The Agentic Insider

TAI - Todd James
===

Speaker 2: [00:00:00] Welcome back to the Agentic Insider. I'm your host, Philip Swan. In this show, we explore cutting edge ideas and trends in AI and data, as well as hear from other industry thought leaders in the AI space. This podcast was brought to you by arid AI committed. To safe and responsible AI innovation. Let's dig in.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Welcome to this week's edition of the Agent Insider with Alistair and me is somebody who's got 30 years of experience in information security, technology, data, and business consulting. Eight of those years. He was an officer serving with the US Coast Guard where he held senior leadership roles in IT, information security and shipboard operations.

He's an active public speaker within [00:01:00] the data and AI community, and he currently serves on the CXO Advisory Board for Sierra Ventures, a major venture for capital firm, and the founder of CEO Insights. Todd James, welcome to the podcast.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, thanks Philip. I appreciate the opportunity to be here.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Awesome. So like everybody else, no different. Todd, what future are you solving for?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah, I appreciate the question. So, uh, you know, I actually founded my new company to, uh, help start and pursue that future. I, I, I view the world through a little bit of a historical lens and recognize that we're going through yet another transition, uh, transitions open up opportunity. Uh, I think it's been called the largest once in a lifetime, uh, global replatforming that we'll see.

We'll see what happens with Quantum. We may get two of those, but right now, uh, it's with ai. So, you know, uh, my, my thesis is companies are really struggling with two [00:02:00] key questions. One of which is, you know, how do I get started? What is it? Don't want anyone to know. I don't know what it is. And then the other question is, I've been doing this AI, but I haven't seen value or scale.

So the future that I see is one in which. AI capabilities becomes broadly democratized. One in which organizations are able to steward their workforces and their companies through these changes without undue friction, I still see a world where there's winners and losers, but I, uh, you know, the future that I'm solving for is to make that journey easier on companies, on individuals, on society.

So that's what I'm excited about, what I'm motivated by and why I left a. Very good job to go out and say, can we do take these skills and take these capabilities more broadly?

So an entrepreneur at least once in their life, right?

it it is, it's funny though, you, you, you go from having all this infrastructure around you to, you know, uh. You work just as much as you do in a large corporate job, but you know, you're, you're [00:03:00] thinking, you're strategizing, you're being creative, you're talking to people and you're like, I gotta go fix a printer, and how do I do that?

So, yeah.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: That's right.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: or or calling my wife, Hey, how do I write a check? I, that discussion actually happens, so, yeah.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: So, I mean, you, you're talking about those two key questions and I think, you know, how, how do you think about using AI to transform businesses so that they've got the right to innovate and compete in the future?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: It, it, it's interesting I, going back a little bit through a historical lens. There's all this hype around ai and it kind of bothers me that, uh, we put so much focus on it, even though I'm a beneficiary to some extent of the change. Um, and I'm enjoying, uh, the, the, the transformation. But you go back and, and I look.

The big events over my career and the big factors that have been used to drive changes in companies and, and I really see several things, three things. One is, is, you know, in the fifties with Duran and Dimming and the, the, the Six Sigma [00:04:00] and the lean Revolutions where we're able to look and say, how do we reorganize people in processes to be able to drive more efficient workflows that you're able, better able to serve customers and reduce friction for your workforce.

And then came along automation. And this is what we've been doing for most of our careers, right? The combination of automation and that process, organizational dynamic to be able to, to knock friction out of systems so that people can re-envision products, re-envision customer experiences and create new business models.

But until recently, um, you know, we got to a point where the. Processes could only get us so far. The upskilling could only get us so far. And deterministic aspects of automation could only get us so far. And we had these big pools of resources that were stuck with dealing with complexity in making judgment calls and decisions.

And as much as we talk about the mystery of ai, the simplicity of it [00:05:00] is, is we are now dealing with that bucket that we haven't been able to solve for. Is people making judgment calls, having to grapple with complexity because it's not an A or B decision. And what we're, we're left with is we have an opportunity to use math to convert those judgment problems into prediction, machine prediction problems.

And in doing so, we can better inform decisions. We can automate in places, we can change the way that we work and we can create better customer experiences. But if you, if you ask it kind of how I view it, it's that continuation of. Kind of the sciences and the automation that we've applied all along, we're finally able to get that last bucket that we struggled to solve for as we looked at driving improvement organizations.

And, and, and that was really around how do we help people make better judgment decisions? And quite frankly, how do we unburden them from, uh, decisions and activities they, they don't want to make? So. That that doesn't sound very ai and it doesn't sound, uh, if that's the word, it doesn't [00:06:00] sound very, uh, exciting, but it's, I think it's hugely transformative.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: I, I think it is. And also you're talking about, you know, in some ways collaboratively solving problems. Okay. Previously, you know, you're talking about, you know, mathematical models to be able to do this. So, you know, you bring people who have a PhD or two PhDs onto your staff to be able to do that, and then they become the, uh, the buffin that you work with to be able to get through this.

