AI Only Helps When You Know Which Hill to Climb

AI is everywhere, yet most teams still feel stuck between exciting demos and messy reality. Jason Rome sits down with Jon Webster, Chief Operating Officer at CPP Investments, to pressure test what has truly changed since their last conversation and what has not. We talk candidly about why generative AI adoption is starting to look like every other enterprise technology rollout: uneven, political, constrained by governance, and full of “we bought the licenses but we do not have the use cases” moments.

We dig into the economics behind the hype: token pricing, subsidised plans, and why clear price signals matter if you want real ROI from enterprise AI. When AI feels cheap, sprawl is rational. When prices rise, leaders have to prioritise, measure outcomes, and decide where AI belongs in the operating model. From there we explore the human risks and skills that get exposed fast, including cognitive load, over-reliance, and the growing divide between people with strong mental models and those who skip straight to prompting.

The conversation also goes deep on practical ways to work better: owning your outline before you generate, using AI as an adversary to challenge your thinking, and borrowing frameworks from great strategy writing to choose the right “hills to climb.” We close with predictions on where the next nine months may go, from more disciplined optimisation to shifts in SaaS, systems of record versus systems of action, and the leadership balance between IQ gains and EQ and empathy.

If you found this useful, subscribe so you do not miss the follow-up, share it with a teammate who is wrestling with AI adoption, and leave a review with the most valuable AI habit you have learned so far.

Jason Rome on LinkedIn: /jason-rom-275b2014

Jon Webster on LinkedIn: /in/jrwebster

Method Website: method.com

CPP Investments Website: cppinvestments.com

Transcript

Welcome And Why AI Feels Different

Jason Rome 0:00

Welcome back to another episode of Build What's Next. Um, Jason Rome, your host. Really excited to be joined by John Webster, who is the chief operating officer at CPP Investments. John is not just one of the smartest, but one of the most thoughtful people I've talked to about technology and especially when it comes to AI. And he's done a lot of thinking and writing in the space, as well as being an early adopter himself, as well as at the company. He and I spoke uh about nine or 10 months ago, and obviously a lot has changed, but also has not changed. And so he and I are going to go through today and discuss what's evolved, what's helping organizations adopt, what's holding them back, where is the real value, where do we think cost is going, and we're gonna project to the next nine to ten months. And then we'll come back and we'll talk about uh where we're right and where we were wrong, uh, just like we are today. So excited to uh to talk through things today.

Josh Lucas 0:49

You are listening to Methods Build What's Next, digital product perspectives, presented by Global Logic. At Method, we aim to bridge the gap between technology and humanity for a more seamless digital future. Join us as we uncover insights, best practices, and cutting-edge technologies with top industry leaders that can help you and your organization craft better digital products and experiences.

Jason Rome 1:16

Welcome back to another edition of Build What's Next. I'm Jason Rome, your host. I'm really excited to have John Webster back. Uh John and I spoke last August, and we were just connecting a little bit before the podcast, you know, a couple months before Opus 4.5 came out, which was a bit of a turning point for a lot of companies and what they were thinking about. Um, and so gonna dive into what's changed, what hasn't, where are we in both the hype and the anti-hype cycles, you know, maybe a lightning round on some fact or fiction and see where things take us. So uh John, uh, welcome back. Would love for you to just give a little intro for those who haven't uh heard you before, and then we'll jump right in.

SPEAKER_02 1:53

Great, Jason. Fantastic to be back. So yeah, John Webster, I'm the Chief Operating Officer of CPP Investments. Um, and for those that don't know CPP Investments, we uh support the retirement security of future generations of Canadians and 22 million Canadians today through investing uh in the public and private markets. We are a global investor, we run a diversified, resilient portfolio. I think we have investments in over 50 countries uh with probably over 300 partners. So uh a complex set of things to look

Hype Cooling And Adoption Reality Check

SPEAKER_02 2:23

after.

Jason Rome 2:23

Yeah. Well, well, let's dive in, I guess, since since last September, you know, when it comes to AI, the adoption of AI, just just what you know you've seen uh, you know, within your company, what you've explored personally, what you're seeing in in other companies or hearing from colleagues. Maybe what's changed and and what hasn't, in your view, over the the last nine months?

SPEAKER_02 2:42

Yeah, I love it. It's a great question. I think um, you know, to your point, August last year, we were probably just you know getting ready for things like 4.5 to arrive. Uh I guess harnesses were sort of probably around on the scene in some form, but sort of hadn't taken off. If you if you'd asked me nine months ago, uh I would have been in the uh we're just about to see like the world of work completely change. Nine months later, I think in general, that's not happened for the most part. Um I was reflecting on my sort of just everyday consumer experience. I was taken by one of the quotes from the, I think it was the COO of Uber over the last couple of weeks, and I think one of the statements he made was we're not necessarily seeing, if you like, the the token burning coming through in the product for our customers and the connection to customer outcomes. I think that's true for most organizations now. I think probably a year ago I would have said this is a normal technology, it's gonna diffuse in the normal way, but there's no doubt I was like deeply, deeply excited by it. I am still excited by it, but it's definitely a normal technology, in my view, and it's gonna continue to diffuse in the normal way.

Jason Rome 3:48

You know, except for the companies that are token maxing, obviously. And and I think that'll be a uh a very brief moment uh as uh as the prices catch up. Though, you know, thinking about price is interesting because you know, the will prices decrease uh compared to how much subsidization decreases is an interesting one to think about. And maybe we'll go there later. But you know, even if you look at the news about what Anthropic and OpenAI are doing, launching their own consulting companies with venture back to be able to drive enterprise adoption, to your point, that signals that this is not a overnight success story that you just kind of pick it up off the shelf. And I think, you know, I'm seeing a lot of still the, you know, we have the licenses, but not the use cases yet. And or we don't have the governance, or we can't keep up with the change, a lot of analysis paralysis. And then companies as well, you know, I think the moment that I see a lot of my clients in is, you know, moving from this experimentation to expectation and figuring out, like, you know, experimentation for people is great. It's very bottom up, it drives excitement. You know, you can you can find that magic moment, you know, the first time you, oh wow, you know, this this changes things. But I think now that's led to, you know, sprawl confusion for a lot of organizations, everyone having too many personal agents trying to do too many things that are collaborative. So organizations are trying to figure out, hey, what do we need to do top down versus what's done within a department versus where are we still empowering people to do it on their own? But but to your point, you know, where it's being used, where it's not being used, I think as well to your point around the consumer side, you know, early on, you know, is a very, it's been a very delicate balancing act, especially for SaaS companies looking at the SaaS pocalypse of, you know, AI washing on one hand versus vaporware on the other hand that nobody potentially uses and getting to those use cases. So I think we're getting a lot faster to the, you know, the the balanced approach of, okay, like what's real, what's not, and where is the hard work to be done? Um I I guess, and and maybe that leads to the next question. You know, if if we look past some of the hype, you know, and and and we look back again, what's what's working now potentially even better than you thought it would? And and where is there still room in in your view when it comes to adoption?

