[DECODED] Your Positionless Future is Limitless

[00:00:12] Phillip: If the first wave of AI taught us to think differently, then the next wave of AI is going to teach us to act differently. Today, we enter the era of augmented capability, where your next teammate may not be a person, but a prompt. And while we hear again and again in the media about the fear and uncertainty and doubt around AI, to a marketer, AI represents freedom from constraints and bottlenecks. Welcome back to Decoded, a Future Commerce limited series in partnership with Optimove. I'm Phillip Jackson, and in this episode, we'll explore how artificial intelligence is dissolving the boundaries between creative and analytical roles. It's no longer art or science, it's both. Strategy and storytelling, data and design, left brain meets right brain, and the marketer becomes something entirely new, a positionless force of execution. Joining me again is my cohost, Pini Yakuel, the CEO of Optimove. And together, we'll trace AI's journey from the early days of predictive modeling to today's conversational intelligence. And we'll ask, what happens when anyone in any role can wield the power of insight, automation, and creation all at once? And you'll hear personal stories of transformation, philosophical takes on generational change, and why AI might just be the most joyful creative partner we've ever known. Let's decode it together. Pini, I have to admit I used AI to prepare for this episode today.

[00:01:55] Pini: So did I. {laughter}

[00:01:58] Phillip: I've been a big fan of deep research. I've been using deep research and 03. But I subscribed to all the services. I'm paying like $50/60 bucks a month now for AI services. I don't know if that's a... Am I the only one?

[00:02:14] Pini: Wait, isn't it $100 for like the top one with the...

[00:02:18] Phillip: 200. Yeah. So I pay... I have to admit, and this is where my ops director gets on my case all the time. Because I want Operator on chat GPT. So I use Operator for some things. And then I want like a bunch of deep research every month. So I have the top plan on OpenAI. But I also use Claude for some things. I use Perplexity Pro for a bunch of data, public data sources and stuff. I pay for too many tools. I pay for all of them.

[00:02:55] Pini: In my mind, it's actually very cheap compared to what it does.

[00:02:58] Phillip: You're so right.

[00:02:59] Pini: Like if you actually look at the value and somebody says, "Oh my God, it's a hundred bucks." I'm like, "Come on." If I would tell you that you have infinite wisdom and it's something that gives you back hours and hours and hours and hours of time. And then it's a hundred bucks per month. I mean, that's ridiculous. By the way, I think they're still losing money on it. Right?

[00:03:21] Phillip: Oh, for sure.

[00:03:22] Pini: Yeah. So it's actually not a lot at all.

[00:03:26] Phillip: Today, we're gonna talk about this, it is expanding human capability. I believe that because I experience it every day. And I used to look for excuses to use AI. It is part of my all day, every day routine now for all kinds of things. It's also done something really interesting, and this is what we'll get into today is sort of how it's transforming the boundaries between creative and analytical roles and marketing here, because I think that's this particular episode on this five part series. But one of those, it didn't take long for me to change my behavior. My default used to be go to Google, or my default used to be to go to Amazon. Wherever it was that I would, like, go to spearfish a need, my defaults changed, and it took less than a year or two.

[00:04:25] Pini: Yeah.

[00:04:26] Phillip: And I have to believe, but I'm an early adopter. I don't know. How long did it take for you?

[00:04:32] Pini: I'm not an early adopter of technology, but I think I'm pretty much an early adopter of GenAI because for me, I've always been waiting for this, right? It feels like, oh, finally you're here. Thank you. Because I think I feel that my dance with Gen AI is a very natural one and we each play a very clear role. I think for example, if you're a person that you're an exceptional writer, it could be a little bit annoying. Could be a little bit...

[00:05:08] Phillip: So true.

