Subscribe
Share
Share
Embed
AI First with Adam and Andy: Inspiring Business Leaders to Make AI First Moves is a dynamic podcast focused on the unprecedented potential of AI and how business leaders can harness it to transform their companies. Each episode dives into real-world examples of AI deployments, the "holy shit" moments where AI changes everything, and the steps leaders need to take to stay ahead. It’s bold, actionable, and emphasizes the exponential acceleration of AI, inspiring CEOs to make AI-first moves before they fall behind.
Paul Roetzer (00:00)
And it seems like what most companies have missed is the bigger question of
How do we actually drive growth through new markets, new product ideas, new ways of doing things, not just 10 % improvements, but 10X innovations. And to do that, it has to be treated as a people problem, a change management
Forum3 (00:20)
This is AI First with Adam and Andy, the show that takes you straight to the front lines of AI innovation and business. I'm Andy Sack and alongside my cohost, Adam Brotman. Each episode, we bring you candid conversation with business leaders transforming their businesses with AI. No fluff, just real talk and actionable use cases and insights for you.
Welcome everybody. Welcome Paul, welcome Adam, where we have an exciting episode.
Paul, please welcome Paul Reutzer, friend of the firm.
Paul Roetzer (00:58)
It's great to be with you guys.
Yeah, we we usually just do this like privately between the three of us So it's kind of fun to have this conversation publicly
Forum3 (01:06)
known Paul now a couple years. He's a friend of the firm. He's an advisor to Forum 3. We are avid fans of his podcast, but I'm going to allow him to introduce himself for those of our audience that don't know him.
Paul Roetzer (01:20)
Yeah, so my background came from the marketing world. spent the first, gosh, I don't know, I started in 2000 in a marketing agency, started my own in 2005 and built that up, became HubSpot's first partner in 2007. And then in 2011, I got really intrigued by artificial intelligence and started trying to connect the dots of what the technology was and what it might mean to the marketing industry. And then more broadly, the business world. And then eventually,
Realized I was going to change everything and created the AI Institute to focus on that. And yeah, and then somewhere along the way it actually happened. So I started the marketing institute in 2016 and then obviously everything changed for all of us in 2022 with Chad GPT. But yeah, we've been at it for a long time. We run AI events, do online education, some advisory work, but yeah, today my life is definitely spent focused around trying to drive AI transformation in companies.
Forum3 (02:11)
It will start because Adam and I have been on our own ⁓ journey, obviously with our book, AI First. And somewhere along the line, we we ended up, think, first developing a digital relationship with Paul. But for those of you that aren't listening to his podcast, it's a mainstay for us. We're huge fans. And so I'll put a plug in right up front for for his podcast ⁓ to our listeners.
Paul Roetzer (02:37)
I'm pretty sure Adam reached out to me on X. That was like the origin of the relationship. think he reached out and he's like, Hey, we should have a conversation. And yeah, I don't even remember that was it was probably 2023. I think. Yeah.
Forum3 (02:39)
Thanks.
Adam Brotman (02:48)
It's right when we were starting to write the book, right? Because I was like, look,
Forum3 (02:50)
Yeah.
Adam Brotman (02:51)
it's a long time listener, first time caller, And I just want to echo what Andy said for our little community, which is much smaller than yours, Paul, and we're part of your community, is that for our community, we say this a lot. But I'm going to say it again.
Paul Roetzer (02:54)
Yeah, that's great.
Adam Brotman (03:08)
There's very few cheat codes in life and in business. But one of them, if you want to stay abreast of what's happening in the AI world from a business perspective in particular, but overall, you got to listen to every episode, if you can, of what Paul and Mike put together every week on the artificial intelligence podcast. think that's what it's called.
Paul Roetzer (03:27)
Yep. Yeah. It's a, and honestly, it's like a bit of a cheat code for us. Like if I wasn't doing that podcast, I wouldn't, there was no, there'd be no way I would stay as tight into everything that's going on. It's like our forcing function to every week, synthesize what's happening. And then I've always said like Mike and I would be doing the podcast if no one was listening because it truly forces us to stay into everything that's happening.
Forum3 (03:28)
Yeah.
Yeah, well, great with that. So welcome, Paul. are show today. I want to frame it. We're we're at the beginning of January 2026. You did an app. You dropped an episode at the end of 25. I think it was your last episode, in which you were sort of laying out the state of A.I. as we're entering into the as we're entering the year.
And so this episode, what we want to cover is really the AI 2026 operating map. And AI is, even in the last quarter of the last, at the end of 2025, feels like it's accelerating it on the exponential curve that we all know it's on. We know that the winners will be companies whose leaders translate those, shifts in AI into
actual, tangible operating and structural changes in their companies. And, um, and we want to, we want to talk about that. So we're going to talk about the operating map and really what CEOs should and could be doing this year to advance their businesses. So, with that, I'm going to start with. I, when preparation, Adam and I were talking, I was like, you know, if I had a magic wand, I'd set this stage and ask Paul this question. So I want.
