor can be the smell and the flavor can be more closer to what the orig the natural food product brings for example strawberry strawberry is not difficult uh to to identify as a strawberry right strawberry jammy or strawberry sweet and all strawberry it's just plain raw but how to get all this right in in the shortest time ai can do that i strongly believe that i know they're already happening you know i i i just thought about this who determines what food tastes good and i never i never thought about that until just now but but i was at a restaurant the other day and you know my one daughter loved the the food that i was eating and another daughter tried it and she's like yeah i don't like it and if we're talking about and if we're talking about a.i and how ai could impact the flavors or the the sense of smell we still need to have a human behind the scenes that that says yes that's that's the flavor we were going for i mean it sounds like the the ai technology is there to maybe augment the simulations and run those permutations and variables but it's not replacing the chef it's not going to replace the food technologies because somebody needs to physically be there to say yes this strawberry tastes exactly the way it should even after we've done all of these things to modify the strawberry yeah the hell does that of the machine take away my job no uh something that we do it almost every day so when we taste a food product uh still in the early development stage we make sure that uh all the comments or the uh ranking of how strong is the the strawberry notes okay how we describe the strawberry notes and how this strawberry uh uh is it a sweet sour and all all these valuable comments are being captured so we do spend time uh process them now ai will help in this manner but ai will not tell you until we feed the ai sufficiently to form uh the big data that they have what is called uh sour that is always a threshold that is good for human being certain sourness a strawberry if it's not certain do not have certain acidity do not carry certain sourness it is not a nice strawberry but a strawberry that has upfront sweetness but in the sweetness it carries an undertone of sourness and also a sweetness lingering behind after taste that is a good strawberry now how can we achieve that until we feed the ai with certain threshold that we determine the threshold then the ai will in return process that and help us for a faster processing yeah i lost you no no i'm i'm i'm i'm thinking i mean i've got food now in my head so i've got all these things going on the other the other piece that i was thinking of is the the combination of foods for example when i eat a peppermint leaf yes i i get that peppermint flavor but when i eat a peppermint leaf and i eat it with some very dark chocolate i've noticed that the peppermint is more peppermint the the flavor of the peppermint is raised you know to the a different level so i i'm thinking another use case here is is certainly the combination of foods and how we combine foods together or how a restaurant would put certain foods on the plate to be more to be more adaptable but maybe to combine together and create something better okay yes in fact uh it's already happening okay it's what we call uh food pairing okay what best who who does who decided that chocolate pairs well with peppermint yes right yeah okay so this one is ho uh combination somehow but what we are talking about could we have new pairing for instance so uh this one actually uh championed by hayson blue mentor a very renowned chef in uk he pairs caviar with white chocolate okay it goes very well trust me i tasted that okay but i'm sure i'm sure there are a lot more combination that we do not know of but as we speak uh what i know there's some groups of scientists as well as food technologists and also chefs who are talking about cultural differences behind each food pairing why does it pair why they are bringing in much more than just why this food pair with this they are bringing in the cultural the geographical the the place where they grow certain food so they are putting all this information and surprisingly uh the according i i did read an article in terms of food pairing they are differences between the west and the east the west would like to just look for a good pairing in terms of two kinds of food that means food a and food b share certain similarity in terms of molecule that's a good combination of food pairing but they discovered in over the eastern side a lot of food do not need the similarity of molecule to pair the food so in this combination the food a and the food b have totally no relationship between whatsoever in terms of molecular level however they are using a third food that means the food number c as a bridging flavor so they able to classify in terms of geographical area the north asia and the south asia and over the european side and usa which country or which continent will have this tendency to pair flavors based on the same molecular structure and which country or continents would like to pair food based on the third party that means the food number c as a bridging flavor to bring the the combination of food together i am blown away with this kind of result but in order to achieve that no way and no amount of human analytical mind and also the time to process all these data you've got to be ai behind this yes i'd i didn't even i mean i'm i'm not a food technologist i just really like food and i didn't even think about some of the the nuances and the the the additional variables that you just brought up that just looking at food at a molecular level and how it impacts culturally and you're not even really talking about you know one point yeah the us versus you know asia pacific but you were also talking about the nuances within that asia-pacific geographical area that's that's pretty amazing that that's a lot you're right it's a lot of data then in fact i i just wish okay they are probably they're already being done how can the nutritional value could impact the food combination at a very affordable way because nutritional foods at times not that affordable that's one then how could when we combine food especially from a food preparator point of view when we combine food how can we make sure that whenever the ingredient we use are sustainable and actually uh much reduced carbon footprint can we do that before we actually start cooking can we do a simple planning when we do the shopping on on the grocery for for for the food can we do that uh only i can only choose from certain uh area or