SPEAKER_02 5:53

Yeah,

Pricing Signals And Scarce Use Cases

SPEAKER_02 5:54

I mean, uh I I think we we are a general most I think most organizations, ourselves included, I think are just about coming out of the sort of era of individual productivity. We talked about that last time. Um, if you'd asked again, if you asked me nine months ago, I would have said we would come out quicker. I do reflect on why that hasn't happened. Um I think there's a few things, and you touched on a few of them, which I think are worth sort of bringing up. I think actually price is interesting, right? I think this technology has not been priced, you know, clearly in the last basically, you know, since it started to come onto the marketplace. And we're starting to get some signals about the price of it. And I think when when you get prices, you can value things, right? And I think up until now it's been genuinely, I mean, if you if in general, if you were a C CTO or CIO or COO, uh, I think it's been pretty reasonable actually to put this hand this technology into the hands of everybody and get a sense of um, you know, what is possible bottom up, where are people going to find utility with it. You know, I talked last time about ultimately for a long period of time, this or for the first couple of years, this was an RD technology. I popped out of RD labs. It wasn't really a product. I think things like, you know, code, cowork, codex have now started to really give you a sense of where the product is heading. But that's a relatively new innovation in the last sort of you know nine months. So I think up until now with a sort of price signal, which was this is actually pretty cheap technology to put into the hands of it. It's been pretty, it's been very sensible, I think, actually to go bottom up. Look, Canada, I don't know where the price signals are going. Like intuitively, you'd say they're going up, and you are gonna have to start to attach, you know, much clearer value if you're gonna spend, frankly, a lot of money on processing power. Yeah, having said that, you know, OpenAI in the last you know week or so started to talk about price cuts, not price raises. But I think that you know the price is likely to go up, and therefore it's gonna you're gonna start to see actually what are the scarce valuable uses for this in my organization. I think that's ultimately really gonna be quite helpful because it's gonna start to attach you to things like where genuinely is the value. Um and you're gonna keep having to come back to. I think with new technology, you've got to come back to in the long run, technologies don't create some sustained competitive advantage. There's something you do that they amplify. Yeah, that might be you're great at design, you're great at being responsive to customer need, whatever it happens to be, there's something that you are you are already great at, and technologies amplify that. Equally, they can destroy it if your competitors do something that you can't or won't do. But I think it's causing us all to kind of come back to the basic questions again, like where is this useful for our primary business? How will it help us make better investment decisions? There's no doubt that the individual productivity is there. And actually, I don't think we should get too tired on the fact that, you know, if 10 everyone's 10% more productive, are you seeing 10% more product productivity across the organization? I still don't stress about the fact about whether you do or you don't. I think it's generally good if people can be more productive and they're more engaged in their work, you know, and they can churn through sort of toil more quickly. But we are getting to the point where, particularly as you know, the price signals look like they're starting to go up, you're gonna have to come back to like, you know, how can this help you do your primary business? And for us, that's how does that help us make better investment decisions? What are those investment decisions and what degree of value do put do you put on the price of the computational power you're putting into those investment decisions? So I actually think the sort of the current trend is is helpful to get us back to technology's ought to do something useful for you.

Jason Rome 9:06

Yeah. A lot of gold to unpack there. I I think just to respond on the price one, I'll probably put a pin in that and we can come back to it in nine months after uh Anthropic and OpenAI, both IPO, because I I think it it, you know, I I saw an article and and will have to fact check me on this, but you know, someone, you know, took the different, you know, pro and business and enterprise plans and and token maxed it and then kind of reverse engineered that to the actual cost of okay, I'm paying $20 a month, but I can spend $400, I mean, I can spend $400 worth of tokens. You know, I'm paying $500 for an enterprise plan. I can spend $10,000 worth of tokens. And so I think there's this question of, you know, there's a, you know, there's the model arms race, but then there's a market share race of, you know, just a land and expand, you know, can we claim market share? You know, which we saw similar things with a lot of the CRM platforms back in the day of, you know, undercutting their prices and then running into margin issues later. But then there's this, you know, what numbers are you trying to optimize for your IPO? And then what's it going to look like when you have shareholders that you have to report quarterly now? How do you maintain that long-term versus short-term vision? And then where do the expansion strategies go from there? And we've seen, you know, there's a lot of interesting things, you know, obviously we've seen the companies toy with ads and pull that back. Uh, and so I think, and then there's still a lot of questions of, you know, these model companies, are they infrastructure? Are they a utility? Are they telecom or are they how much of the app layer are they going to be able to own and the consumer loyalty? And, you know, are people gonna just be able to shop multiple models and bring those APIs in? And so there's such a confluence, I think, right now, as as we sit here in in June of 2026, of where pricing will go. You know, you look at the data, you know, data center issues and construction and the backlash around that, and uh, you know, where the US is with their build out compared to the EU with their build out of data centers. So, yeah, I I think put a bit in that one and let's come back and revisit where that has gone because there's a lot of arrows pointing in a lot of different directions of of where that could go for these companies. The uh I want to come back to the, you know, the keeping people engaged in the productivity.

Critical Thinking Strategy And Cognitive Load

Jason Rome 11:04

You know, we've seen the term AI brain fry kind of come out. We've seen these tools potentially create greater, greater cognitive load. We've seen the the early research studies, I actually haven't seen as many recently, about the impact on critical thinking ability and and people over-relying on these. What's your take there on, you know, where are these tools amplifying and where are they potentially, I don't want to say making us lazy, but where are they decreasing quality? Um, and and what have you seen there? What have you put in place there?