Who pretty. You know, we're, we're not far from being able to have, um, you know, an AI at our fingertips that's usable by anybody that's gonna have, you know, just about any PhD in anything you want whatsoever. And so at that point, you are now able to have a collaborative partner to be able to work with Okay.

And, and grow the sort of, the, the strategy that you want in the direction that you want for. To the company and, and test things out Okay. Without necessarily needing to, um, to, to, to have to go and find the right person who had those skills. If what you are looking for is sort of able to be sourced from the, from the AI side of things.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Which is wonderful. It's [00:07:00] beautiful and it's disruptive. Um, you know, I couldn't agree more with you. The, the other thing I would say too, in. And, and a lot of us are already using some of these, these systems in our, our, our day to day. Um, it's a collaborative partner. I would say it's, it remain one of several collaborative partners.

I mean, I, I have had a, you know, been going back and forth, uh, with an agent recently, just questions, asking questions, looking for input around an area that I was not that familiar with, and talked to a lot of people in that area. And there was consistency. But yesterday I had a call with a gentleman, uh, out of Toronto that hit a totally different vector.

So you, you, you need a little bit of hallucination. You need, uh, a little bit of diversity. I think as you continue to look at your decisions, we're, we're, we're not done helping each other yet.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: No, I agree. And I don't, I don't think there's, the, the, the time for, you know, the lack of existence of human to human collaboration is going away. I don't think, I don't think that's dying anytime soon. But I think suddenly you've now got an ability to, [00:08:00] to do some collaboration that wasn't possible in the past.

Okay. In ways that now, you know, uh, give you unfettered opportunity to, to, to do new things.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: At a pace and with reach, you've never been able to do it, and quite frankly. You are not bothering anyone when you start interacting with that agent at one in the morning when you can't sleep, you are, if you pick up the telephone and call that colleague of

yours.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: Exactly.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: don't have to worry about vacations or the weekends or after hours or any

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah, there's, there's no boundary spaces

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: it's interesting because we are, we're really talking about a dito, a dichotomy here. You're talking about, you know, one of your first, you said, you know, answering Philip's question, you know, what future resolving for is. A lot of this is, uh, two key questions.

First one is, how do I get started? And then suddenly we're talking about, you know, you know, intense collaboration at one in the morning with multiple PhDs to be able to do all of this. And, and the, there's some companies that are like, hang on a minute, this, this. This is accelerating so fast and I, I feel like the train has not only left the station, it's, it's out of the, out of the system in the universe if I break [00:09:00] the metaphor and I, I don't even know how to, how to even get started on this.

So there, there seems to be a very, a very large continuum as to where people are falling at this

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, you, you, you bring up an incredible, uh, point, uh, that, that I think people are missing. Uh, if, if you look at the data, I forget the source, but like you go back to 2023, 50% of companies said they were working with AI in a material way end of last year. 2024 is 70%. You're like, good. Everyone's using it.

We're done. Right. Um, no. Uh, that continuum tells a incredibly rich story that I. Think if you're sitting out there as an organization or you're sitting out at there as a leader, that is the thread you should pull. 'cause what the data is starting to say is those that are leading, those that have figured out not only how to implement an AI pilot and to start to use it, but have figured out how to instill a culture where leadership understands how data and AI can better create processes and informed decisions.

How data [00:10:00] with every. Time you touch it, you have an opportunity to, an obligation to, to protect, but also an oblig, an opportunity to enrich that asset. Those companies by various studies, and I'm doing some of my own analysis now with some of the data I have, but they are seeing, I think it was uh, I think it was Kearney and McKenzie.

They came out with some points and I can't remember exactly which one, but one was citing a two to six x. Total shareholder return for those companies that were really leading and had data and analytics as a culture embedded in versus those that were stuck in pilot stage. And you are also seeing a difference around profitability upwards in some cases of 60%.

Not in all, but if you're sitting here as an organization and it's. The velocity's increasing, uh, it, it's time to stop talking about that and it's time to start moving. 'cause there is a non-linear growth rate of value that we're seeing from those companies that are [00:11:00] implementing artificial intelligence in a way where they're able to start compounding the effect of individual decisions that are being made by AI in a way that's driving true material competitive differentiation.

So that that's the opportunity. Yeah, it's, you're sitting there with Joe's on the side of the pool. Um, it, it's not just that you're falling, it's not velocity, it's acceleration that's happening in front of you.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Yeah, so funnily enough I was going down the same path, uh, question. What I was going in was, you said in. Your advising executives who feel stuck on ai, you know what? What are the diagnostics that tell you whether the problem is data, talent, or culture, or all three are a mixture of, right. What do you look for?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah, I, here, here's my, my take and I'm, I'm probably a little bit, uh, out a pattern on this. If the CEO and the leadership team isn't interested in getting behind it, [00:12:00] then, then it's a waste of time. I have, you know, across the companies I've worked with and the companies I've, uh, advised, if you do not have the buy-in and the interest at the top, which is what sets the culture, which is what determines the allocation of capital, which is what sets the strategy, um.

