The Fine Art of the Judgement Call

Hi, everybody. I'm Diane Brady, Executive Editorial Director at Fortune. What that title means, I'm not sure. But I do know what we're talking about. And I first want to thank Salesforce. You know, I love these conversations. They wouldn't happen without your partnership. So thank you. And let me introduce our fantastic panel. And their names are so wonderful. I feel sexy just pronouncing them. Shibani Ahuja, senior vice president, data and AI strategy for Salesforce. I'm not going to exaggerate because I know these things are chronicled for history, but I might. Okay. Valia Carboni. Did I say it right? Bravo. Bravo. Okay. Chief information officer, Shark Ninja. We should just talk about Europe, but no, it's okay. And then Athena Canjura, CEO, Latin America Foods, and global chief strategy and transformation officer of PepsiCo. excellent so we're talking about judgment and can i just say as a journalist i'm a little tired of the big boys of silicon valley the few that i think of being out there talking about their ai is gonna kill us or it's gone rogue and it's you know um putting in murder tips and i feel like let's speak since we're talking about ai and like do you trust ai can i just start with that like is it gonna kill us let's just do a quick let's just do a quick because these are voices i'm not hearing from, and this is very, this will be very interactive, by the way, so I'll ask you if you two think it'll kill us, but let's start with you. We're starting with a vote that will kill and... Well, no, but just how do you feel? Like, I feel like we're hearing a few voices and maybe we could do with a bit more, right? Sanity from people at the front lines of tech. I don't know if we're going to get sanity from this panel. We know each other. We'll get it. We're going to get the unhinged version of our responses that are peer and media approved, of course. We, so, thank you First of all, very happy to be a sponsor of this event and the partnership that we've got with Fortune. So thank you for that. So I think when we chat about this, it's like the fourth horseman. There's a lot of news out there. Is it going to kill or is it going to solve world hunger? My take on this is in the media, you hear what is presented to us. And I think right now there's a lot of focus on the frontier models. The models themselves are not AI. AI is so much more. So I think that what's dominating the news right now is the news around the models. specifically the testing that is by design and intentional around the models and the models going rogue by design I my personal take I think that's that's great I think that there should be no more news around them sorry what part of that there's a news that our news information available okay there's information we know what the frontier models yeah I think but I'd like to hear more about what are the guardrails that frontier model organizations are putting as they are testing these models to to be able to determine what their limits are that's on one side the other part that we don't hear much about is the organizations that are creating the guardrails or the harness for us to be able to use it in every day the models by design we're testing them to break free and see how far they can go that's hitting the news so you're not just an open door saying come on in what have you gotten if i was 13 we'll take it right you're you that's a very good point i haven't thought about that how about you yeah i Oh, I agree. And that was very well said the way you said it. I mean, we're not, I think there's too much focus on the wrong things right now. I think we want these companies out there breaking things, seeing where it could go, all of that. I mean, we wouldn't want to lose that as a country. So like, I'm a very big supporter, like I want them to be a little bit rogue because I'd rather we find it than someone else finds it. And then as a company, of course, we're putting the guardrails in place. I don't want to limit creativity. So I push our teams, all of our teams, all of our functions are in like let's go but there's always someone that's sort of policing it like as a CIO I have to have the guardrails and security all the things I worry about but I don't lead with that story because I think that also takes away some of the creativity we want to see people and really trying to transform the way that they run their their area of focus so I think it's an exciting opportunity and I think it's all about the way you lean in and obviously it's always a verified like you're not just going to trust anything coming in but I think we've got all the right protocols to manage that. Okay, that's good. Would you like to offer a perspective that is similar or differing? Yeah, for sure it won't kill us. I will start with that. We have 320,000 employees and billions of consumers where we feed and hydrate the population. And we use AI everywhere. So if I felt that it would kill and destroy, then we would be the first ones not to use AI in every part of the operations, right? So I think it's all about businesses of our size being extremely responsible of how you use the technology at every facet of interaction with consumers, customers, employees and in your operations and if you were to do that because this is how you touch every life in the world then I think no matter what the tech companies say that myth will not turn to a reality because ultimately they are not engaging with consumers they are engaging