[00:05:08] Pini: Because you may feel like it has taken your place or a superpower that you have had, all of a sudden, many people have that superpower. So to me, kind of like being a CEO of a company and kind of like managing people for a long time, it's very easy for me to basically direct something. Right? So like give instructions, I want this like this, I want it like that, I want it like that. And when you can do that, you can actually get better results from GenAI. So that's the first thing. The second thing for me is also I see it as in many ways, so I grew up as an elderly child. So my sister is fifteen years older than me and my brother's thirteen years older than me and my parents decided to have me.

[00:05:53] Phillip: Wow.

[00:05:53] Pini: And then I'm born to a 40 year old father, which is pretty rare at the time. It was 1978. So the time people had kids when they're like 25 and my dad is having me when he's 40. My dad was always the oldest. So among my peers, my dad was the old dad. Right? So he wouldn't play soccer with me. But he was this omni intelligent, you know, like to me, was like this oracle. We would walk, we'd have long walks. Like every Tuesday I used to go to his work and then he would take me home and he would walk for like an hour and get some pizza and ice cream and talk about life and talk. And he had this like insatiable patience. He could explain things about socialism and capitalism and cars and social structures or economies or relationships or his youth stories, whatever. He would just tell me everything about the world. And to me as a curious young person, that's how I grew up. I grew up with a very smart father who knew everything about the world as far as I was concerned, right? Of course he didn't, but I saw it like that. If I had a question, I would go to my father and I could get like a short answer, a long answer, a medium answer. I could get whatever I want. And I feel like it's a little bit the same to me, at least with AI. It's like this own intelligence that if you want to satisfy a piece of curiosity, if you want to now consult, have a good soundboard about something. And also because I'm experiencing it when a bit older now, so I think when you get to my age, I know we're not that far apart, right? When you get to our age, you have some life experience. So your decision making and your ability to kind of like say, "Okay, this is good. I want this," or 'This is not good." Your ability to choose, pick and choose between options and understand what you really want is much greater. I think for youngsters that don't necessarily have developed that trust in themselves or in what is real, especially in a world where there's so much conspiracy, so much false information, what is real? I feel like I got a really good compass in understanding what is real. So when I talk to AI, it's all pure. There's no trust issues, there's no problems. I know what I want. That's how I experience it. And I think it's just a treat. Like to me, it's like it's joyful. It's really joyful.

[00:08:47] Phillip: Totally. I have a very similar experience with it as well. And I feel like I have a, to your point as well, the way that I started trying to find the boundaries of it right away. It's like, what doesn't it know? It turns out not very much. And one of my frustrations early on was it doesn't have a lot of context about me. So I specifically spent about six months trying to give it context. So I would go on long walks and when the voice feature came out, and I just would talk to it so that it had all the context that it needed.

[00:09:33] Pini: And now it does, right?

[00:09:35] Phillip: And now it has all the context. What is wild is very recently, Pini, like I said, I used ChatGPT to prep for mini podcasts, just this one. What's wild is I will be in a completely unrelated chat figuring something else out, and it will bring in something else from another chat sometimes, for instance.

[00:10:04] Pini: Of course. Of course.

[00:10:06] Phillip: So the context is leaking right now. That's no longer siloed. And so I was looking for some factoid about Airbnb. I wanted some context specifically about this Airbnb story that I was putting in the newsletter that I also was going to reference in a podcast that we're going to do with Neil Hoyne. And it specifically said, "You should mention this to Neil Hoyne." Like it said that and I was like, "Oh, I forgot that I was also gonna have that on." But it had nothing to do with that podcast. It was such a wild thing. It remembered that I have that coming up. And that is the kind of thing that that is where we're starting to transcend these boundaries of like, that's surreal almost. We're getting now to a place where it is transcending a traditional boundary. And this is where I guess that first question is really here is, you know, how is it transforming a boundary between like a creative and an analytical role specifically in marketing? Because I think there used to be very clear boundaries between this is the creative side of a business, this is an analytical side of a business. But I don't think that there's boundaries anymore in the context windows of AI anymore and especially in my chat context. So are there any boundaries anymore in these roles in the future in an organization or ideally?