Messing around with the data and stuff, yeah, you'll get some incremental gains, but you're, you're not really gonna transform the business. So I think it's all those things, but if, you know, I'm talking to someone and they're not really willing or don't think their CEO's there, it, it's, it's very hard to envision how that company is really gonna be able to transform in a way where they're using data and AI to competitively differentiate.

Point solutions are great, but you're not doing a whole lot more than you can do with process improvement and automation. It's isolated. Where you get the benefit is where you start to look across entire value [00:13:00] streams and you start to think differently around what your next offering or your next investment could look like based on these assets.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: so while you're running things at Kroger. you you clearly embedded AI into everyday grocery decisions, right? So you clearly had the support of the C-Suite to make that happen.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah, I was, you know, it was, it was interesting, uh, you know, the similar journey, uh, at, uh, fidelity Investments, kind of took that playbook when, when I went to Kroger. But early on, when I went to Kroger, and I really appreciated this, I was having a discussion with the CEO, uh, at the time, Rodney.

And he looked at me, he said, Todd, we've gotta get it from where you're pushing your way in to my leadership team and the other officers to where they're pulling you. And what he was talking about is you can't do AI coming from a technology first. You have to come from a business first. And I give a lot of credit to, to him, I give a lot of credit to his leadership team [00:14:00] for how they modeled it.

Um, you know, there was, you know, we did, we had ceremonial events like going up to the northeast to a college that is very well known for a university. It's very well known for its technology and the senior leadership team up there, and being able to walk through and spend dedicated time talking about how this can impact our business.

Visiting that university and talking to leading professors. And while that in itself doesn't change, it is representative of a commitment. That commitment got signaled across the entire organization. And, you know, I saw the same thing at Fidelity. I remember when, uh, I was in a, uh, session and we, we had, we probably had about 250, you think 200, 250 of the top leaders in the firm.

And we're sitting up on, it's been a few years in the building, the top floor where all the conference rooms are at their headquarters, and there's a. University from New England that's known for tech in there. And they're talking about AI and [00:15:00] what it means for business. And then I look in the first row sitting there as, uh, the owner of the company, the chair chairman Abigail Johnson sitting there taking notes.

She didn't have to be there, but she was sending a signal by being there. And so when I talk about the need for leadership from the top, it's that cultural signaling. Then, you know, where, where I think the biggest moment where you start to see that culture shift. I think probably one of the biggest points, at least in my, my last role where I finally felt like we had it, was you go into the board and you go on a cadence and, and, and part of it was, you know, once a year you go in and you give the state of ai.

You may get called in a few other times, but once a year you're there to talk to state of ai. Well, the last year I was there, it shifted. Um, instead of me going in and leading and talking about ai, uh, it was all of the company officers talking about their strategy, which had AI embedded in it. And I was kind of, and I took the second share, which is what I wanted.

This was success for me. You gotta give away a bit to, and [00:16:00] the beauty of that was, was it really started to signal the shift that this was being business led. The executives understood it. They were making calculated investments around it. And they could fluidly and coherently talk about their strategies with business impacts, but in a way that integrated the AI to it.

And you know, I added to the discussion at the end, said, oh, here's how we're gonna tie it all together. So it's not gonna cost you a fortune. Here's what scale looks like. But that moment when you're head of AI is not the one leading the discussion is the moment where your organization is starting to step into.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Great.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, that level of maturity where they understand how it's gonna transform the business and they're not just hunting projects.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: That's great. So, so at Kroger, I dunno if you can disclose or not, but which use case this delivered the fastest payback and why did it succeed where others stalled.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: You know, the, the beauty of Kroger is, you know, you, you go back about 10 years and they've been using, uh, machine learning on like best customer communication [00:17:00] and embedded, I mean, I walked into a, a pretty mature asset there, at least around, uh, the merchandising and the marketing aspects, which if you understand the history of Kroger's Analytics subsidiary and how that grew out of Dunhumby, the fact that they were heavy on merchandising.

Marketing makes a lot of sense. Um, we, we had tremendous value and everything I'm saying here has been written about in public. You can, I can even cite the article, so, uh, I won't step in anything that we shouldn't be talking about. But, uh, you know, pricing and promotion was, was, was a big one. But the, the ones that, that I saw, um, was, uh, we, and, and, and Drove, had a big play and was around this idea of.

How could we start to scale a analytic capability, process, engineering and scientific approach from a mathematic perspective across multiple use cases to drive real value? And we saw an acceleration there. So you want me to tell the [00:18:00] use case, tell the story? I'm happy to do that.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Yeah, that would be great. That

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah. So here's the problem.

