 Ladies and gentlemen, exchange for media is back with the fifth edition of Primetime Awards offering creative advertising in media agencies. A unique opportunity to showcase their best work done on television. Primetime Awards is one of the leading platforms to crown the absolute standard of creative excellence for television commercials. We are also proud to announce the first edition of TV First, a summit preceding the awards that brings together leading experts of television marketing domain to share knowledge on the effectiveness of television advertisements and how television advertising still remains a key consideration for marketers. Please make sure to block your calendars for 31st January 2019. And now we'd like to announce our second Twitter contest winner. May I please have on stage Dhaval Gurnani who is the second winner of the day of our Twitter contest. The most innovative tweets going out here. Dhaval Gurnani, I hope you're here with us in the hall. Are you? Or are you not? So I think Dhaval isn't here with us in this hall right now. We will try and find him outside and give him his gift or maybe call him up a little later. Okay. So ladies and gentlemen, the other day I was reading an article by JP Coppola's who's the CEO of Brands Eye. Where he's written this very interesting piece where he goes on to say that the future of humans is going to be dystopian. Now the future will see humans being replaced by a fleet of slick automatons who are destined to glue weld and solder us out of our jobs and diagnose, account and dissect us out totally. Now that's one part of the argument but this part of the argument is backed by heavy weights. Heavy weights such as Alan Musk or Stephen Hawking or that matter even Bill Gates. But there is a counter argument going on here and as had happened in history even earlier when the Industrial Revolution happened many out there said that machine would supplant man but that didn't happen. And today of course everyone says AI will supplant man but in the meanwhile AI continues on the path of sophistication achieves greater and greater heights but still even today AI doesn't have that one thing that we humans do and that is human insight. So this discussion is going to focus on human insight versus AI in marketing. And moderating this panel is Ms. Bindu Sethi who is Chief Strategy Officer Jay Walter Thomsen India and she's held this position since the year 2011. She spent over two decades in the world of advertising and marketing in India. She's also won the JVP Atticus Award for original thinking. She was largely responsible for setting up Thomsen Social that is the agency's arm that redefined the way communication was presented to promote new health programs in the rural and suburban sectors. May I please have on stage Bindu Sethi. Good evening. Very warm welcome to you. Please join us here on stage. The moderator of this next panel coming up ladies and gentlemen. And while she walks up I'd like to invite on stage the panelists. Anirudh Pandharkar Head Marketing VIP Industries Arvind Chintamani Chief Marketing Officer Colgate Pamaliv Juzer Tambavala Head Marketing Franklin Templeton Mayank Shah Category Head Parley Products Navneet Narula Head of Watson Customer Engagement India and South Asia IBM India and Pradeep Hejmari Chief Executive Officer Adam Wencher's Private Limited. And I'll hand this over to the honorable moderator of this discussion. So exciting times. You know I've been in market research, advertising and marketing for maybe 100 years. Exaggerating of course. And every time something new like this comes up and we have to think through the whole damn thing again. I think I'm a bit tired about it. But anyway with a lot of passion we'll go for this one. There's an exciting and diverse panel so I don't think I will have to... Why is somebody ringing the bell? Is it all already over? Then we can all go home or get drinks outside. And from diverse set of companies so we'll get a good perspective. Our conversation outside has been that since we've got a very good perspective of how creative agencies are looking at creativity and integrating machine learning and artificial intelligence into doing things. I pointed out to Tamara of course just as a digression that the Rembrandt happened because somebody thought of doing copying a Rembrandt which is an insight. And then of course execution followed through AI and she said, yeah, yeah, you can only think of that because you're a planner. But getting back to the subject, what I thought we would do and that's the conversation we had in the speaker's room is that each and every one of the panelists if they could and I think that's the big issue now that we have a background from Tamara and Ashutosh and of course I believe a completely another point of view from Mr. Sodhi which always happens and I always enjoy hearing about the Amul case again and again. What we would do is get everyone to speak about machine learning and AI in the way that they are grappling with it in their particular organizations, how they're using it, what is the issue, what is the struggle. They have what AI means to them, how they're trying to kind of address because how they're defining it and if this is the conversation we had, if they had a bit of money and time, what is the first thing to do? How do you get started on that? So that's how we'd start so that everybody has a good sense of what's going on. I'm going to start with you. So I think you put forward a lot of questions but I'll just give you a quick perspective of what the way I, I mean as an organization, we view AI and I view AI personally. To me it's essentially about