[00:11:48] Pini: I think, basically, I think you used the word roles, and I think you will start having roles instead of positions.

[00:11:58] Phillip: Mhmm.

[00:11:58] Pini: And your position could be a designer and a design team, and all you do all day long is you basically get orders to do a banner, do an ebook, do this thing for the website, do this image asset for that, for this. Now we have a big show. We have to create tons of digital assets for that or visual assets. So that's the only thing you do where it becomes, if you start to have a role and not a position, it means that I still believe that people will have core capabilities. So like I'm really good at, it's like 80% of what I bring to the table is design, right? I can tell you that your design does not meet our standards. I can tell you that your design is not going to be something that sells well. I'm 80% I'm a designer, right? But if I need to analyze some data to improve, I don't need to go to data department and ask for help. I can do that. I will ask their help. I will ask their help when there's a high stakes job. For example, if we want to really analyze deeply how we spend our budgets and what's the assumptions under that. And if there's a specific elaborated model of data analysis, which requires a data specialist where on their side, 80% of what they do or 80% of their core competency is around data. And they went to school and they started engineering and they have experience since they can write SQL and they understand all of this and they know how the math in some predictive models actually work. Great. But it changes, right? So for a lot of the tasks that are easier to do and I can be trained to do them by AI or I can be consulted on how to do them better or advance my capabilities by AI, I can close that gap. And then I may not call you that often. I'll call you socially, but I'm not gonna call you because I need your help with data. And I will when I really need a specialist.

[00:14:15] Phillip: This happens all the time, to be honest with you. So the more that you work cross functionally with other people in an organization, you learn shorthand and you learn organizational, I'd say, standards. Or you learn working styles. You develop rapport with people. You sort of learn to anticipate what role players will want from you, and you deliver that before they ask for it. That is just something that people learn to do because people want to reduce some friction. Also generally in human relationships, we typically want to kind of please other people. Although I've worked with plenty of people who I would say they definitely don't want to please me. {laughter}.

[00:16:22] Pini: Yeah.

[00:16:23] Phillip: But for the most part, I think other people are trying to work well together. And so as time goes on, most people are trying to deliver better quality work and first quality work better over time. And so you learn to anticipate what other people need. I think to your point, that is something that now we can accelerate. And to your point, then the creative role does become more analytical necessarily because the tools allow it to become that way and vice versa almost. Do analytical roles become more creative then?

[00:17:04] Pini: Yes.

[00:17:05] Phillip: Is that possible?

[00:17:06] Pini: Yes. A hundred percent. A hundred percent. I think it's again, the question is... Let's go back to the assembly line analogy. So ultimately a piece of work needs to get produced, right? That piece of work needs a little bit of data love, a little bit of creative love, a little bit of administration love, a little bit of strategy love, whatever. You got a few things that need to happen to produce a piece of work. And now the question is, if you're in data, maybe you can do some of all of the data and some of the strategy.

[00:17:46] Phillip: I see.

[00:17:46] Pini: It's simple task because you sit on top of the data, now with GenAI, you can say, "Hey, we're looking to do this and this and this in the business. This is what I see in the data. And let's think about a good strategy together." And you can do that with AI. And then you can learn about how to come up with better strategies with AI. So you'll do those two things. And the designer as an example, would take care of the visual assets and the copy and also maybe some of the admin. So how do I do it? How do I set it up in that system? Oh, it's pretty easy. You go to this page, you paste your work over here, and then you're done. So now from four professions, we're down to two. So from an assembly line of four different departments, we went down to two because one person took care of two and another person took care of two. So now basically we're faster. We can accomplish something much, much faster. And you know what it reminds me of? I remember, by the way, I'm not sure I mentioned it, but I'm an industrial engineer by university and academia. And later I've mastered in operations research, but the beginning was industrial engineering. And I remember one of the cool researchers that we saw when we studied about assembly lines, there's a really cool research that actually looks at what happens in a production line, assembly line. Well, you train an employee, you train Charlie Chaplin not only to work on his part, but...