We actually may do this. ⁓ We've been talking with Paul about doing a CEO summit. And so I want you, to imagine where we're in Las Vegas. We've got an audience of 50 CEOs of, let's just say, Fortune 1000 companies. And so it's a small intimate gathering in Vegas, which is a little bit of an oxymoron.
Paul Roetzer (05:06)
You
Mm-hmm.
Forum3 (05:28)
And it's 2026, it's January and I say to you, Paul, what's your advice for these CEOs? What should they be doing this year?
Paul Roetzer (05:36)
Yeah, you know, it's interesting. I have a talk actually in February to a private equity firm and it's like 90 CEOs from their portfolio companies that we're doing this talk for. And it's part of their conference. It's like their portfolio event. And so the talk I'm doing there is the AI innovation imperative. And I think that
That is probably the primary thing that I'm advising CEOs on right now is I think after Chet, GPT and for the couple of years following, there was so much focus on optimization, efficiency and productivity gains and treating it as a technology, like an incremental technology solution. Like how do we layer AI into things so we can gain 10, 20 % efficiencies?
And it seems like what most companies have missed is the bigger question of
How do we actually drive growth through new markets, new product ideas, new ways of doing things, not just 10 % improvements, but 10X innovations. And to do that, it has to be treated as a people problem, a change management
Like this isn't just go buy Co-Pilot and give it to 10,000 people and think we've solved it. It's you have to do the hard work of holding their hands and like...
getting them to realize the value and getting the first few use cases and first few wins and then take it across the organization. And I know that's what you guys do all the time. Those conversations you're having too. So I'd say innovation and treating it as a people slash change management opportunity, not a technology problem.
Forum3 (07:08)
Adam, would you either compliment or what would be your answer to the same question?
Adam Brotman (07:13)
Yeah.
Well, I mean, I couldn't come up with a better answer than that. is the answer, right? So Mike, I have a question. I'm a little bit of a cop out instead of me answering it. But Paul, my follow-up question.
Forum3 (07:22)
Yeah.
Adam Brotman (07:25)
Because Andy said to me, what would be your one question? And I would say, because I liked Andy's question. You just answered it well. And I'd be like, what do you think the number one thing? I'm sure there's a lot. If you had to boil it down, Paul, what's the number one thing from getting in the way of companies doing what you just said and seeing it that way?
Paul Roetzer (07:44)
Yeah. Yeah, I will.
Like, I guess the one other thing I'll say before I get to that one too is, ⁓ I don't want to gloss over the optimization thing because most companies haven't even figured that part out yet. So it's not like we're RDA. That's table stakes. Everyone's figured out the optimization. Let's just move on to innovation. What I'm saying is it needs to be simultaneous. And to answer your question, Andy, the thing that's preventing all of it is a lack of understanding at a leadership level of what this stuff is actually capable of.
Forum3 (07:59)
That's it. ⁓
Paul Roetzer (08:10)
and then how to infuse it. And so I think what happened, least a lot of the companies we've talked to is the leadership, especially at the CEO level, just assumed IT, CIO, CTO, like the technology people would figure this out. And so they didn't take the approach of empowering the leaders of all the different departments to say, okay, here's what you need to know. Here's the technology you have access to. Here's what it is able to do.
now figure out the uses in marketing, in sales, in customer success, in finance, in HR. And so when you throw it to the technology people, it's not like they shouldn't be a part of the conversation, but they're going to look immediately at risk and safety and security. like all of those things is their true domain. Their domain isn't innovation and growth. That's not what they're there to do necessarily. And so I think that's really the biggest challenge is this lack of awareness and understanding at multiple levels.
that have led to a lack of vision and execution to take advantage of what these things are capable of doing.
Forum3 (09:12)
Can you ground that in an example?
Paul Roetzer (09:14)
Yeah, I I think the story we all hear all the time is like everybody's buying the copilot licenses or getting the Salesforce instances or, you know, getting Gemini or chat, GPT enterprise licenses, and they're, buying the, by the thousands. And that's why these companies are, growing with these tools and they're just providing them to people. So I have those conversations all the time with friends who are in marketing or sales or customer success.
And they're like, yeah, I have a copilot license. No idea what to do with it. It summarizes my meetings for me. I use it to reply to my emails sometimes. It's like, okay, have you ever built your own copilot? Have you ever built a custom GPT to like help you do your job? And they just stare at you like, what even is that? How would I even do that? And so you and I, like we live in this bubble where we assume everybody else knows what these things are. What's a gem? What's a custom GPT? How do I build one? And the reality is the vast majority of knowledge workers have no idea.
how to do those things. And so that's again, the example we just see over and over and over again, is people assume everyone knows what we know. They know what a reasoning model is. They know why chain of thought matters. Like what is the value of having that? What is deep research? Like how do you use notebook LM? Like notebook LM to us is like, it's amazing. Like for the last year we've known that it's got all these amazing use cases. You go talk to the average business leader in any C suite and you ask have you ever used notebook LM? I'm like, what the hell is that? How do I even get to that?