if I look at what humans can do, humans can acquire and capture data, they can process it, they can reason with it and they can take corrective actions. This is typically what for me an AI tool should be doing and is doing right now. I think the only thing that today that AI can't do is really bring in the emotion, emotional aspect which humans actually have. I think from a business perspective, the big thing that we are really grappling is, I mean from the financial services business, I think today with customers, obviously their attention spans in terms of even where products are concerned is obviously getting more and more challenged and it's very difficult today for us to really get one step ahead of that customer to figure out what's the likely next move that he's going to make, what's the kind of product that we can really offer, what's the kind of pricing that we can really offer for that customer. And we're really building and investing a fair amount of time, money, energy, people resources in trying to build that platform for the firm to make sure that we can at least get, at this point in time, we'll write probably one step behind from the customer to get in line with him and then hopefully in future get ahead of him. Pradeep, would you like to take? So I've got sort of a couple of observations, you know, basis my experience with intelligence per se and then of course we expecting the machines to sort of come into it. One is that, you know, it has to be taught first, right? It has to learn and even before that is the process where we need to bring in things that it can learn on and that clearly is a hugely human effort. You know, so both of these in my view will, you know, need to coexist because there are, otherwise, you know, there'll be all businesses looking like one another if you're going to, you know, use sort of regression based modelling. Very clearly, you know, innovation will work absolutely in the opposite side because by all measurement sciences terms it's an outlier, right? And nobody's going to catch it. So if that were to happen then, you know, technology and so-called artificial intelligence will be the reason why a lot of innovations will never see the light of day. So I really believe that we are still and we can constantly see how human capital creates wonders. So I am a firm believer that that cannot completely be substituted. But yes, the computing-based strength which humans will struggle with when you're working to find something that humans will apply bias and ignore, that is something that a machine can definitely pick up and throw at you, right? So it can really enable us to do something at a far greater scale and at far greater reliability. However, the human will play for years to come, definitely a very significant role and so will human insight. So I mean, I picked up on something that Ashutosh said which is the fact that it's a tool. So in some sense that's what you're saying, that it's the tools and you need a lot of human thinking, effort and skill and insight to improve them, to train them so that they can give back to you. At least something that's closer and unique. One of the things that I think I grappled with, which you touched upon, which is a very interesting point, is that if you're all going to go to the same algorithms and the same machine learning patterns and big data pattern-seeking, then the brands that come out are the strategies and the way they engage with the consumers that they're going to use, are they going to get replicated and therefore where is the uniqueness? Because in the madness of human error lies some amount of distinctiveness and uniqueness and therefore where does all of that stand? So just one point, I know we'll get to the discussion later, but it's just troubling me as you mentioned it. While I said it can sort of work against innovation, if you have humans who can get the machine to tell you what would happen if this would be done, it can give you a fair prediction. So that's where I think it can still participate. However, it can never take out the way that a person can solve the problem. And it's all on the past, so if we are all talking about the future, it's what we will decide to do with it. So it can in fact give you things not to do other than to do. That would be a way of using it. So while we have been talking about human insights and artificial intelligence, I'll just give you an example from my industry which is luggage. Everybody is talking about smart luggage today. At the end of the day whatever Bluetooth enabled backpacks or trolleys and stuff, airlines don't carry them of course, because we know how delicately everything is handled by the airline handlers. Most of the times it's about the customer dissonance. It's probably over-engineering. Most of the times we need to be careful about, rather than just thinking, okay, this is the problem that the consumer has. We have trolleys bags at VIP. We have tried doing trolleys bags, just follow the Bluetooth chip and stuff like that. Biometric locks on the luggage, but the poor airline handler does not understand what a biometric lock is. And it creates a lot of customer dissonance and bases these kind of intelligent tools that we try to apply and over-engineer. So there are things which we need to be cautious of. Yes, of course there are positives about it, no doubt. Predictive modelling has helped businesses in my zone and we have been able to garner shares, get in new consumers and all those things have happened. But I think there's a flip side which we need to be very, very careful about. So the way we look at it is, I mean, I would rather start with, you