SPEAKER_02 11:33

Yeah, I mean, I think there's I mean, there's a there's a few like there's a few things. I was in a conversation the other day and um it was about ta ta talent and how we should think about you know skills of the future and critical thinking keeps coming up. And when we talk about critical thinking, we we we talk about as if it was like a strategic choice, like there's an option not to build critical thinking into your talent development process and in terms of your hiring processes. So I must feel I get a little bit frustrated with like you know, you know, we need to really double down on critical thinking. It's like every organization like needs to double down on critical thinking. They always needed to double down on critical thinking. Like, you know, the opposite of not doing critical thinking is just being stupid on its face. I think that's one where actually we should probably just be honest with ourselves and say, have we done a good enough job of building the critical thinking skills of you know ourselves and our colleagues in general in most organizations? And I think the answer is probably not. So I do think, you know, I think critical thinking has to be on the agenda, but not because actually AI is in some sense the issue. I think AI is highlighting the issue that we haven't always done that as well as we should. I do think there is I do, I mean, I see a legitimate divide between I was talking with someone about this the other day, right? The the the whole loop piece is coming around now, right? You've got Boris Cheney talking about loops, is this is just about what loops you build. You see people posting all the time, I think uh on things like you know, hill climbing, everything's now just a hill climb, you can sort of just point, point the machine, add a hill, have a hill climb. And I don't want to sound all trite about it, but actually that you know the most important thing in strategy is to like find what hill to climb, yeah, know when to get off the hill that you're on, you know, abandon the set of hills and go somewhere else. Um, and I think it's just bringing back to the fore the need for strategic thinking. And what I find is there is a real divide between people that have a set of frameworks, mental models, and how they use the machines constructively, and frankly, people that have probably not equipped themselves as well with a set of mental models and ways of thinking and how they use the machines constructively. Again, I think you're into an amplification issue here. I mean, I think AI is absolutely amplifying that divide, but the divide was there, and you know, we need to sort of make sure we equip people with the skills to close that divide from the outset. So, what's my best advice to people nowadays? Read a book. Read a book, you know, read a book on strategy, go go talk to somebody who's got a track record in strategic thinking or a track record in doing something that you're in, a genuine track record in doing something you're interested in. Find out about how they think, find out about how the way they approach problems. I think that the the primacy on being able to really structure your thinking and then point the weaponry in the right way at the problems you need to solve. So I do think there is a, you know, there is a growing divide on these things, and I think there is a real danger of, in some sense, the sort of thing we've seen with social media more on the sort of consumer level. Yeah, that it's sort of it's taking attention in ways that are are unhelpful. I think there's a there's a further risk that it actually not only does it take attention in ways that are unhelpful, um, it can actually cause you to act in ways that are unhelpful. I mean, you just think about something, you know, I have a calendar thing, calendar skill. Yeah, it does me a sort of daily, weekly, monthly perspective on my calendar. You run that five times on the same week, you'll get five different slight orientations in terms of the thing that you you should be focusing on. I thought it was probably great for the first week when I had it, and then I realized I run it five times, I get five different orientations. Actually, I should probably just look at my calendar myself, make up my own mind and where I should be spending my time and attention over the course of the next week. I don't have any answers actually, Jason, about what we do about these things. I think, you know, first of all is to call the sort of concerns into consciousness and hold them as sort of primary risks and issues organizationally to think about and deal with. What that means more generally in sort of mass consumer land, I think generally more difficult to tell. And I think there's a lot to unfold in this, and there's a lot of work to do to make sure the models are pointing us in good directions, not sort of, you know, good hills to climb, not bad hills to climb.

Jason Rome 15:14

To your point there about uh strategy books, I think actually your initial opening thought there was a uh, I think tying back to Ramel, I don't know if it's as the crux book or good strategy, bad strategy, but hit his point that, you know, a strategic pillar is only viable if the opposite is a viable strategy as well. And so to your point, you know, critical thinking is not a strategy because not being critical thinkers is not a strategy. So you can't have your strategy be we want more critical thinking, which I uh, you know, I I I've always loved his writing and and his uh very directness. And so if people are looking for, you know, two good strategy books, I I know you're a fan of playing to win as well, but you know, the good strategy, bad strategy, and the crux would would definitely be too, where he articulates that well. It's interesting. And I was looking up too, because I I think uh, you know, you mentioned social media there. And it's interesting because you you have seen some of these leaders, you know, putting out there that we would love to slow down with AI. And and even they can feel the speed. But I think the balance of the conscious, you know, you know, it'd be good to slow down and the game theory of if we slow down, someone else will take our market share. Uh and and to your point, you know, we're still wrestling with the last technology of social media and the impact it had on our brains and on our children. I I know the UK just uh they've got the ban going into place that I think just came out. And so they're taking action there. Obviously, here in the US, you know, there's been a lot of litigation with with Meta and YouTube and the impact there. So, you know, to your point, I think that's you know, well, everyone wanted to slow down until the the US actually banned the export of uh mythos this week. But I'm sure by the time this podcast uh ships, that that will have been solved. Uh but I I think a lot of interesting points around this pointing at the hill.

Practical Team Habits For Better Work

Jason Rome 16:48

And and we've actually seen this in the um product engineering space. And you know, we have a lot of clients asking us about how to how to really rethink their development lifecycle right now. And and one of the things that I talk about to your point around critical thinking is because of some of the agile frameworks and thought leadership out there of the last couple of years, you know, we took what was supposed to be this product management role, which was really about judgment. It was really about trade-off and it was really about decision making. And we split that one role into five or six. You know, we had a PM, we had a PO, we had a Scrum Master, we had a strategy team. Sometimes we had a data analyst who actually looked at the feedback. We had a researcher who did the research. And so we created these really long, like very heavy handoffs of some of these roles. And, you know, we took people that, you know, were supposed to be machines themselves and running a very complex process that involves a lot of critical thinking. We tried to turn that into a factory where the people were gears instead. And I think that's one of the hard things to your bride about critical thinking is I talk to a lot of organizations and when they say, Hey, I want my product team to have a more commercial mindset, I say, okay, is that a is that a permission issue? Is it a time issue, or is it an upskilling and a talent issue? And the answer is yes, that that they usually say to that. But you know, we haven't given people the time to do this work. We haven't given them permission. And then a lot of times we haven't given the framework to succeed in because we've given them jobs that demand demanded just blind task output. And I think when we come into companies, especially looking at how AI impacts really complex end-to-end workflows and we're trying to reverse engineer those, you know, there's a lot of people at different points in the workflow that has no, they have no visibility of what upstream and downstream looks like and how things actually happen. And, you know, AI collapses all of that. And that's that's one of the struggles I see to the point about critical thinking. You know, and then even internally, you know, I know I've gotten, you know, I've already started getting a little bit allergic to clawed created slide content. I've become really allergic to AI generated copy. Um, and and I've started kind of publishing some guidance. And one of my things I point teams at is own your outline. Like you can bring in AI research, but until you've written two pages of like what you're trying to do, the outcomes, your understanding of the pain point, what you're trying to accomplish, don't let AI touch it. Because if you introduce AI before you have shaped your thinking, and and this is one of the things you said to me last time that that really has helped me here, which is, you know, thinking is the act of like codifying your own knowledge. I mean, writing is. And I think that's such a key thing. And AI has made it so easy to skip step one. But then everything after that is biased and there's downstream impacts of that. And so that that's one of the most practical, you know, I'd say, you know, three pieces of practical advice here. You know, one is always own your outline for clarity of thought. Two is, you know, don't expect anyone to spend more time reading something than it took you to write it. Uh, you know, that that's that's what I tell people is like, don't, you know, I will spend as much time reading it as you spent writing it. Uh, and so if it's one prompt and an email to me, then I will scan the title and then I'll plug it into my own AI and ask it what it says, um, which I'm probably just reverse engineering your prompt at that point. And then the third one is, you know, when you think about using AI, I always say one third advise you, one third assist you, and one third be an adversary to you. And I think that's probably one of the ones that people don't do enough, which is these adversarial AI agents kind of fact-checking them, challenging them along the ways. Have you seen any other kind of practical guidance for people as they've wrestled with these different ways of using AI that you'd offer to folks? Great question.