with consumers through us so we have a big responsibility Good. Okay. So I feel reassured. Let's go to judgment because it's become a word that's sort of more in vogue now and I understand why. So I want to unpack it a little bit because these are tough decisions, right? In terms of what do you, in essence, sort of outsource? Where do you sort of find the right line with regard to where human judgment is? And it's one of those terms that can be overused a lot because you actually need a playbook to make decisions. So give us a sense of your own playbook. Maybe start even on a personal level. You've talked to several hundred CIOs in your previous role, right? How do they define judgment? How is the term evolving with a technology that can now take over things that we used to do ourselves? In two minutes or less. In two minutes or less. Okay, we're going to talk really fast. Can you follow along? No, no, no. You get an extra minute. No, because we paid for this. No, I'm kidding. I'm kidding. Never this is pure judgment no yes this is pure uh okay i asking a question on judgment and ai i actually kind of go back to the definition of ai ai is i've i've said this many times it's the smallest biggest word or or two initials is like people are just saying ai ai ai to everything data center is ai like the model is ai we're at salesforce because now so you're to your point i just spent the last two years advising c-suite leaders now i'm implementing ai for our own organization i'm actually breaking it down to say what is ai and i'm breaking it down to four modes ai is assistive ai which is like the day-to-day slack bot gemini that everyone has and you've democratized ai then you've got um like citizen built ai where you've got finance people that are now being builders and hr people being builders and we want to encourage that and you're giving them maybe co-work cloud co-work for that and a more advanced model then you've got functional ai where you might maybe taking a level three process in an organization and you're automating it end to end and so then all and then the last one is to me uh cross-functional enterprise AI where you've got agents that are coordinating across maybe like the lead to cash process end-to-end multiple departments judgment in each of those varies and so i think answering the question in two minutes is hard because i think that it should vary as should the ROI as should the controls as should who are the builders determining the right of an appropriate amount of judgment as you're moving from what is maybe like predictive AI through to generative AI to the end being autonomous well let's talk about before we move on let's talk about your you're now implementing this through a tech company right and you're all i think if you're all incredibly you all have tech embedded in your DNA what are you making how what are some of the toughest decisions you're thinking about because there's the readiness there's also an evolution right as we learn more we can do more give us a sense as to what you're seeing right now in this role and where are you seeing I hate the term human in the loop, but where are you sort of treading more cautiously in terms of trying to get it right? Okay, again, maybe I'll give an example of it. I think in the early days, it was a lot of conversations around what can we automate and what can we just let the AI do? And that went out there, and people tried that. And our earlier versions of harnesses were still very probabilistic, meaning you ask a question, it's going to give you an answer. Ask that same question, different answer there. Judgment should not have been left to the AI. there should be a deep number a deep amount of human in the loop i think now where we're getting to where we have the ability to have both probabilistic but deterministic guardrail we're now able to give a bit more judgment or autonomous action to an agent and instead of the agent early days doing the work and there not being anyone checking it but like assuming that it's right now i think we're getting to a state where we've got agents that are performing the work and perhaps a lighter validation or spot checking from a human to ensure that they are looking at yeah it's and it's lightening the load on the human to do less of the admin work or the check and more of the strategic insights work do any of you have digital twins yes you do how is mine's really boring it's like a tamagotchi pet that spews my LinkedIn profile back at me and i'm like oh i better rewrite that give me the experience of your before we i mean the playbook but what is your experience because the digital twin in a way in essence you're trying you're hoping it captures your judgment, because that will be your avatar for making, so let's go to you. How is your digital twin? Do you like her? Actually, I'm going to use my colleague's digital twin. I think it's where we spent a lot of time. So our commercial org, and we actually shared some of this work recently with Salesforce. In our commercial org, we're a very rapidly growing business. So Shark Ninja, 25 new products a year going into 15 countries. And we don't want to just, you can't just keep scaling with humans so one of the things that we did was really get you said with Cubans humans humans I'm sorry that was my forgive my Boston I love you but no no no it's not forgive me I'm just you know with human making sure okay so it was really how do we replicate I don't know where the Cubans came from we'll talk about