Adam Brotman (10:19)
They haven't heard of it.
Yeah.
Paul Roetzer (10:37)
Like, what is it?
Forum3 (10:38)
And so if you're talking to the, if we're in Vegas with the 50 CEOs, I don't know what notebook LM is playing one of the CEOs.
Paul Roetzer (10:44)
Mm-hmm.
Forum3 (10:46)
It's 2026, I get AI's critical. What do I do?
Paul Roetzer (10:50)
I'm always a big believer. You just have to show them a use case that is relevant to their life. So if it's a publicly traded CEO, pre-build a notebook LM that has the earnings call transcripts from their top three competitors for the last two years and have it run an analysis on them, have it build a 10 minute podcast talking about key takeaways from Q4 data. Like just do the thing where they're like, I had no idea I could do that.
That is something I spend 20 hours a month doing. Here's the challenge I have. So like the one I love to do for CEOs and actually I talk about on episode 189 of our podcast is a project I did over holiday break using my co-CEO GPT that I built. And again, I have no programming ability. I'm not a coder. I've never written a computer program in my life, but with a simple prompt and some knowledge base about our business, I build a custom GPT that functions as my co-CEO.
And so as I'm thinking about our growth plans, I have finance questions, legal questions, HR questions. I can't call up my advisors or my staff in the middle of a holiday break and ask these questions. So me and Co-CEO do all this planning, save myself hundreds of hours and probably tens of thousands of dollars. And I arrive at a plan by January 2nd, when I come back to business, that I would have never in the past been able to create over a 10 day break. And so I just, look at these really tangible things and I
I think that anyone who's leading transformation or advising on transformation, you just have to know what are the pain points for these CEOs and then show them how AI can help solve those pain points.
Forum3 (12:22)
Yeah, great. Adam, anything you want to add?
Adam Brotman (12:24)
No, I mean, it's you just said Paul that advice.
about showing them, think it's like Andy and I have this mantra that Andy started about show, don't tell, show, don't tell. it's it's and in fact, I hope I'm not speaking out of school, but we recorded this thing right before the holidays. Andy and I did is supposed to go on this pretty large, largely distributed platform where they're like, OK, Adam and Andy, this is, you know, you get two 40 minute episodes to like boil everything down. You
that comes from everything you guys have learned and everything you wrote in your book and everything you're doing. And Andy was like, I think we need to just show use cases. I it's exactly what you just said, And I'm excited for that to come out, but not because it's so brilliant as much as it's just an example of what you just said. Your Notebook LM example, over the holidays, Andy's probably laughing. Like, what did I do, Andy? sat there and I put, you know, we focus on restaurant brands, primarily. We restaurant and retail. I focus on, I took, I took
the last earnings transcript of every restaurant brand that I, it was like 19 of them, threw them into Notebook LM and basically said, I want to talk about the consumer. What can you glean? By the way, it is now Gemini 3, I believe under the hood of Notebook LM. They just switched it from 2.5. And I was like, what can you glean about the state of the consumer right now? And Andy will tell you, I sent this report out to Andy and our whole team and said, this was amazing, but it was, was,
Paul Roetzer (13:23)
Sorry.
Adam Brotman (13:46)
Because I'm in the bubble like you, right? Like I know how to do this workflow that most people don't know. So we end up showing, telling. So I love your example there,
Paul Roetzer (13:57)
Yeah.
Forum3 (13:57)
Yeah, so let me actually pause because it's a good transition to the next part of our what we had planned to talk about. Really what both both of you did over the holidays and you just gave illustrative examples of Paul, you talked about your strategic planning process and how you were able to do strategic planning in in a for twenty twenty six in a way that two years ago wasn't even on the tape, wasn't even available.
Paul Roetzer (14:23)
Probably six
months ago, honestly, like some of the stuff I was doing, I could not have done six months ago. Yeah.
Forum3 (14:25)
Yeah. Fine. Six months
ago. And you were able to do it and it blew you away. And you worked, you were working and had a different workflow with the tool with larger language models, whether it be Gemini or Claude or OpenAI in a different way to do your strategic planning. And then Adam, over the holidays,
Paul Roetzer (14:32)
Yeah.