know, unlike the market here of biscuit and confectionary would start. A little technical thing that we basically look at AI as three building blocks and three building blocks. One is the data acquisition, where you try and probably gather as much data about your consumers as possible. Then there is your algorithm or fuzzy logic as we put it. And then there is your application. And in good old days when we learned AI it was called robotics because it was more oriented to processes and process controls. So these are three basic building blocks. Now a lot would depend on what exactly are you looking at doing? Why are you using AI? Depending on what brand objective is, what you are looking at doing. So basically going back to the marketing funnel, so whether you're looking at driving awareness, whether you're looking at driving, let's say, consideration, or be it, you know, the final stage where you're talking about purchase. What exactly are you looking at doing? Depending on that, you know, there would be varied uses of AI. So just to take a case, let's say if I'm talking about the biggest biscuit brand that we are, biggest cookie brand that we have, Pallaji in our country, the consumption is varied. So, you know, I see consumption happening from probably, let's say a six month old to 85 year old and across various socioeconomic classes. It becomes really difficult to talk to them. So, you know, the way I would probably look at using AI or the way we plan to be using AI is, can I do mass customization when it comes to communicating to them? Mass customization, you know, I tell that you're talking about a segment size of one, you're talking and, you know, deciphering each and every consumer, trying and understanding what he wants. And how do I, you know, probably communicate to them, sell my brand to them and make him or her consume my particular brand. So, if I have, you know, so when I talked about data acquisition systems, the moment I have the big data and the moment I have the algorithm which tells me, you know, how is, so if it's consumption driven, if it's consumption that's what I'm targeting, then I would be looking at, you know, how they're consuming it, how much are they consuming it, when they're consuming it, or be it, let's say, communication where I'm looking at building some kind of a repo, some kind of relevance for my brand, I would be looking at, you know, how do they resonate with my brand? Okay. And since I have a diverse set of consumers, you know, how can I really break them down to various segments and probably customize my communication for each set of consumer. And, you know, with AI, I have a capability where I can really, you know, break them down, or as I said, mass customization, break them down to the sample size of one, and have probably a customized communication for each of them. So I think, you know, a lot would depend on how we can do that. And coming to the third building block that I talked about, which is, you know, the application part of it, where, you know, as earlier Pradeep said that, you know, the system needs to be learned. So we're talking about self-learning system, you know, self, so basically what we talk in terms of, you know, machine learning and stuff like that. Once the system starts learning, you know, then it needs to be put to application in terms of, while it would tell me that these are the varied consumers, these are their, you know, these are their requirements, or this is what I need to do. If I need to resonate, you know, or if my brand needs to resonate with them, deciding on the communication and then talking to them or, you know, putting that across to them. So the third part, which is the application or the communication is something that is, you know, going to be very critical. And that's where I see, you know, probably some more time before we can really graduate and use AI. But the first two part, which is the data acquisition part or, you know, probably assimilating the data. And secondly, which is applying the algorithm or machine learning is something that I think, you know, can pretty much well be done by AI. So when you're coming to the point where, you know, what is the role of AI that we would really see in the immediate future? DC role of AI in immediate future, at least in terms of, you know, understanding of consumers and probably, you know, reaching them, but talking to them or what to talk to them or how to communicate to them. I think, you know, AI has some more time to catch up because that's something where you'll require really human intervention. And as, you know, I would more than agree with what, you know, Deep said that I think, you know, it's that little imperfection that makes us, you know, different and brand different from, otherwise, you know, you're talking about cloning or everybody doing the same thing. And there won't be any difference between two brands. And from your sense, from my sense of what you're saying is that it's a great way to distribute communication in a customized fashion to amass a lot of people which you weren't able to do. It goes for that problem. It's much more communication, distribution through media which, like we've seen that era go from Dutertean to multiple channels, digital. And now you're being able to customize it because of the mobile phones. And there are two case studies. The one that comes to my mind is this international case study and you can look it up. It's on YouTube of how, and it was interesting that how acts picked on things that you could not use to create clusters such as fragrance, such as taste, such as sensorials which they got. And then they created some one-lack