that's me forgive me sorry but as you were so one of the things that we wanted to do is it's a founder led company so you know mark and some of his key people they've been running the business even as we've gotten to seven billion dollars kind of doing it the old way so we knew this was like a very ripe area on how do you make better judgment calls on where should we move inventory because we're like about scarcity with inventory and a lot of what we've built now is to replicate sort of Mark's our CEO's brain and some of his uh chiefs that he's got underneath them in that commercial org and really allow them to have a digitized way of automating some of the insights and where we saw opportunities to move calls like where do you move the inventory and it's really helped us and all of that to say because we want to satisfy that consumer there's nothing more frustrating you go to our site we're out of stock or we don't have inventory at our partner stores like it's very frustrating from a consumer perspective so to do this at scale so you don't have to pick up the phone and call Mark to say, hey, what should we do? It was a really interesting place for us to start the digital twin sort of mindset so that we could do that at scale as we're going to 15 new countries. Yeah, I know you guys have been on a tear. Yeah. I think I own a bunch of your products too. I feel like I should be wearing the t-shirt. Well, yes, there's a limit, but yes, they're very good. And also, let me go to PepsiCo because I think about the amount of, I've always thought of it as an innovation machine in terms of even just the way you highly localized test market you're head of Latin American foods I mean how are you let's talk about your digital twin what purpose does it serve because i want to in an i want to kind of crystallize how we think about judgment in our own lives first before we start you know democratizing it scaling it building infrastructure around it yeah and there's a consumer digital twin and then there is operational and engineering digital twin right so and i built a digital twin when I took over the, and I came to the company as strategy and transformation lead and I started kind of adding responsibilities, so processes, capability centers, the technology, all the business programs, right? So, it was very hard for the teams to be able to understand kind of the full thread of why strategy, business, transformation, delivery, service, processes, tech would sit together. So, the whole concept of the digital twin was, how do we synthesize the strategy of the company through creating a consumer digital twin for me, so they would be able to do much more what we call FAQ for the rest of the XCOM and the organization. And then of course it has proven to be very handy since I got a second job with Latin America. But just a bit to anchor to your earlier question about what do we see the biggest challenges. I think for a company which operates pretty much in every market besides North Korea, you see various levels of adoption, change management willingness and pull from the markets so you have markets in Asia where they say I want to be all agentic China, Indonesia, Malaysia South East Asia where they say I could care less if I have 1000 employees or one employee just want to be agentic so their level of acceptance of the agentic transformation is massive and then you have a market like the US where they say, no, no, no, no, no, I want to control everything. It's like, no, I want the agent, but I want the human as well, and I want the agent. So I think what we see is a big divergence in terms of how employees, especially, and then subsequently consumers, accept the fact that there are elements of the business that are agentic first, right? And therefore, the guardrails, to your point, are very localized. I don't have the same guardrails in China, versus the U.S. Well, it's interesting. Before I go to Spine, I want to hear from the group, obviously, all of your own experiences, but that makes you an excellent test kitchen in a way about the sort of human AI judgment. Do you find the decision-making or the results in a place like China any different than here's? And which do you think, now that's our future, and I mean, obviously, markets are contextual and localized, but since you now have experience with people who've embraced it, and people who resist it, what's been the outcome difference? In China, 99% of the cases, they trust the output of the agentic, you know, either system or the agent they have within them or the broader context. They trust the output. They don't second guess. They believe it has been vetted and validated already. They don't go back and say, okay, show me the traceability because in reality, China is much more of a semi-closed ecosystem, so they know they have been used to that type of cultural pivot comparing to everywhere else, which is much more of an open ecosystem. So there is distrust in the technology, in the output of the technology. The more you go to the Western world, the more the mistrust comes into play. And therefore, you need to prove the stage gates. You need to prove the traceability. You need to prove the source. You need to prove that there is no bias or there is less bias. So they need the proof points. yeah this is when people always then say okay now that you have given me the proof points then i will trust the output this is a this is such a true like we're finding it ourselves too of you know i think there's such a