SPEAKER_02 20:10

You know, I think my view at the moment is we're we're still very much, well, you know, we're like in the foothills of practices forming as well. You know, technology and practices co-evolve. I think people, certainly lots of people I work with would probably just find me quite boring on this point because I'm sort of like, I don't know what the best practices are. I still don't know what the best practices are. There are some things I'm coming to dislike, which is as a sort of like gut instinct signal. So you're you bad to your point. You can tell AI written copy. AI written copy in itself, I don't mind. But is there like an original idea somewhere in this, you know, AI written copy? Um and so, you know, once you get through the sort of the M-dashes and the load bearing and sort of all the other stock phrases, so can you find something that is genuine? You go, okay, I hadn't thought about the world in that way before. Um, which I think comes back to like own your own outline, right? So I think there is something about just, you know, are you trying to put down some original thought? Now, you know, I also like got to be open-minded on original thought. I find most of the great thoughts come from a good structured framework and a problem, and you put the two together, and you can you can get like quite a degree of originality out of thinking through new problems in a structured form. Um, but I do think like the own the own your own outline is really good. I spend a lot of time writing with AI. I find that very productive. But you know, my my mode of writing with AI is to take a problem I don't know anything about, bring some frameworks that I do know something about, and try and help AI, you know, bring to the fore how this way of thinking helps me look at this problem. So, yeah, example, you know, there's a lot of talk about loops. Loops are my current favorite thing. Everyone's talking about loops. It's like, well, actually, let's go back to John Boyd's like his real definition of the UDA loop. Let's talk about you know creation destruction, his later work. Let's say what he, you know, is it stuff people are writing about loops? Does it actually stack up with you know, probably what the the best thinker around loops? Has thought about. And actually, I think when you when you stack up on that, people are so obsessed with the hill climbing side, it's like actually, you know, Boyd said, you know, later in his career that actually orientation is the most important part of the UDA loop is the stance you take, the representation you take. Actually, you're willing to throw away your representation and look at the effectively look at things in a different way. That's almost the antithesis of a hill climbing loop. Yeah. Sits in in one frame. Uh I think it was back to like you used thoughtfully, the machines really can, I think, bring different perspectives to problems. And you never won by doing the same thing as everybody else did. You had to do something that people can't, won't, or to some extent didn't think of. So I guess my other piece of encouragement is for people to just think about what you can do with the machines in that way. Don't just bring the frame that you are used to. How can you reframe the problem? You know, even ask the machine for five or six ways of looking at this problem differently that might like, you know, elucidate something for you. And the main, I mean, I I you know, I do think the fine, I mean, actually, my most important piece of advice at the moment to people is go and read something. Read something you otherwise wouldn't have read. Yeah. Some of the time you would have spent wrangling with the machine, go read a book by someone who's thought long and hard about a problem or a frame of reference, read that book and then try and bring us some of that back into your dialogue with the machines. Because I think that's uh that there's so many ways of looking at the world that you simply won't discover if all you do is prompt.

Loops Reframing And Learning Beyond Prompts

Jason Rome 23:21

Yeah. I mean, to to your point there, you know, I I find, you know, there's there's time spent with the machine, there's time spent reading. And I've actually found, you know, now I actually need time just for my subconscious to operate with with no input. You know, I I um, you know, it it has become so much of an information overload and the ability to do so many different things that, you know, I I find I just need more time just listening to music and and letting things zone out and being able to kind of think critically and and synthesizing and letting it all come together in my brain. And then that that's when really good ideas come up. I mean, and that's one of the things I I love about these tools, you know, I um and I'll say I think this is important for everyone, you know, because I I think it's impossible for organizations to keep up with all the things that are changing and the expectation of trying to train people on the tools. And so, you know, if you want to stay up to speed in the space, you have to be using the tools on the side. You have to be personally investing in learning them. And so, you know, I've taken that dive. You know, I have the personal licenses and building some side projects, but you know, such a cool thing, you know, I'm I was playing some music with a friend this weekend, played a little gig and saw a pain point that that we had around managing the set list, managing lyrics, you know, staying in touch, learning new songs. And I was like, oh, I can build that all into an app with offline access. That's a progressive web app that manages all this like really cleanly in production 48 hours later. But then I had an idea driving into work this morning and between my car and getting up here into the office, you know, I was able to work with Claude on my phone and push code. Uh and that that is pretty cool. But like it was that time where my subconscious was actually thinking through the user experience in the car when I wasn't reading anything, when I wasn't looking at anything, when I wasn't in front of the screen, that generated that moment. So I think for people really taking space, walking, you know, getting outside, you know, whatever it is after the book, because I I think we're so over inundated more than we've ever been. We need to take that time. I really like your point on original thought. And I pulled up the uh the Newton quote, you know, if I've seen further, it is by standing on the shoulders of giants and kind of the concept that, you know, is there anything such as original thought, or are we always reformulating and building on those that have come before us? I want to go back to something you said, the this concept of of the calendar agent that you built and and the lack of uh, you know, the lack of consistency that you're getting from it. Cause I I think that's one of the other things. And and perhaps the models solve for this, but what I've seen, you know, the models feel so much like magic and they're able to get to pretty good, you know, call it 80, 85% accuracy so quickly. I I've found I think a lot of organizations underestimate what it takes to get to 95% or 99%, or even four nines or five nines, and when they need that. And so one of the things that we've started seeing is hey, do we actually have a probabilistic problem versus deterministic problems? Now, you all work in a space that is very probabilistic of how you guys think about investments. Obviously, it's it is impossible to know the outcome, and you guys have to consider hundreds of factors, both within a company and from a macroeconomic perspective. You know, have you seen places where we're trying to use probabilistic models to do deterministic things and guiding teams, you know, when to when not to use AI where we've overapplied it potentially, and there are simpler, cheaper solutions in the market?