So here's the problem we're asked to solve. Um, so, uh, when you, when you wanna do a pickup order and. Feel free to do it. It's still a very good company and I hold stock there, so feel free to, uh, uh, get on your Kroger app and make an order. But if you're gonna make a pickup order, you get on your application.

But then there is an individual that actually has to run around the store to pick all those items. And the challenge was, look, we, we, we wanted to get, we had about three hour lead time between the time you made your order and when you could pick it up. Like, that's not, that's not the customer service that we wanted.

We wanna get it back to two and continue to drive it down. So working with the operations team, working with the technology team, working with the science teams, uh, that we had, you know, we, we got talking about it. We said, look, if we can reduce the amount of time it takes that individual to run around the store and do the pick quarter.

That allow us to reduce the lead time. So it became a story of how do you, [00:19:00] for stores that are, you know, 2,500 stores separate layouts, how do you refresh all the digital images? Keep 'em up in real time. And how can you do realtime routing for an associate where you can either optimize for speed or even optimize for cost?

We, we biased towards speed, but the algos could go either way, but basically created a advanced routing algorithm that kind of optimize that trip around the store and reduced about a 10% percent reduction in. Uh, steps taken distance travel, and if you think about that across the number of times that you do it, um, across a number of stores, it had a significant impact on customer experience, on associate simplification of their work and from an economic perspective, but about.

Getting into that, we realized we had to harden nose algorithms anyway. You're doing, I think it was like 200,000 calculations a second in the stores to be able to optimize that. So we had to harden those. But we also said, this is routing. If we can route an associate around a store, we can route other things.

So we started thinking about what engineering [00:20:00] patterns, what did we learn from a process and how not just to deploy. Which was important, but how to support, uh, from an o and m operations and maintenance perspective after deployment. And where can the scientific approaches, it'd be different data, but where can the scientific approaches be reused?

And so while that first, uh, application, um. Drove great value. It, it took about 6, 9, 9 months to build and then it took us about six months, which was super aggressive to roll out across 2,500 stores. A lot of credit to store associates and, and tech groups and the scientists involved, but it was, it was, it was difficult the first time.

The second time, uh, we had a pilot up and running and we're routing trucks out of, uh, a distribution center. We said we, we can route an associate around a store. We can route a truck from a store to a distribution center. We could start to be, instead of thinking about a problem, let's think about a routing center of excellence and a routing capability.

And what that enabled us to do was [00:21:00] speed the value there. We're up in three months outing trucks on both of those. Um, they are both on track. Uh. Both those initiatives and the other projects in those areas across 18 to 12 months. When I last look, it was about a 2.6 x increase in value delivered from an EBITDA perspective.

And they're on track as long as they deliver this year to be the highest EBITDA generating, uh, uh, sciences, uh, in place.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Wow,

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: So I gives you a story of speed to value, but also, uh, the, the ability to maximize value and underlying, that's not just the use case you could have solved for, which was an impressive use case, but it was the idea across the tech teams and across the science teams that.

We can use this in other places. So what engineering patterns, speed, real time science. How can we take the learnings from science? And a lot of it was around operations process. What works in [00:22:00] working with the business and what works in the support? And that made a big difference.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: So, you know, when you're thinking about things like this, we're, we're obviously as we move into an AI where everyone is, uh, you know, or there's simply a, a very huge fear where people think their jobs are going to be made obsolete, um, and be replaced by, by millions of agents that can do their job pretty well.

And, and I think that we all understand that there's a. There's a change happening from where humans are, the operators to humans sort of becoming supervisors over, over, over a identified infrastructure that could drive this sort of thing. But how, how, you know, do, do you bring the workforce along in conversations like this?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: it's, it's interesting. I, I would say the, the cases where you had, there, we all have stories of sciences that were highly performant and never got implemented, right? Uh, people have a vote. Um, at the end of the day, it's like anything else. If you're gonna deploy something there, there, there has to be value in it for the [00:23:00] recipient. a a, a few things come to mind and I can give you two great examples, but, uh, and I'll start with the, the Kroger, or I'm sorry, the Fidelity one where I really learned the lesson well. 'cause we were all experimenting. I said, I'm gonna, I'm gonna try this, we'll see if it works, and if it does, we'll replicate it.

But we started getting, you know, you wanna reduce the distance between your tech teams and your business teams. You wanna get that intimacy and, and, and the simple way to describe it, the people that are gonna be impacted by it. We brought 'em into the project they weighed in, that resulted in a better, better solution. It resulted in a faster time to value. It resulted in less friction from the workforce. 'cause the people that are gonna be impacted by it, a group of them are in there saying, no, I'm looking at this. This is how it's gonna change. It also gives you that over the horizon view, where you can start to work with the people whose jobs are impacted to be able to determine. What they can do now that they don't have to do that task. And I think it's best summed up, I was living in Boston at the time [00:24:00] and I flew down to check on the project. It was being done out of, uh, uh, uh, fidelity had big operating center, still does down in, uh, Covington, Kentucky, which is suburb of Cincinnati, just outside.