type of consumers and they created one set of coffee, cut it so that one-lack people would get in different ways. Right? I'm sorry, I think I was off the mic. Arvind, what are you going to say? He said a lot of things that for large CPGs, you know, I can do for, you know, very simply when for us in this country we have a billion potential consumers and, you know, two billion potential daily interactions and we're selling out of six million stores. The amount of data that is generated potentially captured is immense. There is no question that whoever can use that data will stand to gain it. So the question of the utility of an intelligent system is, I think, redundant. The question of human role is, I think, going to be the differentiator. Whoever does it smarter, better, faster in a world with fewer resources will benefit. So, I mean, we're doing a lot of experiments in very early stages. A lot of machine learning, whether it's in HR recruitment now with surprisingly good results on fit with organizational culture. In internal sales systems, you know, all the salesmen now take orders on electronic systems. They go through 100 to 100 SKUs. So the prompting of what that store might need based on what that store has been buying or not buying, you know, is basic machine learning. Intelligence goes in there, et cetera. In terms of, you know, programmatic buying on media, et cetera, that's obviously been happening. And various other experiments, right? We're trying to put cameras into stores to see how are people coming to a Kerala store and asking for toothpaste. You know, what do they say? Coolgate do? Yeah, Saada do? We've got a lot of data, but we want to see how it changes when the child comes in, et cetera, et cetera. And then, you know, where do they look, et cetera. So there's an enormous opportunity for us to gather data and then see what we want to do with it. So emotional inclination is hope and intent is high. I have a very pressing question which I'll ask after Navneet, which what does, and you can touch upon it if you like, what does that do for, what is the ambition for this? Right now I think the current ambition is distribution of communication and customization and things that we weren't able to. It's kind of answering you a wish list of how do I reach what part of my communication to which target audience in the best possible way without boring him on IPL or whatever it is by learning data. What would be the wish list? What would be the ambition for the next term and therefore how would you skill the organization? After you answer the first question then you can answer the second also and then we'll go in the reverse order. So guys, I believe in a simplest of the term you would have heard everyone what AI is all about, the technical differentiation. In simplest of the term, if used effectively, it can give you more time that you can spend with your family. That's personal. It can help you come to a decision point and decide quickly which you can defend either in your boardroom or in front of your manager that why you took that decision. Now, to me, in the simplest of the term that's what AI is to me. Now, what role does it play? It can play multiple roles. One, it can play a role of an intern. What it means is I can just ask a simple question which I would have made 10 clicks to get an answer for. To me, that's the basic level of AI that's been infused in most of the apps today which is more like an intern. It can play a role of a colleague where it can alert you if it detects an anomaly on your side or something is not happening the way it should. It can alert you either it could be in marketing terms, could be a result of a campaign which is not going the expected way or in a customer journey if you have a graphical representation or use a marketing automation technology and if someone is stuck in one part of the cycle for too long it can alert you. Final and the most interesting one to me is it can also act as an advisor. That means you give that tool an ability to make changes on a near real-time basis if he believes that action or rather if he believes the action on one of your sites got an audience could be moved faster following some other route. So a decision that you created could be overruled and the machine can make the decision for you if you give him that ability. To me, mainly those are the three key areas where I believe AI plays the role today and it will anyways enhance getting better and better as we move forward. So like we were discussing what do you feel should be the next ambitions of marketing organizations and how would you skill yourself to go because I think I don't even know if you're skilled to do some of these things in the perfect way. We have people who are doing machine learning analytics, are we doing it ourselves both from a sales point of view and also from the communication point of view. Ashutosh spoke about how mind-sharing is engaged with it from the communication point of view. I'm not sure we've taken the first step because even the boost example that you showed Ashutosh was created that was created by the creative agency. And creative agencies are feeling embarrassed but actually it's their product that's going out there and shining through a communication distribution system which is far more efficient now to be able to put that out and therefore get better results in terms of market share and consumption and so on and so forth. So the question would be for a marketing organization where it can use communication in two ways both in terms of sales something that you spoke about Arvind and in terms of distributing communication which is what Amayang's emphasis was how do I reach that guy with my message even on a one-to-one engagement basis. How do you scale? What do you outsource? What do you keep inside? How do you see the role? Because one of the pressing questions coming from a creative agency is how does the creative agency then evolve? It's not good enough to be a social media agency or a digital agency you need to be a creative agency in this world of AI, machine learning, digital and whatever else it is so if there's any thoughts that you'll struggle with, grapple with one of you to kind of speak about that. So let me start with a scary fact first. 