push like we want to transform as a leadership team and how do you like mobilize an entire organization and when you think about what we do it's a very creative innovative company there's a lot of hands-on you you walk around the office there's hard products with wires coming out so people are just very used to relying on their own instinct we do so much consumer research but you know as you think about that scale 25 new products coming out a year like I don't see a world where you're not leveraging it but it has been a hard transformation for people and I think every article you read it's how do you bring people maybe in the western world I fully agree a lot more on the journey of change because I think in many ways it's like insecurities come out like can i do it is it going to replace me and like my advice i always tell people is like get in the game like you just got to be in the game and then trust me like the opportunities will come but i still think as a culture we're still there's a lot of resistance on there's a pride there's a pride we can do ourselves and before i go to the audience i want to ask you know i always think of money money money money and we had somebody in our office who i you know wasn't me but it was um you know i think about token maxing and they used perplexity compute to basically give us a list of restaurants in Dallas and you're like you could have just called like it just was a bad use of compute because you're kind of showing off so there's there's now concern of course about the price and so part of it is we want to double down but at the same time it's getting expensive you're on the front lines of that is there anything about the economic picture that's impacting how this plays out right now um can before we get to that one i can i come back to the judgment one please yeah be the financial judgment let's talk about the judgment yeah so judgment just judgment as the word as we were preparing for this and i was like judgment to me i think judgment is a combination of uh having the facts and the data but also feeling right it's a combination of the two and as i think about ai and the interlock with judgment i i i think about where ai is enabling us to improve our judgment like the the thing is you'll get asked a question and you'll have a feeling whether you've got the answer to that or not then you try to collect a couple of facts or data points to help inform what your decision will be you make your judgment call with with ai and i know this because we're modeling almost like a digital twin for robin who's our cfo ceo of the company we're going we're not going to replicate everything that she does but what are the data points that she needs to make a better judgment call that otherwise it would have been impossible for any human to compute the correlation between this over here this over here this over here that's happened across the organization now we're able to do that part with AI to improve her judgment call so i think that that's that i just wanted to say that that's a fair point there's that combination of both like feeling and fact okay switching to to money okay money well but not no but in the sense that i think it in there when times of scarcity and times of plenty impacts our judgment around how we deploy experiment and roll out right it's what we are going through at Salesforce uh it's then it's like to frame it it's we have to democratize AI for everyone everyone should be a builder but wait either we don't know where the value is coming from how are we going to identify the cost structure or the the ROI on this again kind of coming back to that framework that i talked about people are trying to associate value with colleagues getting either Gemini, Slackbot, or co-work. I'm going to stop wasting your time. That is something that you just have to do. It's the time and place. Give people access to that. Now, pause on it for a moment. Recognize and meter and observe who is using it. Start to look at not just who's consuming it, but what are the skills or agents that they're creating? What's the reusability of those skills? You're going to identify some power users. That's a value in itself. It's how you identify value. It's efficiency, effectiveness, experience, risk and control so if letting giving everyone access to these tools is allowing you to observe patterns in your organization to extract those super users that happen to be buried in finance that's one element of value that you've been able to identify elevate them get them going and doing like tiger team things when you get to mode three where you're automating a process like a level three process within one department one function do not do that until you actually have what the outcome will be you should be able to measure that when it comes to mode four which is like the cross functional, have every person in every function that is signing up for an agent that's going to go across, those are the ones that you want to start to measure the value of mode three and mode four relative to the overall cost of token consumption. As far as it goes on token consumption, we've kind of opened up the taps. We've said, everyone try it. We went through what a lot of companies have gone through to say, oops, that's a lot of spend. But we're now going, rather than turning it back off, we're actually, again, looking for patterns of who are the super users, Smart token. Exactly. So we're spending more time on analyzing what models are they using? What