And I went down, I, I spoke to one of the women who was, uh, on the business side and said, how's it going? And she said, Todd, you know what she said, I used to be a transaction processor. Now I own a micro bot. I make it better and it makes me better and I have more time to do what I really want, which is to create a delightful customer experience.

So she was getting rid of the mundane that kept her from being that advisory resource. She was receiving a call and she was able to help solve problems now and free her up to use her natural ability. We, we took that at Kroger too, and like, before we started projects, one of the things when I was, uh, leading data and tech over at 84 51 for the, the data scientists working in Kroger, I said, we're gonna go out and we're gonna [00:25:00] visit and we're gonna get on site and we're gonna, we're gonna see and do the jobs with the people before we do it.

So like, even I, as chief data and tech officer would go out. Do some of the work that we were about to impact, uh, one to signal. Um, but I learned a lot. Um, and we learned things like solutions we talked about would require people to have three arms or you learn things that, uh, have an appreciation for how hard and complex the work.

I also learned I'm no good at any of those jobs in the store. So if data science stuff didn't work out, there was no fallback, the stores weren't gonna hire me. But, uh, We would get the teams out there and then the leaders, the vice presidents running the initiatives for me would be out in the stores a lot.

And I also made a point early on that if I was gonna meet with a VP of data scientists, I wanted the business, the person from the business that was attached at the hip with them to also meet. I wouldn't, I wouldn't really for a fact, but I wanted to reinforce that tech and business were in the boat together. [00:26:00] And when you knew the one day I knew I had it, I went out and we were testing that routing solution in the stores at Amelia's store, uh, east of Cincinnati. And, uh, I asked a technical question and the director from the business answered it. It was about enriching the data for future potential uses. She answered it and I asked a business question and my BP of data science answered it.

I walked outta the store going, this thing there, there, this is tough. It's hard. Things will go wrong, but this is gonna work. And that was the biggest. So a long-winded way of saying you have to bring those people in. One, they have a vote on whether they accept it. Two, I think we have an ethical obligation to do it.

And three, you gotta figure out what's next for that workforce and having them as part of the loop. As you do that, you get as many insights from them as you do, as you think more broadly about business priorities.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: I couldn't agree more. And I know that you, you know, we've talked in the past, you know, you, me and Philip, you know, about, about all of this. And I think that when we think about, you [00:27:00] know, coming to the fidelity when, when you think about what a financial advisor looks like now compared to what it looked like 10 years ago and what it's gonna look like 10 years from now, um, I, I think managing the transition to a future that, that maintains that.

Dignity and that purpose and contribution that, that one wants to feel as well as revenue. I mean Yep. All service to, to the God of revenue. But I, I think that's, that's important. And as you say, you know, integrating people into this and having that engagement really makes a huge difference. But also, you know, if you involve those people, I think you can discover.

You the, and the purpose and the contribution through those conversations, rather than believing that the ivory tower has all the right answers to, to the right way to do this.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: the beauty with this, I, I call it business intimacy, reducing the distance between tech and, and we probably not for this call, we talk about the structural ways to do that, but, uh, one of the, uh, what you start to see, usually the first idea [00:28:00] comes from leadership. 'cause money comes from leadership, right?

So bright minds, we get together, you have those discussions. But after that first project, and as you start to look at that project going. The direction of that project. And then the follow on projects is usually a product of organic discussions. Between business people who now understand technology better and technology and data scientists who now understand the business better, that's when your best ideas start coming up at the ground level and they start working faster and they start being able to drive more value, but also do it in a way that is more consistent with not being told to, but getting the opportunity to.

And, uh, it's, it's not something, yeah. Not something that you can make happen. It's something that you can set the conditions so that it does happen.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: Mm-hmm.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Yeah. Yeah. So retail margins are res, razor thin, right. And you know,

we all know that.

Yeah, exactly. Uh, yeah. Aon, we've got financial [00:29:00] services that you've worked in. They're highly regulated. You know, you've led AI in both. Right. Which industry constraint was tougher to innovate around from your perspective?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, it was different. Uh, it was different. Um, the how is different? I wouldn't say the difficulty levels are different. The, the how is different, uh, when you're operating in a private company with large margins. You have flexibility and you, you think in terms of how can you change and disrupt and redefine an industry and you can approach it with a appetite for failure and patience

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Right.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: and a grocery retail.

And, and this is probably, you know, one of, I think one of my, it was, it was before the first. Christmas that I was at Kroger, I was having a discussion with Rodney and I'd been there a few months and he asked, how's it going and what's different? He asked a very similar [00:30:00] question, and I must have been tired 'cause outta my mouth came.