2.5 quintillion bytes of data is created every day. I mean I actually googled how many zero does that mean and it said in British calculation it's like 30 zeros after one and I'll give you a scarier fact if that wasn't scary because you work with data every day. 1.7 megabytes of data would be created by 2020 for each individual on this planet per second and if you're a marketer you would be thinking wow like you said chief some for some who can use this effectively it will mean a lot but for some it may mean that you really have to work doubly hard to keep your company flow. Now why I gave those facts I believe while people will keep learning as they take this journey the three effective ways we see right now people are trying to address it. Most of the applications from tier one vendors now are infused with AI. So a mid-market organization who doesn't have a lot of money to spend on data science can get those apps and get on and start delivering on some results. I also see in the marketplace where people do outsource end-to-end marketing including the data work that's also there in the market and then finally as you would have seen where the technology the agency and the brand work together in a collaborative way where a technology is extended to the agency and they together work on a same platform to deliver results and I see that's where I see a lot of customers succeeding working with the brands. I'm not sure if did I answer what you wanted me to do? In some part yes. I think we'll get the details and then I'll come back to you later. I want some of the other answers. Sorry, I wasn't speaking to my mic again. Ashutosh keeps pointing me you weren't speaking to me. So you're you're saying how we should organize potentially to Yeah, because I'm the the thing I'm thinking and the words I'm hearing innovation, marketing, communication at the heart of everything is this piece of communication through which we engage with the consumer who will then buy the product because ultimately what are we here for? We are here for behavior shift. We say something to the consumer so that the consumer or all of us as people shift from one position to the other and somebody gains for it and what we say is as important as where we say it how we say it all those four boxes that are Ashutosh pointing Now what we say has two aspects to it one aspect is what the machine will tell us what we should say and the other aspect of it is what sharp human insight observers will say combined with the learning that they have from the machine is this is what we say Now in that scenario where there will actually be an exchange between what the artificial intelligence and the data crunching puts out to us and what we put back to that machine and see what the results are and that will impact on all the things which have distinctiveness and softer aspects which actually will lead to human behavior change because that's what we are chasing and therefore how do we structure for it because we haven't structured we've structured for that but we've structured for that in another era and this era and arena has completely changed in terms of what learning and skills that we require to that it's very easy actually to use all of this data and change the quantifiable meaning reach X amount of people reach this message in this forms to different kind of people everything that we can measure but things that we can influence so there's learning and there's influencing now influence is again constructed out of two things in my view one is hearing what you have to say and the other is what are you saying that I'm hearing for me to shift that behavior so how do we structure for that it's a difficult question and you know I think we need to even fully understand the potential for what is possible because from stuff you read some strange patterns are being identified that humans are not able to understand how the machine they've programmed is getting to it there was this piece of data that's a scientist fed in some 2000 profiles from harmony.com faces and profiles into his algorithm and after that for every face that he put in the program was able to identify the gender preference of that face to a 95% accuracy and the scientist had no clue how the program was doing it and had to pull the plug on it so we don't know what is possible fully yet at all right so in the absence of that knowing how to structure will be hard but I think the eventual joy will be from getting to nuances and to then use it to influence so for example Colgate's launched this toothbrush in the Apple stores in the US right now which has a camera on the head which as you brush transmits the images of the parts you're brushing etc. onto your phone and then you open your app and it'll tell you you've got a 85% score you miss these areas etc. Now once you get that for millions of people for millions of occasions and you realize that right handed people seem to be avoiding brushing that part of the gum line and left handed people are doing why etc. then you will design solutions