are they using it for? How do we publish recipe cards that are markdown files that enable people to optimize use? But here's another thing that we're doing is we're giving everyone access to a dashboard that allows them to see how much they spent. Self-regulate. We've given them a budget. You are now a leader. You now have a budget. You manage your budget. And when you run out of that budget, like it's teaching people how to regulate and use the right models for the right reasons. It's like giving an allowance to your kids. It's an allowance, yeah. Be responsible about it. That's very smart. Now, do you have any questions? Just raise your hands or your own experiences. I can go on. Okay, Marjorie. Do we have a mic? Do we want a mic? You can just project. We'll give you a mic anyway because you deserve it. You deserve it. You deserve it. So we started an AI lab almost nine years ago, and so we have a lot of product from it. And one thing we've done on the judgment side is we've taken how we organize the agents and then we have added a layer where we have mind mapped 200 people that come from different places with different experiences. And we have embedded in the system the thinking process of those people so that, I don't know if anybody else has done that. Are these personality types? It's not personality types. personality types it's like so i was mind mapping it's like what's the hardest thing you ever worked on how did you approach it what is the mistakes you make what wouldn't you do again you know that kind of thing and then there are people that um come from very specific areas of experience so they were in politics or they ran a business so you're getting very different inputs yeah but then we have aggregated that and that really has made a big difference in the output and um and and we We've done this in a bespoke way for each of our clients so that you can take the thinking that you were saying of mind mapping, Mark, in that case, and then you can add that to it because it doesn't leave your system. But that may be one thing to think about as people try to deal with the judgment question. Yeah. Yeah. So anyway, just wanted to- Excellent. Do you want to hand it back to Hanneke there? Yeah. I was super interested, Shark Ninja. How are you leveraging AI in your innovation process? Because you guys do it fast. You do a lot of it. There must be AI at several steps. Can you talk a little bit about that? I always laugh. I'm like, how much more can we do? But 25 new products is a big. And they're new products. They're not new colors, new versions of the prior. I actually think there's a fascinating opportunity across the board, all the way from the way we do. We do a lot of consumer research. Like, our mission is finding everyday problems and solving them around the world. So it's how do you really get more of that intel in an automated way as we go into 15, 20 new countries a year? There's a lot of localization of needs, obviously, size of apartment versus house. So I think we're seeing it a lot help us there in more consumer insights and research. And we're starting to see it play a role in the way we actually do the engineering and design work. I will say the tech is coming along. It's probably not where we started because the technology wasn't ripe enough. but the research side and then the transformation that we're doing and the way we run a variety of functions to bring products to life and get products out into the marketplace i think that's probably the biggest pockets that we've over invested in but one other thing i wanted to add to this whole value conversation I think it's something we've struggled to when I joined we we launched 200 pilots we gave everyone licenses uh go use whatever you want yes we had a cap yes we've done the P&L here's your little P&L everyone gets to be a leader and it's been great I think you need, every company had to do that because you had to get everyone in the game. But I will tell you, even for us, and we're, again, innovation, like, is in our blood. Like, that's what we do. Every day is a new idea. Like, what else could we be doing from an experience or product? The thing that really helped us propel the change and drive value was we did something back in April. We called it Jailbreak Live. Jailbreak Live. Jailbreak Live. Not from Cuba. Just Jailbreak Live. Oh, no. I'm never going to live that down. I love Cuba. I'm Canadian. I'm allowed to go there. And we shut the company down for a week, and we just wanted everyone in, and we defined 20 big hairy problems that were kind of slowing us down, all the innovation, like even us. Did you solve any of them? Yes, we did. I will tell you, 20 went down to like five that are super active that live today, and it's less about 20 to five. I celebrate the five. The five are game changers for us, and I'm so excited about the work we're doing. Do you want to share one? Yeah, one was around this inventory allocation opportunity we had. We go for a lot of scarcity, and it really was bringing cross-functional teams together to work on one common goal, like how do we do demand planning? In a way, when you've got a growing D2C business, by the way, if you find us on TikTok, we're number one or two around the world, any country you go into, we're like one and two. So like a lot of demand, and then our amazing partners with the big retail shops across the world. So it was our opportunity to bring these folks that, you know, a little controversy between the team sometimes, where