certain list that are all sustainable within 300 kilometers radius not more than that could we do that could we narrow down 250 kilometers radius so this this all this information uh uh are useful it's not that a chef doesn't care about global warming or environmental awareness but we need the tools to enable them to do so well i'm glad you're thinking about this um because you've taken it one step further it's not just looking at how ai can impact the food the taste i mean we've talked about taste we've talked about texture we've talked about how the food changes when it's cooked or and you've even referenced that how it's cooked differently whether it's in a fryer air fryer whether it's an induction versus gas versus electric but but now you're taking that one step further and how ai could impact a more nutritional value at a price that's affordable for everyone and i i think that's just an amazing concept yeah so this is all about how can we do it better that's why i'm in this space of how can we search for a better way to do things better way to to cope better way to manufacture food that's what future food all about i think that what i've learned today is that ai is not a trend in in this space in food technology it's something that's really going to help the food technology space evolve probably more quickly in the next couple years but i'm seeing it move away from a trend so trend for me is there are pockets where maybe this is being used and i i could see it evolving into something more of a permanent reality but what i've also learned today is that you are not going away that just because ai has a a variable calculation that says this is what it's supposed to be we still need the the human interaction we still need you as that food technologist to ensure that that's the right combination to ensure that it is sustainable to ensure that it is going to provide that nutritional value and it's affordable yes i strongly believe that technology is a double-edged sword it's just how we're going to use it and i believe that with ai is really about it should be helped us better to to deal with food especially to be honest our food system is a broken pieces it needs to fix i i'm i'm i just hoped uh there are people are working on it i'm working on it in in a very small way but we just need to fix our current food system it's not that we got not enough food it's about how we're going to distribute the food it is totally different topic i understand but it's really about the entire food system that how can we make it better for for everyone on this planet so as we we wrap up where do you what do you think is next i mean you're in you're in this space what's next for for the future of food where is this going at this moment we are looking at how can we do something a very small way of doing it meaning by changing the way people eat that is something we i'm doing now and what's next as to answer your question it's really about making it much more affordable but still maintain the integrity of the food after all we are not human not talking about if i will save the planet i'll pay through my pocket without tasting it good the taste is still the paramount of every food technologist to make it tasty to make it delicious and get to to safeguard the integrity of the food itself so the future is not not to happen totally entirely new food but how to have a similar taste better quality better nutritional and much more affordable that's wonderful do you have any final thoughts my thought yes i do hope that how can we distribute the food much better food wastage is the my main concern is how can we fix that food waste yes you mentioned earlier about changing the way we eat does that somehow connect with your comment about wanting to reduce the the food waste yes because when we choose certain ingredients to go into preparation of food it begins from there are we able to make a better decision so that we do not waste food during the processing that's one can we select certain raw material and can have a better shelf life so that we do not waste food can we make sure that when the expiry date is being uh stated as of today uh this is going to expire how how can we change that is it because the the law said so or because they have gone through a series of tests that's that's a lot of uh because we have to abide by certain food regulation and that's where the food weighs i witnessed that we throw away food because it's just stated that it hit the expiry date without even assessing whether it is still good or not so this is something is yes we are abiding by the food law that's fortune no doubt but how can we do better how can we prevent this from from the very beginning but if we come to that stage how can we process that kind of thoughts by not just deciding to throw away the food all these uh uh i have honestly i have no answers for it but it's really can we not look at it in in a different perspective and upcycle it into a different differently so it's really about how to re reduce is the the way stage to begin with but before we reduce the wastage how can we do a better selection from a start and that's something right that's something yeah no i i i hear the passion i i think this is in it i i understand it's there's i i think what we've learned today is there's so much still to be undiscovered around food and so many areas within the the food life cycle that can be improved in pond and not just what we're talking about at the earlier part of the conversation the impact to the consumer which increases profits or helps reduce the affordability for that consumer the taste the texture the alternative proteins trying to feed more people around the world but i think you've also hinted that there's a responsibility you know part to this there's the distribution to individuals that maybe perhaps can't reach the food that we're we're producing and more importantly the the waste of the food that you've mentioned and here is i think another another really good use case for ai yes yeah okay i really appreciate your time today i know that you and i are completely in the opposite time zones of the world at the moment uh thank you so much for staying up late to be a part of this and i would love to have another opportunity to talk with you in about a year from now on the same subject because i have a feeling based on what you're doing things are going to be moving quite quickly within the food technology space okay i'll i'd love to yeah i'd love to yeah great thank you so much for your time today thank you marisa thank
2022-08-04 00:03