better or it'll tell you more things like that or for example you know Who does it take to come up with that insight? So that's the question the asking of the question the imagination of the solutions that are possible when data can be patterned in this way I think that is going to increasingly be our jobs to say that to imagine the business problems are the obvious starting points communication is a solver of a business problem now that we know what our business problem is can we imaginatively ask questions that can drive influence I think is where we'll have to find ways to organize Bindu also you know as I shared earlier when I spoke earlier that you know while we look at it for communication but that's not the only thing there's a market here and not just as a advertiser it's not just the advertising or communication that we are talking about it's also the purchase and as I talked about the funnel so there's a huge amount of data that's out there you know when you're talking about organized retail and you know moderate formats about consumers how they are buying it and I mean also talked about you know now going directly to retail outlets you know being present in let's say lacks of retail outlet and going there and trying and collecting data but even otherwise today as we speak you know there is a huge amount of data consumer data that's already lying with organized retail on how they are buying it why they are buying it when they are buying it and you know there is a huge scope of putting you know AI to use there in terms of understanding how your consumers are buying your products when they are buying your products with what they are buying your products and you know running algorithms on them to better understand their requirements so while we as probably marketers more skewed towards advertising are talking about you know pieces of communication that we should be or you know how to target them how to probably influence them I think even you know down the funnel when you are talking about actually purchase act and you know the advocacy part of it once they purchase your product I think you know we need to look at the data which is already there out you know in public or you know with organized retail and probably work on it so to my mind I think you know I would not restrict use of AI just to communication or advertising part but even go beyond and probably you know go full 360 degree of marketing part wherein you are also talking about purchase you are also talking about advocacy and all. Yeah absolutely I didn't mean that I said you since you spoke about it all I was saying is that then if you are going to use it in all parts how will you structure for it will the structure change do you have any sense of it I am sure you all starting of it everybody is having this conversation about how will marketing structure itself in a way organization structure themselves in a way organization structure themselves differently Google structures differently and so on but organizations by and large structure themselves having a marketing department a sales department I mean my sense is this data management department will become bigger and then you will need a human insights department that will kind of train the data and I am just saying if you can take a punt at and we will start from this side of how would you structure your organization to make the best use of this new tool which is vast and has immense possibilities to give you stuff because they will come a time like it came to us that everybody had distribution muscle and everybody had so many things and so it became parity and so how do you structure for that kind of yeah so I mean with the increase in data the kind of cuts that are available most importantly I don't think there would be any greater change in the marketing structures who to think because ultimately the brand teams on the marketing teams have a business problem to solve and that business problem can be very well now defined by the amount of data that we have and the inferences from the data that can be drawn ultimately that's what a marketer's job is to solve a problem on a business front or the consumer front and the intuitive part of marketing may slightly come down because we have a lot of data which now helps us getting predictions forecasting and stuff like that but I am saying the basic structure of marketing in the next 3-4 years probably will have adjuncts of analytics to it but the core of marketing would still remain the same you asked the most difficult questions but a couple of observations I think one is that we should never try to get people buying into something out of fear and I think all of us fear this world with what quintillions of data 2.5 quintillions of data what registered for me was 30 zeros now I am immediately going to go to Amazon, hire a server and start gathering the data no but you know I think the important thing and I think people use this term very often data is the new oil and the other day I was talking to a friend and he said you know he read some article and so I said you know I just got one perspective on it oil has many properties one of the biggest one is that it's combustible and it's really slippery so that's the one thing that we should really watch out about data and the thing that alters the viscosity and the explosiveness is the humans and if we put it in the right place like the gasoline tank it can really help the person in control to do wonders so you know as much as I have you know sort of been learning and I don't think and that's why I was wondering you know where does intelligence meet learning so I think we're still all learning