SPEAKER_02 26:28

I would say I have, I mean, actually, I think most people are finding, I mean, certainly in our organization, you know, we come with a sort of, I'd say, a sort of pretty mathematically lit, you know, literate group, and they're very and they're very grounded in the maths in that sense. Uh, and I think they're pretty unlikely to want to use like, you know, non-maths when maths is is is is required, right? So I think actually we're fortunate, I think, given the sort of research heritage and the professional heritage of an organization like ours, we don't really step into those concerns. But what what I what I do think people are starting to um again, this comes back to where you can find the value, you know, where the machines start to play in spaces that are generally additive to you. You know, if you're thinking about scenario analysis, yeah, you know, you think almost every human scenario analysis you've ever seen, it's like, you know, paint one of five possible futures, you know, some sort of path that threads through a few of them, then they sort of diverge in some way. You know, being able to look at scenarios through lots of different lenses, look at it from the geopolitical lens, you know, look at it from the sort of regime economics lens, I don't know, look at it from the um, you know, power lens of institutions. The ability to look at problem, you know, provide different perspectives and different frames onto a problem, I think, is something machines are actually brilliant at. You know, you can point them in a direction and sort of have a very different set of perspectives where you can then start to think about how you integrate. I think people aren't genuinely are now starting to use it in the way it in like a good way. I mean, I think of these machines in some sense as, of course, you know, with tool usage, they can run, they can write code, they can run code, they can do all those things. But I sort of think of them as abduction machines in some sense, right? They play brilliantly in that space, you know, hypothesize with me about this from this particular perspective. Um, and it comes back a little bit to like the practices. And I think these practices are co-evolving as people use them all. I am sure there are going to be organizations that walk into attempting to do something, you know, which is, you know, must be deterministically repeatable and consistent and use these machines. It reminds me a little bit of the there was a TED talk by Malcolm Gladwell. He was talking about this sort of famous bomb site, the Nord, I think it was a Northern bomb site. And he it the long story of the long and short story was like don't build things that shouldn't be built. Yeah. And I think we are a little bit in the phase at the moment where we are building many things. All of us are building many things that in the long run probably shouldn't be built. You know, I I knock up stuff all the time. The sum contribution of what I knock up to to like humanity is like pretty limited. I mean, I did a PhD back in the day. I was told there's two sorts of PhDs. There's like genuine additive PhDs that contribute to the sum of total knowledge, and there's longer service awards. I sort of see like what's going on with us building things a little bit in the same way. There's things that people are building where you're gonna, in time, we're gonna get it. That was great. I'm delighted that you had that idea and you were able to build it. I do think a large portion of what we're building, I mean, it's entertaining, it's fun, but the sum contribution to humanity in the long run is limited. And I think actually back to the sort of price signal, let's see where it goes in nine months' time. But I think that discipline is really helpful inside organizations. It's back to like, you know, the things that you don't do are as important as the things that you do do. Yeah. When everybody in principle can have a go at anything, it's encouraging for everyone to have a go at everything. And I think we have to come back to that actually, what's genuinely valuable. I, you know, you could argue with three years in, shouldn't it all be a bit clearer by an hour? But it is only three years in to what is probably a you know 10 to 15 to 20 trans 20-year transformation journey. I'm perfectly comfortable that it's not clear, but I think we also sort of just have to be very cognizant of where the real opportunities are and also candidly where the real risks are.

Jason Rome 29:58

Yeah. And and and that's where, you know, there's the um there's a proximity bias of a lot of the people, you know, writing and managing the zeitgeist, if you will, of the moment, are the people building the technology and and using the technology every day. And so, you know, but probably worth a degree of skepticism when when people say, you know, 90% of code will be auto-written and you know, 50% of this job will be eliminated. And you know, we've seen a lot of people have to walk back some pro prognostications that they've made over the last uh six months to a year when it when it comes to that stuff. You know, to it's a it's a very wide consensus, right?

SPEAKER_02 30:33

I mean, you can sort of you can judge the degree of certainty by like how wide the consensus is. And this consensus is so wide you can, you know, you could put a cruise ship through it sideways, no problem.

Jason Rome 30:42

Yeah. And to

SaaS Future And Build Versus Buy

Jason Rome 30:44

your point though, about you know, this makes it accessible for everybody, you know, did the world need 10,000 new personal habit trackers? Maybe not. I think to your point on the price is like, when is it worth, you know, signing up for a $5 a month habit tracker app versus a $20 a month clause subscription where you manage your own? You know, and it's it probably will depend on the person and and how much you can use for that. I I do think we will eventually settle out on this concept of, you know, personal or small batch software, you know, where and I think that's one of the things that excites me is, you know, I I I you know, I personally believe the the SaaS pocalypse has been a little bit overblown. Um in it, you know, there are a lot of very different classes of SaaS products, you know, where yeah, if you're just a data visualizational workflow on a on an Excel spreadsheet, basically, yes, you probably are in trouble. But there's a lot of systems of records, a lot of really complex business products. And, you know, for for some of these SaaS products, just the cost of paying two or three people to maintain an in-house one is probably less than the cost of what you get from that vendor and the innovation that you get. But I do think there's a market gap of, hey, I have a team of five people that does this thing. Let's build personal software or a set of agents for them. You know, I think you and I have talked about before, though, you know, how ephemeral are agents. Like, is it worth an ever improving an agent or is it just cheaper to rewrite it when the new model comes out? I I think that that's one of the things that we've talked about versus there's been a lot of uh anthropomorphizing of agents a little bit versus the, hey, you know, it's you can it it takes five minutes to to kind of give instructions to one to have it to do something versus the ones that you build over time. I'd be I'd be curious your thoughts of you know, where does this potentially like balance and land out where you know in an organization people will have kind of personal software versus enterprise software and and and and what you've seen of what's that balancing act of like top-down versus bottom-up for companies?