Uh, and this speaks to Rodney's Grace, uh, uh, outta my mouth came uh. Basically describe what I said about, you know, in financial services, you, you can be really strategic and you can be patient, you can think about how disrupting industries over time. Um, but I said here, you have to hide your strategy and your tactics.

And it took me a little bit longer to, to figure that we had a really good discussion. I, I, I, I stayed there after that comment, but, and it wasn't out of frustration that I made the comment, but when you have. Public company pressures and you have tight margins. It doesn't take a lot to get a company out of alignment with its financial performance.

So there has to be a very rigorous expectation. So for me, the transition was, um, less around the regulation. I mean, the regulation and financial was tough, but you gotta be really careful when you're dealing with food and people [00:31:00] too. So, and, and medicines and healthcare aspects of the pharmacy. So I saw those pretty consistent, but it was around.

How do you take that strategy and back it through tactics and into operations for really rigorous edge execution, and it doesn't do you a whole lot of good to share that strategy broadly. You share it with a few people that understand it so they know where you're going. But on a day-to-day basis, it's how do you drive to get the result in time and how do you work as quickly as possible?

Which by the way, when you can be super strategic, it breeds one cry kind of creativity. You have lots of constraints. It unleashes another kind of creativity. So I, I, I appreciated both environments. and they're both, they're both hard.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: and then when you're looking at manufacturing, which is big, where you live in Ohio, right? So they're. It's highly regulated financial services, highly regulated insurance, highly regulated, you know, and [00:32:00] we're in a world today where countries, states, here in the US are there, are sprinting toward drafting AI rules.

In your opinion, what single policy change would help Mo would help enterprises innovate safely without stifling progress.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah, and I, I, I, I had the numbers on me. I, I, uh, they flushed outta my head. But you look at the, precipitous rise in state level regulations. Um, and, and, and, and I'll say something. I was talking to a, a colleague, uh, in, in Europe the other day, and I say it used to be we, we kind of laughed at the regulatory patchwork of, uh, the eu and to which he responded, you're not laughing now.

And, uh, you know, there's, there, there, there, the, the, the states And what they're doing I think is with good intent. Um. But it's very hard to operate without some overarching level of regulatory framework [00:33:00] across states. And I, I, I, I do, I am optimistic. You, you've seen, you've seen positive development in, uh, some of the agencies.

You haven't seen it from a, like a legislative perspective across the board, you are seeing more at a federal level around protection of children. You are seeing more around fraud. You are seeing more around DeepFakes. That is fantastic. But if they could harmonize the patchwork framework of the states, you would reduce costs, you would increase the competitiveness, and you would make it easier as you learned which regulations were working and not.

To be able to turn that dial accordingly as a, a, a governing body than when you have every state do it and it, and they can't possibly talk to each other. That's not a ding on the states, but they're, it's not like there are a bunch of experiments that are being coordinated. It's just a bunch of experiments.

Cool.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: Absolutely.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: I think we saw yesterday once, uh, Alistair, um, meta [00:34:00] was saying they're not gonna adhere to the e, the EU AI Act, I Sure. But I mean, it's, there's a, there's, there's enough of these shots across the bowels that are being portrayed right now to see if they can encourage, um, uh, regulation to be diminished. Okay. Uh, or to be less restrictive. Um, I think with the a with the eu it's a little bit more difficult to, to, to make those shots 'cause the EU is resistant to those.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: I think the system, as much as we complain about it in the United States from time to time, is a more flexible system for change.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: It is absolutely. I mean, you know, there was enough enough regulations put forward over, um, over, uh, you know, Gavin Newsom's desk in California that were rejected, um, last year. And then, uh, you know, Texas just just approved a a, a whole host of fantastic. You know, regulations closer to the EU AI Act. So, you know, it, you, you, you people can be surprised as to where this regulation can be coming from when you

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: You, you raise a good point. There's no political thesis here, right? It's, it, it's really all over the map. [00:35:00] Which I think is one of the reasons why like, when something gets so messy and so visible, it, it, it, in this system, I actually think it pushes it to a resolution. I think we're more likely to see action outta the federal government because it's getting so bad.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: Well, I, I think you also end up with a thing where it says, well, no, no, no. I, when I say transparency, I mean this. No, no, no. I, I, no, I mean this, no transparency is really this. And now you have, you know, you know, it is the same thing. If you put 10 architects in a room, you get 23 different opinions. Um, it's the same sort of thing here where you end up with a, you know, with a series of, a series of different opinions on, what do I mean by transparency?