and most examples that are here are all you know more unsupervised machine learning or retargeting which is if you really ask me at the moment it's fairly branded you know I sometimes feel it creates post-purchase dissonance you know they target you for the same brand if you ever looked at it and decided not to buy it it's again time and again in your face so I mean jokes aside I think that's one that we shouldn't do right and the other thing I think intrinsically that we should be conscious of is that culture determines a lot of things right no AI tool told Google to structure its organization the way it did it defined its culture right and then they leveraged what came out of it so if we understand that there'll be a lot of conversations of consumers out there and we are always going to try and appeal to them we've got to see how we can make this a companion of ours on that journey to be better to be more specific to do really better creatives you know which people will really want to like flip the page on right and then come into the world that we're putting out an invitation card for so I think that's the real challenge because you know we are always going to be bombarded with a lot of media none of us can predict what will happen in the next five years but it's what we decide to do to ourselves that will either prep us or pop us out so that's exactly the question I'm asking is do you have any sense of what we should do with ourselves right now so I think one is that you know we should definitely not ignore this aspect right and take very keen interest in it to ask the most difficult like you are you know and the most human questions because at the end of the day let's understand one thing right we can be learned but we become intelligent when we hang around with learned people right because we take from each other's experiences we do more research and experience some things on our own and then become intelligent if machines don't do that with one another they will not develop that intelligence so it's also about our ecosystem you know there are marketing clouds where I don't know how many people will ever put the data out there but unless this thing which is meant to learn which a human will tell it what it needs to look out for is meant to play in various playgrounds it's never going to develop that intelligence so you know I think culturally we need to prepare ourselves and be open and know that it's a powerful tool but never make it a crutch don't try to defend your job using it but let it sometimes give you results that you want to be surprised so be open to it is what my you know submission would be difficult to comment after everybody said so much about that's the thing about long plan so I just pick up something that he said about culture right I think and you pose the question the panel that well what do organizations need to do is marketing as a team and a structure so the first thing I'd like to say is that you know I think skills can be acquired but orientation can't I think using AI tends to be today it's become a fashionable term and think about it in a marketing conclave today we're talking of a scientific tool called AI things are changing it's about the orientation of the organization two is that well am I going to use AI how am I going to use it but let me tell you it requires a ton of investment it requires a ton of ability to have the ability to capture good data today a lot of data and organization tends to be very very muddy and which is not going to give you the results that you're actually looking for getting the third thing that I want to say is that this is a it's a long drawn process it's not something that well I'm going to put up an AI tool and the next thing I'm going to get the right prediction of what I'm going to be able to sell it's not going to happen I think brands like Amazon have put in tons of money so you know I think one of the questions you asked about it what next so you know Amazon is actually testing this whole concept of saying that well one is to recommend a product on their platform which they're doing very very well they're now saying that well can we predict a future buy and actually send the product to the person's house right and what if he doesn't buy this is fine no problem we'll still give it to him free right but they're so confident of the fact that well they are today using data that's available to them not just of user behavior on their platform but also external data you know in a way they're getting into a position to actually predict so correctly I think the last point that I'd like to make is that you know this whole puzzle of saying that well is marketing now no longer going to be an art is it going to be a science I think today the science is giving us an ability to make sure that the art works in fact in fact that's a good note to wrap up and see if we can take some audience questions this is an interesting article which I again recommend everybody reads which was in the mint in which the author says and this is the headline and it's a good place to stop it says for the future to be a for the future of AI to be bright the skills needed are human which I thought was fantastic can we take questions so we run out of time we're out of time so thank you everybody I hope you made some sense can I have Mr. Pradeep Diwedi to give up the momentous please get the moment presenting the momentous Mr. Satya Brattadas CEO of Media Keys India and Pradeep Diwedi Sakal Media Group CEO once again we'd like to thank the Honorable Moderator and the panelists for their time and their participation here thank you very much we value your participation truly