SPEAKER_02 32:32

Yeah, I think you know, I I think there will be. I think not just I think to your point on the sort of small team level, like you know, if you're running a startup 5, 10, 15, 20, 30 people, I can imagine you building flavours of your developer experience, flavours of your sort of full process. I I can legitimately have lots of that running sort of agentically, and uh, and I think that's probably that's probably right. I think from a um you know big enterprise perspective, I mean you start to hear people talk about things like you know, classic systems of engagement, CRMs, those sorts of things that I think are you know more susceptible to um, I guess the the argument around the SaaS apocalypse. You're looking at systems, you know, systems of record. I think for a period of time, systems of record actually not, yeah, and actually because actually often the systems of record are much more about the coordination of the organization around a cohesive perspective of what's happening. Systems of action starting to operate on systems of record for sure. You can see more of that happening. I mean, take things like underwriting, where you may or any company may re underwrite an investment periodically. You might do that more continuously to keep your information more up to date with the way the world's changing. So I think anything where you've been uh sampling, you can see sort of moving into uh into systems of action, you know, external services like audit and valuations, those sorts of things are classic. I think those are genuinely will be sort of value chain disruptions that will happen. Um, but remember, software is also very nichey as well, right? So, you know, it's like provide, you know, providers will seek the price they think buyers are willing to pay. Um, and let's say you happen to have a sort of niche CRM system for investors, that is something that's probably you know more susceptible than it would have been in the past. You know, certainly we've done our own sort of look at the economics of our technology landscape. We do think there's a set of services that we buy and procure that if their economic, if our providers' economics don't change, then the opportunity cost of us building it is actually worth doing. So I think there is a there is a will be a shift in the build and buy dynamics, largely back to just price and supply and demand. You know, but the the basic economics driven by price. Now, interestingly, if if the if the prices go, you know, if the prices go one way, as we were talking about earlier, that might bring it back into an equilibrium. But you've got to assume that if you know if prices change, then you can start to look at the opportunity cost of build versus buy, and you will start to make some change decisions. And certainly we've done our own analysis and we think there are a set of things that we are likely to do, something different like build versus buy in the past, if the economics stay in the same form they are at the moment. Yeah.

Jason Rome 35:00

I I I forget where I stole this from, but I I love this quote that there's a there's a lot of big SaaS companies that have hostages instead of customers. And and I think there's, you know, there's a lot of people that have been kind of shoehorned into a SaaS product where they they suffer from a Goldilocks, where there's either a lot of features they don't utilize or um they're they're under featured. And so, you know, you never want to be someone's biggest or their smallest customer in some of these spaces. And and to your point, I think especially those places where people are really underlied, underutilizing features. And you've had a lot of companies with these feature sprawls without repackaging things over time, you know, hey, let's just simplify, you know, to your point, our CRM. Like we don't need all the bells and whistles that we get with some of these complex things, and we want to be able to control our views and customize that really easily. Um, and and for me, I I know we weren't gonna go down the SaaS pocalypse topic today, but here we are. You know, what I see is there, you know, because SaaS has been such a growth market for a long time, a lot of SaaS companies' strategy is the average of what their top 20% customers have asked them for over the last five years, instead of saying, hey, we have a vision actually of how your business should work. And so pivoting from selling technology to selling a belief and your value, you know, a lot of people talk about modes these days, but you know, your value being we work with a hundred companies to do this thing, you know, be whatever process it is. Do you trust us to know that better and to encode the way of that work into your company and integrate with your ecosystem versus, hey, we're gonna provide some tables and some workflows and some you know reporting on top of a database that integrates painfully with a couple other things in your company. So it'll be an interesting one to see how that continues to shake out. I I think, you know, there'll be disruption, but I think to your point, the the build by analysis and and you know, also there's there's a lot of companies out there that don't want a big technology team, that don't want a big technology department, that don't want to have to think about that stuff. Um, I want to pivot quickly, you know, you you've talked about a couple of times now, you know, finding, finding where technology can kind of amplify your

Redesigning Enterprise Workflows For AI

Jason Rome 36:58

team. And, you know, with that, one of the things we've seen, you know, organizations are moving faster or slower than others. And, you know, we've talked about, and a lot of people have written about, you know, using AI as an overlay on an existing value chain versus rethinking and business process re-engineering. What have you seen? Like, what's been easy and hard around, you know, either across companies or even within the company, of being able to extract maximum value. And hey, when it makes sense to, hey, we're going to use AI for this task and you know, hope for 10, 20% versus, hey, let's throw the whole thing out and start over again. And the ability to do that, you know, people's ability to kind of see that. What are you experiencing there?

SPEAKER_02 37:38

I mean, I say overall, notwithstanding, I think we're all still coming out of the area of individual productivity. I think we're for I've probably said before, you know, our our our value chain is really information processing, and these aren't information processing tools right here. You know, the last three years we've had machines that can read and write unstructured language and then you know start to do flavours of reasoning and now sort of you know tie together tasks and tool usage. Uh, why wouldn't that be genuinely valuable in a sort of value chain like ours where we're sort of processing a lot of information? So I think I feel we're fortunate that the argument around you've got to rethink your systems of work, you've got to sort of redesign with a you know, the analogy of think of the factory floor before elect you know, electricity and sort of redesign the factory floor. I think people understand that and are willing to move into it. So we're already underway with a number of sort of system of work redesigns. Uh that's that's proving valuable, but many of those are in uh you know things like valuations, for example. They might be in like due diligence. Now, these are all important parts of the ultimate investment decision-making uh process. But I but I do think it's fair to say we're we're not, I don't think I'm certainly not quite clear yet in the long run whether this technology will allow us to make better investment decisions. You could sort of say, well, if the quality of your diligence is better, you'll sort of weed out, you know, you know, more bad opportunities, you know, earlier, if you're able to sort of, you know, revalue more frequently, you'll be able to take out. I think all of that's true. And you could say, well, they're all contributing to ultimately being a better investor. I guess I come a little bit back to sort of like investment decisions when I think about do you take this opportunity or this opportunity on a relative value basis when the investment thesis is going to unfold over a five, six, seven year horizon. That I think is just a little bit more challenging intellectually to think about how you determine you are making better investment decisions. Uh, I think we've got great examples where we're already on the in so things like you know, private markets, we're better able to, I would say better able to analyze the skew of upsides and downsides that we're, and that in itself, I think, can lead you to better investment decisions. So there's, I think there's evidence pointing in the direction that it will make us a genuinely better investor. But if you said hand on heart today, could I be certain? No, I'd I just I do my usual too early into this, we're still sort of discovering, you know, what's possible. But I think we're fortunate our value chain is um is as well tuned to using this machinery as anybody's value chain. I think if I was trying to figure out how to do all of this in a factory floor with, you know, supply chain planning, and I'd look much older and much greater at this point in time.