And now we're into philosophical discussions on ethics, which never really go very well from a regulation perspective.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah. Yeah. You nailed it. Yeah. I have nothing to add to that. You should have just answered the question for me.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: There you go. Well, well how about going back to your Coast Guard days, right? You, you've, you've, you learned mission critical risk management. You had to as part of what you were doing there, right? And, you know, [00:36:00] and so how does that early experience in your career influence your stance on AI safety and velocity?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: I, uh, first of all, through my late teens and through my twenties, I was in the service. It was very formidable. Um, you know, I, I look back and I've had the privilege of working for. Incredible companies, ethical companies, um, good companies. Um, coast Guard stood out, uh, as just a absolutely fantastic place to, to, to contribute, but also to kinda sharpen your various skills for the rest of of your life.

I, I think from that, uh, the big, there's probably two vectors I'll take here. One. The Coast Guard, you know, unofficial model mottos, basically so that others may live. And underlying that is the idea of [00:37:00] the importance and the agency of others beyond yourself. So Yeah.

I think that framework is something that is still instilled in how I view the world and how I think about the world. Um, the other thing that really came out of the Coast Guard, and then I'll tie all this back to AI was, uh, when things got really, really bad and decisions got really, really critical, you got that half second of panic without every letting anyone see it, and you suppressed it. And then you started doing one of two things or three things.

You would, you would endeavor to make good decisions, you would execute on your rule set, the skillset that you've acquired. And you would, uh, utilize your network, uh, whatever you had on the ship or whatever you had through the radios, you could get back, but you put the three of those into action as quickly as possible.

Uh, that seems to have worked through about just about every decision [00:38:00] I've made and when I made good decisions. There's a combination of the, a thoughtful focus on am I making the right judgment, fully employing the skills, and then using a network to be able to kind of test and make sure that. Rarely is your idea the best idea without contribution.

So, you know, that idea of, you know, kind of bringing those forward as you make decisions, but also the, the idea of, um, you've gotta be concerned about the, the greater agency, uh, beyond you have, have impacted. I, I think, uh, there's been a few things that have, uh, that, that, that's played out on me in several ways.

One, you take very serious the, the implications that. Your science, your solutions, your products have on individuals, and that's always a question you ask as you go through, right? Are, are, are we meeting the expectations and are those expectations consistent with. The experience, the impact, the dignity that we want, our customers, our clients, [00:39:00] our associates to feel, which goes back to too, if you're not thinking about that, you don't put your teams out in the stores to work with the people being

impacted. so so that plays out. Um, the other thing that I've kept is you can get in popular culture, dragged left and right by different ideologies. Someone's drowning. You don't care what they look like, where they came from, who they are, what language they speak. You get 'em outta the water. Um, I think with ai there's been some big mistakes where people have tried to put different differentiations, every human.

Are you treating every human fairly? If you're gonna use their data, is there a virtuous value exchange back to them? And is it consistent irrespective of who that human being is? And that's been something that I've stuck with through the end, right? At the end of the day, you're deploying some of this solution.

I, I see a victim in the water and all I care about is getting them out. I see a person consuming this, don't care where they're [00:40:00] from, what they look like, who they're at. To treat them fairly with dignity in a manner that's consistent across the board. Now, there are deviations, like if you're doing healthcare stuff, they're, they're like, I don't, I don't believe I get uterine cancer.

I'm pretty sure I can't. So you got, I mean, there's times where you do that, but that has been the starting point in my decisions.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: That's great. So we're gonna start winding down here. We're gonna get into some quick fire

closing questions. That's right. You worn? You have worn us out, man. That's, uh, fantastic. Well, I would, there, there is actually one thing I would like to ask before we get into our rapid closing. So, you know, nobody can predict the future and especially in ai, we sure as hell can't, um, predict the future.

As you look beyond time, whatever period it is, whether it's three months or three years, what do you see that's currently over hyped in terms of AI trends that will quickly be, that will quietly become indispensable to us? And which features do you think will fade away?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, I, I think [00:41:00] right now I don't think we've fully gotten our arms around how real the mobility solutions are.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Mm-hmm. Did you dig more into that? What you mean?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Yeah. Yeah, it, it's, it's interesting, uh, back to that meeting where, you know, the Kroger leadership team, we all went up to that university. We had an opportunity to talk to one of the professors who was leading in mobility, you know, the robot dogs. And he explained to us how different and how much more complicated that problem was than large language models and some of the other advances. And we should

expect slower. It is more difficult, it is more complex. I'm surprised by how quickly that is moving both through industry. I think some of the conflicts we have going on, starting with, uh, Azerbaijan and Armenia, looking at what's happening in Ukraine and Russia and looking at what's happened in, uh, the Middle East, the.

Change in the, the, the, the [00:42:00] speed of change in warfare has also kind of changed some of the, the dynamics around mobility that'll bleed into other live areas of our lives. So that mobility, I think we're gonna see a lot more solutions, um, that are coming out, uh, in that space. Much more than if you would've asked me a year ago, I'd

say now it's still a

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Yeah, we, we are seeing it too, right? Because, you know, you, you've got the, you know, right now, you know, when you're looking at, um, open ai, you know the current models that are available, likes of 4 0 0 3, 0 3 Pro, you know, GPT five. Could release at any time, right? Anytime between now and Christmas, right? It could do as early as tomorrow, right?