[00:19:25] Phillip: Oh, yeah. Yeah. Yeah. Yeah.

[00:19:26] Pini: You train that person to work on on his machine, the machine before, and the machine after. So think about that. So instead of each person working on this one machine, you just do your department, you work on yours, the one before you and the one after you. So they did this research when they trained employees to do that and it was magnificent for the production line. Bottlenecks disappeared, the yields went up, quality went up, everything, all of the numbers shot up to the sky because they took a traditional assembly line and they taught people how to be a bit more positionless.

[00:20:05] Phillip: Wow.

[00:20:07] Pini: It's an eighties research, by the way.

[00:20:11] Phillip: That's incredible too because that's I think that that's just again, it comes back down to I mean, we could talk about that probably for hours on its own. Again, it comes back down to the human desire to have a little bit of empathy in that. It's like, oh, I understand now what someone else does, but it's also everyone wants to reduce friction...

[00:20:40] Pini: A hundred percent. I think that's one of the benefits. If I now get you, because I've tried to do some design, maybe I'm gonna be less prickly or less of a... I will understand. When you tell me take time, otherwise, you'll know, what's the problem? I just need to add a line. I mean, why does it take them two days? Because you never tried it yourself. It's not only that. There's more to it.

[00:21:08] Phillip: I wanna ask you, AI in the public conscious has been here since about late 2022, since ChatGPT, but machine learning has been around for quite a bit in sort of data science space. When did Optimove start this AI machine learning journey? Because I have to believe that that's been around for quite a while.

[00:21:33] Pini: Yeah. So we've been there from the get go. I started a company out of university. I was doing operational research and I was doing it with my co founder at the time, Dr Shachar Cohen, who completed his PhD in data mining. So we were like saying, "Okay, we don't want to do this production lines, assembly lines, inventory management, supply chain management," all of those fields that felt to us a bit more archaic. We said, "Let's do something in tech. It's not the computer science department, but we do have data mining. So let's do it like that." And that's how we got started. And our entire kind of like, you know, data mining engine was built on top of Markov decision process and MDP approach. Again, I don't want to geek out here, but it was pretty cool in terms of how it could predict future behavior of customers. And it did it at the time for like being that this is like 2009, we're just getting started. And at the time it was also very efficient from a computational perspective because you didn't have the cloud, these things were like really expensive. You didn't have big data databases, all of those things.

[00:22:53] Phillip: Sure.

[00:22:54] Pini: So that approach was very efficient. And we've always been kind of like trying to do things in a smart way. At the time, it was mostly about all of AI and machine learning at the time. It's all about data, right? You call it predictive models, predictive analytics, and machine learning, but it's always around, here's a dataset. Can you tell me which person in this dataset will buy coffee? You know, that's basically the problems you solve, right? Can you tell me who's going to be a VIP customer after analyzing only the first purchase? So questions like that. Today with Gen AI, what actually changes a lot is that the computer is now talking back, so the computer can write, the computer can reason, the computer can be creative, and the computer can also lie. Because if you don't allow yourself to be able to lie, you're not going to be able to be creative. I think the big change, yeah, that's something I thought about at some point because the evolution of computers, the first computers before GenAI are more like a person on the spectrum. So it's kind of like Dustin Hoffman and Rain Man, right? It's like, you know, error.

[00:24:23] Phillip: For sure. Yeah.

[00:24:24] Pini: You didn't input the right data. You didn't input the now you can input the wrong data and then the computer will give you a wrong answer.

[00:24:31] Phillip: Something's coming out. Right?