Jason Rome 40:09

Yeah. I I think to your point around, you know, are we making better investment decisions or just avoiding, you know, missteps is a good question for anyone when they look at a process of, you know, is this an opportunity to raise the floor or extend the ceiling potentially? And, you know, I'll I'll say in my career, you know, especially when I think about digital transformations and new product launches, just avoiding anti-patterns greatly creates a chance of success. And, you know, I I think at the end of the day, a lot of these AI tools, you know, they're really good at guaranteeing you at least get to average. You know, they're they're, you know, hey, they're gonna get you to a C plus or a B minus. And if you use them to build your slides, at least your slides won't be horrible. Uh, like there's there's no chance that they will be really bad, but they're not gonna be great necessarily. And I think there's a there's a lot of places where, you know, just not being bad is is a pretty good thing. I've done a lot of customer experience work over the years and a lot of great customer experience is being forgettable uh and just not being memorable and just not messing up.

Leadership Empathy And The IQ EQ Split

Jason Rome 41:03

You know, I I want to get your thoughts on what role leadership plays, you know, across all of this AI. You know, I heard something the other day on a podcast that talked about, you know, AI adoption in the company will go as far as the CEO goes with it and embraces it. But then, you know, I've seen some companies where, you know, leadership role is really about culture versus it's about gold, you know, guardrails versus it's it's about training, you know, and as the the hype and anti-hype cycles come and go, you know, how how do you see leaders supporting this and and what you know, and what do leaders maybe not have to do versus what they have to do?

SPEAKER_02 41:36

I feel I'm you know, maybe I mean Jason, you and I spoke beforehand about, you know, I'm I'm I'm definitely on the roller coaster of AI. You talked to me three years ago. I was up here, probably when we spoke last, I dipped a little bit, but still on the sort of on the mountain side. I'm probably a little bit in the valley at the moment. And and the reason I say that is, you know, I remember standing in front of a a group of our leaders, our managing directors maybe three months ago, just three months ago, and said, look, I probably never stood in front of a group of just like managing directors from across the business and said, you really should use this technology in your spare time at the weekend, you know, really get to know and really understand it so you can have a sense of what your colleagues, you know, need to learn, need to develop, where the weaknesses are, etc. Um, and I said, I've never said that about any technology in my career before. You know, when cloud started coming out, you know, whatever it was 15 years ago, I didn't stand in front of a bunch of business leaders and say, you know, you should be playing around with AWS at the weekend and sort of getting to grips with that and sort of say, you know, go and talk to your technologist about how this might be useful in your business, etc. I probably, I probably still will stand by that statement, but I was probably a little bit wrapped up, well, I was probably also just a little bit wrapped up in the hype of, you know, Claude Code and Codex and the things that I felt I was able to do. So I think it comes back a little, I mean, maybe there's a something in here about empathy, right? Um, like most of us, here myself included, will have made sort of predictions about, you know, how straightforward somebody else's job is and how this might be possible to do this in the future, and how, you know, I've been frustrated with tech in the past, and I can build something. You know, we've all made perspectives on how work gets done. Often how work of our colleagues and other people's work gets done, you know, and certainly providers that we've worked with works get done. And I think there's probably something we're missing, you know, back to like what might this lead us away from. I worry actually it'll lead us away from empathy, right? The fun so if I take some of the things I do now, and we all do them. We get sent something to your point on, you know, how long you spent to write it, we'll say how long I spent to read it. If I can say, I'll stick it through my special AI slot of remover skills, I'll sort of get the essence out, I'll determine whether it's worth my attention or not. You know, in the past, I might have rung someone up and said, look, what are you trying to get at here? I think you're trying to tell me something, but what am I trying to get? What are you trying to get at here? Let's have a conversation about it. You wouldn't bother, you know, sending me something unless there was a point behind it. So I do still think it's important for I think I still think it's important for leaders to get to grips with this technology and understand it, but but to understand it, to understand what it can do, to understand what it can't. Do to understand where you still need to engage with people and still get to the bottom of things and still understand why people are concerned about things. And these are you know, these are the cup the phone, go and see somebody get to the heart of the matter type conversation. So that's a long way of saying it's not about just how far you climb the hill on using the technology, it's how you know it's it's that and how well you understand what you've got to keep doing and doubling down on. I mean, I said before, you could argue this is an IQ disruption and we're just bundles of IQ and EQ. And if IQ is being disrupted, you should lean into your EQ. I think that basic logic genuinely still hands. I think in the last probably year, I might have got, even I might have got wrapped up in the sort of the excitement of the IQ half of the equation. Yeah. But I think as a leader, you've got to understand the technology so you can know what's happening to that side of the equation. But my intuition says we should all keep leaning into developing our EQ, our empathy, our understanding, our points of view, our perspective. So I think that's still right. And so I think, yes, it'll go as far as leaders take it, but I think the reason leaders should use it is so they can really figure out the full rounded picture, not just the IQ half of the equation.

Jason Rome 45:06

You know, one of the things that I try a lot to do is backtest a lot of things that that we try to build and figure out what works, what doesn't, to your point around your calendar, what's what's the degree of variability? And so I I did this with AI-backed research, right? Because that what's one of the things is how how does AI compress discovery and research and desk research and all that stuff. And so I took 12 different kind of example research programs and kind of plugged the brief into Claude. I gave it a script, and then I had it benchmark against the actual findings and figure out, hey, what what's the gap? Like, what did it find? Like, what did you find that I didn't? What did we find that you didn't? And where did we align? And what is there stuff that neither of us found potentially that you would frame this differently? And it was about 65%. And that 35% to your point, a lot of it was empathy or narrative and storytelling of the ability to kind of pull one really poignant story out of a of user experience, kind of frame that into a ha, aha moment and use that as vision and and setting. And so, to your point around like empathy, especially for leaders, being being able to understand where the technology starts to break down and what's hard and what's easy. Because like lovable is a good example where you know, we see a lot of companies where the exec team gets their hands on lovable, and all of a sudden now instead of the product team getting an email around, we want this feature. It's here's this lovable thing. Let me know when it'll be in production. But, you know, unless you've unless you've tried to take a lovable prototype and then actually get that over and and migrate it to a database and you know, go through it. You know, I did this one with one of my projects, and there was a lot of like hard-coded, like just slop everywhere. And it took a really long while it took a really long time to like unwind that. I'm not integrating into a bunch of enterprise systems. And so, you know, being able to to know what it can do and I think what it can't do, and exploring what that last mile looks like, which is which is where a lot of the the problems are here. Uh I want to look at, you know, let's get back on the phone nine months from now. Let's

Predictions For 2027 And Closing

Jason Rome 46:58

maybe do a little uh lightning prediction round here. What so what what what's nine months from now? That'll be what early, early 2027. That that will get on the calendar here, first quarter. Uh, what do you think will be the same? What do you think that will be different? You know, where where are you in terms of a uh, I guess you're an investing company. So what's your buy hold sell right now uh across everything?