does the ball get dropped? We don't have inventory where it needed to be. And that was one of many, but it was like, I do think every company has to find their, what's going to unstuck you a little bit. We call it unstuck because I think you do drive value when you can drive the right folks. Will you spark a thought that I go to Athena with when you mentioned TikTok and being in the media business, right? I think about, I think PepsiCo, many companies in this room, you know, you think the direct-to-consumer, when you can create AI-generated advertising, you can create avatars now that are the perfect genotype of the, you know, actor you're trying to portray. And yet, I feel I can spot it. I feel there's something that people value about the human creative process. how I mean how are you thinking about it as you would have outreach in multiple markets multiple touch points a it's a clear use case for AI and yet we know there's something that makes us just think uh you know I don't want a movie made in five minutes I kind of want the filmmaker to go on the streets and shoot the real thing yeah well people care about experiences they don't care about products and we are we as every other CPG company we are predominantly a B2B to C business. Right. I mean it's very rare that you find direct to consumer type of engagements in our industry. Partly because consumers don't want to buy a kind of Pepsi standalone. Right. But they would like a full category experience. Right. With the products. Partly because the economics don't work in terms of the business model and partly because we do have multi-channel approaches so i think the power of the media and what we call kind of the the channel mix and the omni-channel experiences social commerce being one of those agentic commerce being the the next frontier is how do you translate that experience to a viable business model where the consumer benefits and you benefit financially. You cannot make the P&L suffer every time just because your consumer wants another experience. Yeah, which consumers do want different experiences. Let's not kid ourselves. So what is that intelligent layer that you need to pair every channel, every point of interaction, every point of sale, that you need to layer on top to make sure of that trade-off where demand planning gives you visibility, the go-to-market connects with the supply chain, the consumer activations connect with experiences, and the product portfolio management then becomes in the service of the consumer. So this is for us what AI is all about. Because if I go channel back only, if I go only consumer back, then we have seen again and again and again for a machine that has fixed cost in the system, we are not born in the digital world. there is a physical infrastructure that we have to deal with, the maths don't work. So this is where agentic and AI gives you a lot of flexibility to take the fixed cost out to create superior consumer experiences and as much as possible create agile supply chain. So I think that's the beauty of the physical and AI world together for businesses that require a physical infrastructure. If I could just add to that. Please. We're a big, I mean, I think one of the things I love about us is the storytelling. like we're big storytellers about everything we do because we're creating you know we're solving problems we're also creating like these very delightful moments and a lot of what I talk about TikTok like it's not AI generated but the insights that we can get from what is trending absolutely is a feeder into the way we run the business but I do struggle like I struggle to like I prefer authenticity over don't automate it for me just because you have a better way like it that's not going to drive my brand love for you, but I think where I find the benefit is our ability to harness all of that knowledge in real time to say, hey, we're about to pitch a new media campaign and we're seeing this trending in our social channels which are extremely active. That is a great use case for me, maybe not the content itself. Yeah, I think that is where judgment comes in. Do you want to add something? AI and media, it's not a new concept. we wouldn't have Jurassic Park if we didn't have AI and media we wouldn't have Jaws or all of those things. I think that dog reminded me of Velociraptor meets Collie don't you think of that mouth the one downstairs the robot yeah I digress it's not a new concept I think that you may we're not going to tell our marketers don't use AI because we want original and creative thought like actually AI is accelerating them but to your point like I think the values of a company the storyline the how we engage the market that still has got to be uniquely ours but ai is going to be used and has been used i love jurassic park i wouldn't want that it's fantastic much better than whatever they used to be more of the gargantuan you know i have a question i'm we raise your hand but one of the things i'm curious about to get back to judgment we mentioned the zeitgeist in the u.s which has been relatively negative as business leaders who understand the value proposition what is going to change the hearts and minds of people who are nervous who maybe don't you know have the level of sophistication about seeing how it plays out what would you what do you think should be done and what are you doing to help shift that narrative now I'm looking at you first okay all right I think that the starting point is there's two ways that you