[00:24:34] Pini: And you're both fine with it. Right? You're both kind of like saying nonsense and you're both happy. So it's really funny in that sense, but this is really cool, so I think that's the big shift. But to us from a DNA perspective, we've been doing AI before it was cool. And we've been always trying to make predictions or organize data, make better business, allow our customers to know their customers much, much better by using a lot of these techniques. And obviously the dominating technique, which is the deep learning, the neural networks, that also existed in the world of predictive analytics. But to be honest with predictive analytics, while it's beneficial, it's always been kind of like less used because it's really opaque and it's very hard to explain. So if a customer tells you, "Why did your model predict that this customer will churn?" And I was like, "I don't know. It's a neural network." It's kind of like asking why did Chad GPT just gave me this answer. It's like, it's in the deep learning network, the neural network basically said that this word should come first and this word statistically should come second and this word comes third and this... That's how it compiles the sentence in a way that's, but you don't know why exactly like the neural net is convoluted and complicated. It's a function of millions of parameters. But yeah, we've always been excited about things like that.

[00:26:17] Phillip: When we're now thinking out into... So that's sort of the past in the history. Now we have all of these new tools, which are natural language based.

[00:26:31] Pini: Right.

[00:26:33] Phillip: So the interfaces have changed dramatically. The interfaces are also being sort of standardized and kind of they're moving up out of the UI/UX that I think we had a lot of control over, and now they're very free form, which is another, I think, an interesting evolution and challenge.

[00:26:56] Pini: Right.

[00:26:56] Phillip: This, I think, means that everybody has their own preferred way of engaging with AI. And I think that that means that everybody has new skills and a different type of a skill that they need to adopt. I think that that also makes things like best practices really challenging to teach and to enforce. So what kind of skills do you think marketers do need to develop to officially be able to collaborate with AI technology? And I think maybe that's a general question, but then I think I'd want to ask you specifically within marketing stack tools also that use AI.

[00:27:32] Pini: I think there's definitely going to be a shift that, you know, would be partially generational. So the kids that grow up today, or kids who are, let's say, you said the 2022, right?

[00:27:48] Pini: Yeah.

[00:27:49] Phillip: Let's imagine somebody being 20 at the time, now they're 23, they're getting their first job. Maybe a lot of the way, maybe even in college, they start to solve some problems with GenAI. So now the way they approach every problem for them, it's natural, right? It's very, that's a go to. As opposed to a 40 year old where they've spent a lot of years kind of like understanding UIs and looking for answers in a certain way. But at the same time with the 40 year olds, you get better decision making. And the way they interact with AI can be more responsible and maybe even more, maybe they can judge better if it's wrong, maybe when they should ask another question, but we probably don't know yet. I think on the UI side, the jury is still out in the sense of is all UI going to become a like one chat prompt? That's a question because if I ask you, if you fill out a form or you run through this process, do you want to do it in a chat or do you actually want to have an organized form that you go through step by step? I don't know. Maybe everything will become chat. I mean, I'm not sure about that. I think it's still kind of like being decided right now. And in terms of skills, to be honest for me, it's about the attitude. So there's a famous saying in HR, which says "Hire for attitude, not for skill." Because new skills can always be acquired if you have the right attitude. So I think you need to have the right attitude of like openness and curiosity and like this good, better, best mentality. We have that in Optimove, one of our values is good, better, best. Never let it rest until you learn your better is best.

[00:29:48] Phillip: I love that.

[00:29:48] Pini: So this is, I think that's the type of mentality you need to have, especially where there is a shift in technology, big tectonic shift in like a step function in technology. I think that's the type of a mindset you need to have in those times, in these times today. And if you would, you'll be fine. You'll find your way and you'll find the new skills that you need to hone and exactly how that works.

[00:30:16] Phillip: I can't think of a better place to leave it. There are, for sure, risks ahead. There's opportunities ahead. But I think that if we keep that in mind, what you just said, I think that there's, I think, an unbelievable amount of upside for anyone in a marketing centered career. Thank you so much, Pini, and stay tuned. This is gonna be an incredible series. We have much more ahead for you on the season of Decoded with Optimove.

[00:30:46] Pini: Thanks, Phillip.

‍