SPEAKER_02 47:16

You know, I did um for a bit of fun. You remember that um that that the song called uh Trust Me on the Sunscreen? Yeah. I did a token, I did like a token, trust me on the token version of that. I haven't got it here, but one bit that in there was it, there's a there's a point in that song where it says prices will go up. And so like first thing is prices will go up. We'll get back to working out, you know, we'll we'll we'll be forced to, helpfully forced to uh work out where we want to use this technology and what the value is worth the price. So I think that I think that's gonna be generally helpful. Um we will hear a lot about loops in the next nine months, and everyone will go hill climbing. In nine months' time, we'll hear a lot about were we climbing the right hills or not. That's that's my that's my second.

Jason Rome 47:58

We've gotten a lot of we'll be in great shape, uh, but uh you know, we might not have gotten anywhere important. So it's like my my dad used to tell me when I ran cross-country in high school. He's like, it's called cross-country, but you always end up exactly where you started.

SPEAKER_02 48:09

So I think we will we'll be we'll be looking back and going, yeah, the next fascination was loops, and we'll be talking about hill climbing and actually we'll be asking questions about did we did we did we climb to the right hills or not. Um maybe the uh I'm genuinely hopeful that the other thing we will see more of in the next but in the next nine months is more of a differentiation. I mean it's almost like you know, take like a fable five and opus 4.8. On the things that I did, I couldn't tell the difference between them. Right. And again, I think when we were being very undifferentiated in the application of the technology. Yeah. And I think starting to get to be more precise and differentiated and optimized into the application technology. You know, I'm a big believer like premature optimization is the root of all evil in community, you know. I think Donald News said that in computer science. I think that's right, but I I don't think it's premature optimization now. So we're starting to get to the point where necessary optimization, you know, it's no optimization. I think you're starting to see uh more diff, you know, more differentiated, thoughtful application application of the technology. Just because you can build something doesn't mean you should equally, in the same way that just because you can max out a you know a coding test doesn't mean you know what to build next quarter. So I'm I'm genuinely hopeful we'll get back into a, in some sense, a more thoughtful discussion, particularly around product and what to build.

Jason Rome 49:20

Yeah. Yeah, I I like all those. I think to your point around price, it'll be interesting to see too how much room does that open for more of the the kind of small language model of companies that you know are are much more focused. And, you know, I I've even seen, and if you kind of watch, you know, Claude or Codex as they work, they are calling cheaper models to do cheaper things. You know, they're they're trying to improve some of the to your point. I haven't seen a huge output difference, but I have seen my token, my token stretch a little bit further with some of the models. And depending on who you talk to, either you use the most expensive model because it's more efficient and you do less work, or you use the less, you know. I think we'll start to see some some wisdom. And I think there's gonna be a whole lot of organizations with token cost dashboards and management and the the CFO probably get a little bit more involved than than they are right now. I I think the other thing that we'll see there too is probably a little bit more focus on the UI and the application layer for using these tools to make it a little bit, to make it a little bit more achievable for someone that doesn't want to have to go and to you know, Vercel and GitHub and Superbase and like, you know, all these tools that are recommended out of the box and and there's work to do there. I think that I would encourage everyone to go learn that stuff, but I think it's also probably still intimidating to to a group. And one of the most common questions I get from folks is, you know, how do I deploy? Where do I control? How do I work with agents? You know, I can I can email them, I can, you know, you know, tweet at them, you know, I can see them on a dashboard. I think we'll start to settle on the the UI and orchestration layer and the app layer a little bit as the uh as people start differentiating on the experience as as we start to solve the adoption problems. I I would be a sell, I think, on the job pocalypse side of this. Again, I think we're seeing that, you know, the the jobs report data and and we're seeing actually a lot of you know open roles for product. And as technology solves more problems, we'll probably need more engineers who understand this technology. I think with I think SaaS will be really interesting. I think we'll see a bifurcation of different approaches to, you know, SaaS trying to identify, you know, I think some will build agents. I think some will just say, hey, we're just gonna, you know, hook ourselves up into an MCP layer and you can kind of plug your agents into what we do well. I I think you'll see others that have a, you know, bring your own token policy versus trying to like orchestrate on their own and just kind of plugging in. And so I think you'll see a lot of people trying a lot. I don't think nine months from now we'll have solidified the exact perfect business model of, you know, are people paying for outcomes just because I I think just contracting and procurement-wise, we're not that sophisticated still, even though a lot of people talk about paying for value instead and it'll take time. Yeah. And and, you know, I I think I think those would be my three. Uh again, to your point on on the models. I think I said this before we started. Most people don't have a model problem, they have an operating model problem unless you're doing PhD level thinking and and research, which, you know, a lot of us are trying to respond to our emails, right? And so uh any other closing thoughts? Anything we didn't touch on, John, today before we we go through. We said we weren't sure where we're gonna go today, but you know, I I think you know, wherever you go, there you are, and and and here we are at the end of it. So nope.

SPEAKER_02 52:21

I think there's plenty of hills to climb and there's plenty of hills to find, and there's plenty of hills to avoid. Yeah, and we'll find out in nine months which which ones we got right.

Jason Rome 52:29

Yeah, I think maybe uh maybe that'll be our title here is uh, you know, finding the hill to climb with AI can can be our podcast. So we'll appreciate it, John. We'll we'll talk soon and and uh maybe maybe even sooner than nine months, depending on uh what happens next. So appreciate it. It's been a pleasure, Jason. Thanks so much. Thanks.

Josh Lucas 52:45

Thank you for joining us on Build What's Next Digital Product Perspectives. If you would like to know more about how Method can partner with you and your organization, you can find more information at method.com. Also, don't forget to follow us on social and be sure to check out our monthly tech talks. You can find those on our website, and finally, make sure to subscribe to the podcast so you don't miss out on any future episodes. We'll see you next time.