can you can start with let's and let's just contextualize it with colleagues in an organization there are colleagues that are afraid that they're going to lose their job and that's the it's a real fear there's two ways of doing it either pretending like that doesn't exist and like just like talking about the the beauties of ai or starting with colleagues like starting with them sitting with them getting them involved getting them into the not having these councils that's of senior vice presidents and presidents that are talking about ai and ai strategy but actually starting with the colleagues and so that's actually where i'm starting in our cofo organization cfo coo organization is i don't care what my org structure looks like. I've come in as the AIs are. Literally, that's and I'm going, I actually don't want an org chart right now. I want to find those that are independently passionate about what they're doing. I want to borrow them. I want to start tiger teams. Those tiger teams are going to run a bunch of experiments on ideas that they have. We're going to get that You create FOMO within the organization. Yes. Everybody wants to be on a tiger team. Yes, and that's the, I don't like, again, I actually don't want people reporting to me because then people are like, oh, this is mandated. I'm like, can I borrow? Can I borrow? Can I borrow? I've got a guy who I just got from another team member. He's a manager. I put him in front of Robin, who's our CFO, COO. And he's like, what? I'm going to be working directly with her on building out her AI twin? I said, yeah. And imagine, now he's going to get profiled. And he's like, OK, I'm game. That changes the mind. And he's the one spreading the word about what we're doing, how we're doing it, why we're doing it. Also, I was in marketing just before this. We, early days, created a channel that was bite-sized AI. We gave marketers access to all the AI tools possible and said, have at it. But your only ask is publish your hacks in here. And all of leadership was on it, and we, like, were supporting and promoting all the hacks that were happening. And that was, it was, like, homegrown. Okay, any tips that you both have? Yeah, I think two things. I think you've got to do this organic, like, get everyone excited. And I do think there's a little bit of a fear factor in a lot of humans, and it's natural, right? So I do think we've got to bring people on that journey. We've created these AI pioneers, so they sit randomly in various teams. They don't report to me or any of the other transformation leads, but we're trying to get the excitement going, and I'm trying to demystify. A lot of us are out there trying to help where we see people struggling. But I also want to take it to a personal level. I have a 22-year-old who in his senior year at Colby College in Maine, lost tuition, billed a paycheck. Yay. Yeah, but I'm looking at it through his eyes, and I'm like, you know, I do a lot of mentoring a lot of coaching and now I've got my own kid I love to see like what are you gonna be and I think for me he interned with us this summer He's got a different last name. So actually no one knew he was my son, which was brilliant And he said to me well, I don't know everything. I said all you have to say is yes I've got it because I think that's what I mean by that is with the accessibility We have with these great tools You don't have to be an expert in anything like all of us are figuring it out and you can figure it out pretty quickly by leveraging these tools and like my advice to him is like go big like go after all of this like I think back to when I graduated finance major and now I'm a CIO you know very strange journey but it's like I was very prescriptive in what I did and I think the brilliance of these young generations coming in now and for ourselves to reinvent like you can go do anything you want to do like you can learn it pretty quickly and I think it's just having that curiosity gene, that willingness to learn and be vulnerable. It's okay to be vulnerable. As a CIO, I say all the time, I have no freaking clue, but I'll go figure it out. And I think it's okay. I think we got to make it more, like demystify it a lot more than I think we have. But it's interesting you mentioned that because, you know, learning is a kind of messy, deep process. And I actually hear more CEOs now telling me they read books. I had one in particular, head of a big bank, who said, getting all these AI summaries I realized I wasn't as good at connecting the dots because there's something about reading the document that gives me the nuance and the time to like I'm paid to think yeah and if you're paid to think it means you have to do the work to then formulate like for me the act of writing is inextricably linked to the act of thinking but the act of research isn't i mean i find like the socratic you know what's what's happening who's you know it's So it's interesting, all the world's now B+, A-plus becomes a much higher bar, I think. What are you finding? Yeah, I think for the front line is our sales supply chain people This is a massive unlock. They love it. They love it because now they say, you know what, I don't need a computer science degree. I don't need to be, oh, okay, I can use it in my day life and I can use it at work. So full adoption, they love kind of the access to the technology. It makes them more effective, more efficient. better in their job however there is obviously a big disruption in the professional