And it could, you know what's happening there? They're already working on six, seven, and eight right? In various levels and as all focused on that mobility. This, you know, spatial awareness and, uh, uh, everything that's concerned around robotics. So you're absolutely right, and that's the direction that things are going.

So. [00:43:00] You know, safety is going to be even more, um, issues there. It goes back to Asimov if you're,

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: exactly.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: IRobot foundation, et cetera. It's, uh, Yeah.

Laws,

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: 10 years ago if we were talking on a podcast about this, someone say, these guys are nut jobs. Why you listen guys, science fiction, who are these weirdos? I mean. But it's real now.

riverside_alistair_lowe-norris_raw-video-cfr_the_agentic insider_0025: It, it's, and what's what's funny is there's a, that we were talking on, that we were talking on a podcast the other day with, uh, with Robert Scobel. And I, I think that there's a, there's a transition state between, okay, we will take a chat GPT Foundation like model and stick it inside a, a, an autonomous robot that then can do the ironing, do the cooking, do the cleaning, you know, put the laundry on and everything else, which, which isn't that far away.

But there's a step between that where you take the autonomous robot and you connect it to an army of humans, one of whom is a babysitter. One of whom knows how to fold laundry, one of whom does this. And so you then you, you have a human controlled drone, to your point about [00:44:00] war warfare bleeding over, you have a human controlled drone that is then using the intellectual capability of somebody who is trained, certified, whatever, in how to be a babysitter, how to do this.

And suddenly then you've got

different

personalities.

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: care. The implications. We can stay on our houses with three flights of stairs now into our however long. I mean, it's, it's gonna have a tremendous impact. I, I think the other one's probably a little less exciting. Um, uh, we have, uh, AI driven digital customer experiences. We're gonna see that converge with

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Yes, that's going to, that's going to happen, Todd. Absolutely. I mean, hyper-personalization, especially in the world of ENT systems, is going to happen. That's just a matter of time. Uh, we're already seeing it. Um, so it is exciting times. All right. What book would you recommend to people to read if they're want to, you know, to that's influenced your thinking on AI and AI leadership?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: [00:45:00] God. Uh, there's one by, uh, lemme pull it up. I'm, I'm bad with titles, but I know the book. It was written by, uh, Microsoft, CTO. It's a really good book and I'll tell you why. Reprogramming the American Dream we're dated on some of the ai, but how he's talking about the social implications of being able to create opportunity beyond the hubs into, uh, areas, uh, uh, like Appalachia and other areas.

I think, uh, shows how we can advance science while also, or advance our science, while also thinking about how do we bring the whole of society along with us? Not in a way that's about handouts, not in a way that's about UBI, but in a way that's about work, meaningful work with dignity, and being able to address those areas.

I think it has to be done intentionally, but I think it'd be done profitably in a way that, uh. Uh, sir. So that's a, that's a great book both around. It's probably a few years old on the ai, but it's really good around a vision for how we might wanna think about the future.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: That's great. Next [00:46:00] question, what contrarian opinion about AI do you hold that most peers would disagree with?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, I've been saying for years, and it made me unpopular. Uh, I, I believe that this chief data officer and chief analytics officer role. Is going away. I believe it's a transformational role. I think as I call it, the exciting stuff. And the analytics will become embedded in core of business just as operations management and some level of finance acumen is.

And I think the data component will get pushed back down into tech, and it'll be about how do I keep the, the, the pipes together, people in those roles, especially with what they pay. And I was in those roles just a, a few months ago. Um, don't like to hear that, but, uh, I, I think it's. If you look at what you do in that role, that that makes you successful, it's all transformation.

It's like 70% organizational dynamics, 30%, and it's about moving from A to B. It's a transformational role.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: And last question to you, Todd, you'll be so generous with your time. [00:47:00] What do you like to do for fun?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Anything with my family, I'm pretty easy there. So, uh, uh, when I have, uh, family together in the house, my heart is happy. Um, and so that, that's the most important thing to me.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: And how do people reach you with Aurora Insights?

riverside_todd_james_raw-video-cfr_the_agentic insider_0024: Uh, they can look me up at, uh, ww dot aurora insights llc.com. Uh, easier way to connect with me is, uh, you can also hit me on LinkedIn, but, uh, thank you for, for asking, but we're out there and uh, would be happy to help.

riverside_phillip_swan_raw-video-cfr_the_agentic insider_0026: Todd James, thank you so much for your time and being part of our podcast this week, and join us next week for yet another great episode. Thank you.

Speaker: That's a wrap on this week's episode of the Ag Agentic Insider. Thank you for listening. For show notes and more, please visit AgTech Insider Show. To [00:48:00] learn more about arids visit arids.ai. We'll see you next time.