population whether we like it or not and actually it comes a lot from the tech companies who go to the boards and who go to the chairman they say oh by the way you know you can kill 60 percent of your finance organization or hr or because we've done it listen to us and then they come back and they say oh why aren't we doing it so a bit like realization of what is happening in the market so for the professional population is a massive disruption massive disruption in terms of how you rethink the job the skills the taxonomy we've been in the past years through a massive massive redesign of the taxonomy of every function it's because not because someone told us is because we were seeing that the people who were coming is you need to do it otherwise i feel i will be out of a job when we have the financial planners now not doing forecast on top line on bottom line or middle of the P&L and they have the AI forecast doing everything for them so what is my job then is like why do i need to be a financial planner if this is not something on my responsibility so we have had to go and redesign all that population which they made them one much happier second at least reassure that we're giving them the opportunity to come to speed with the latest technology and redefine the profile of what it means to be x but i do have to say that there is at least and i talk to many of my peers and other CEOs at least 30 percent of that population that will not make it and that that they are resistant to fully leveraging this capability. And they are more than happy to say, I'm going to do something else. This is not for me. So I think the responsibility of the organization, every organization, do you give your people the chances to upskill? Do you give your people access to the tools? Do you give your people the career, architecture, and opportunities to move up by leveraging the tools? But also, you need to make it very clear that these are the expectations of the new world where you have to adopt now these new realities. Do you have any cool new job titles yet, any of you? I keep waiting for people. You'll redesign. What is it called? Legal engineers. Legal engineers. Okay, that's kind of cool. We have to sum up at this point. And I was like, so you're all global companies and we know these play out differently in different markets. The regulatory environment, sovereignty. I mean, we could, let's just talk about that over lunch. But let me give you each a chance to give some final thoughts in terms of let's go back to judgment. you know, that the skills, the power, kind of the way in which we have to also lead the roles in which we're in advice you would have. I'm going to start with you, Athena. Yeah, it's, if you were to think the population and the composition of the population in terms of GDP contribution for the lower and lower middle part of the population, this is a massive unlevel growth. So if you were to think creation of wealth in these lower parts of the population everywhere in the world, this is a massive opportunity for the world. So that gives me a lot of optimism because we sit in the company and we sit in the societies, communities that we operate, what AI and agentic can do for those communities. I love that. I think for me as a technology person, I actually lean on the human side a lot. I think we have a lot of work to do to ensure that we're bringing people on the journey and giving them the opportunities and democratizing the tools and giving people the education to to want to learn because not everyone's it's not natural for everyone is one thing you realize um and it's how do I just make it more accessible how do I help that person through because I think everyone's got it in them and I think there's just this like I don't know if it's an insecurity or lack of confidence whatever it is and I think we just as leaders have an opportunity to really try and transform and I'll tell you it's not age related I've seen some of our older folks do better than the younger and some of these things where they're like you know what screw it throwing it all up in the air teach me something new but i think there's just so much around the psychology of what we need to and i think humans are actually needed more now than we were needed before because there is a lot around the human judgment around the things we're doing not in the technical sense but if you want to stay true to your consumers you got to ensure that like we are staying true to the consumers and i think we're still going to play a very key role yeah last word goes to you thank you um we're all leaders in the room like we're in this room for a reason go into your organizations, find that singular person that's tucked away, doesn't matter what function they're in, pull them up and say, can we build something together as a senior leader with them and see what that does to your judgment as it relates to AI, how you're hard coding, again, probabilistic deterministic elements of it, and just run that experiment and see what happens with that. Not just in the thing that you build, but what it does to that individual and what that does as a ripple effect for the rest of the organization. Just try the experiment. I'm constantly saying what would need to be true all the time. And not necessarily the youngest. I'm still high potential. Yeah, super high potential. We can all be picked. Yeah, we're going to Cuba. Please join me in thanking our fabulous panel, Dina Valli of Shiboni. Thank you again to Salesforce.