Forum3 (14:47)
Like you zoomed through what you did, but it's actually a really good illustrative example, which is why I'm highlighting it, which is you took the 19 public restaurant companies and put them into Notebook LM. So you knew what the tool was and you knew how to use it. And you did a state of the customer analysis based on 19 data points and financial results from public.
publicly traded restaurant companies. And you did that for fun while hanging out with your wife and daughter and around the Hanukkah candles and Christmas tree.
That would not, you would not have done that last Christmas.
Adam Brotman (15:24)
No. I mean, I was playing with notebook LLM last Christmas, I, last Christmas, this is how fast things are moving, reasoning models because of deep research were maybe on my radar. I don't even know if they were by last Christmas.
Forum3 (15:25)
So the.
Adam Brotman (15:36)
It was probably not until February. Yeah. like I, yeah. So like I was so, but yes, Andy, to your point, like I, the, the advanced workflow that Paul and I both talked about, which by the way, anybody can do. It's not, I'm not a coder either. Like I, it's all in just natural language, right? But it was, but it is advanced. It is, I mean, it's semi advanced. It's not advanced as in, it's beyond the reach of the average.
Paul Roetzer (15:37)
I mean, they'd just come out in September. 01 came out September of 24. So it was just like brand new.
Forum3 (16:03)
Yes.
Adam Brotman (16:04)
executive or leader, but it's beyond how 90 % of people are using AI today.
Forum3 (16:10)
I highlight those as two illustrative examples, know, beyond the one that Paul mentioned that could be useful examples of ways executives are working with AI in new ways and the unlock that is available to the C-suite and trying to make that point to the 50 CEOs that are in our fictional audience in Las Vegas.
With that, let's transition the conversation because both of those things are illustrative examples really of what I call the future of work. This new workflow, which I think will come up. But let's talk about the 2026 AI operating map. And Paul, I'm going to turn it to you to talk about because you highlighted this. is familiar territory. Should be familiar because it was just covered in your last episode.
Paul Roetzer (16:41)
Mm-hmm.
Forum3 (17:01)
Like, let's start with like, I'm going to try, let me start with Paul and then go to Adam. Sort of what is your, what is the state of AI in enterprise as we enter 26?
Paul Roetzer (17:12)
I think more organizations are trying to look at how do we actually scale this across all departments into all workflows? How do we solve for its impact on organizational structure and staffing plans? The thing that happened kind of mid to late 2025 was a lot of more CEOs talking publicly about the expectation of efficiency gains from AI. So as AI got more agentic, as it got more reliable,
There was people like Andy Jassy at Amazon where like, Hey, we expect fewer humans. Like we're just going to come straight up and say, like, we think we're going to become much more efficient as an organization. And so that was what you saw a lot of CEOs starting to say and earnings calls when they were being asked about it. And then just like in employee memos, they were just addressing the fact that we're heading in this direction. But at that time, agents weren't reliable. They're not fully autonomous. And so I think there was a bit of a misalignment of
what an AI agent is and was at that time and what it's going to become or what we think they're going to become. And so that's probably the first part of it is agents kind of becoming more real. There was a lot over holiday break. We talked about on episode 189 of the podcast again, this idea of like Claude code. Like it was just everywhere on X. was like, if you follow AI people, that was all anyone was talking about was like,
blown away by the capabilities of this
Forum3 (18:38)
you
Paul Roetzer (18:38)
thing to all of a sudden build apps and just solve business challenges from a programming perspective, from a coding perspective. And so you can feel that we're moving pretty quick toward agents becoming more reliable across different business functions, different industries, becoming more autonomous to a degree.
And what I think is going to happen is organizations are going to start having to deal with the reality of what does that actually mean? How do we manage that? We just saw a pullback from Salesforce actually. So again, like we see it like we're getting better and better and more reliable, but then you have Salesforce basically saying large language models aren't reliable enough yet. So we're going to actually go in there and start doing some more deterministic programming of the agents and the workflows. And so I feel like we're going to go through a very messy.
12 to 24 month period where these agents still can't be autonomous, they can't be reliable, and yet SaaS companies are infusing them into their software as though they are already autonomous and reliable. And then they find out they're not, and there's this miseducation of consumers, misexpectations when you buy the products. And so I feel like that's kind of where we're at with agents is they're getting really good in some very specific areas. They're going to become more general and autonomous.
but we're not there yet and so it creates all this misaligned.
Forum3 (19:58)
Adam, you wanna, Adam, do you sum up where would you think we are in terms of, mean, Paul just laid out both where we are in terms of 2026 and then went deep on agents. Do you wanna give your take on the state of AI in 2026?
Adam Brotman (20:14)
Yeah.
I would
say I really like Paul's main point of this topic of productivity or this word ROI. thought, I'm sure Paul saw this as well, because I think our algorithms on Twitter or acts are very similar. saw that whole, the whole, listened to your, I started listening to your latest one 89 episode. And I, I would say.
Forum3 (20:24)
Ahem.
Adam Brotman (20:37)
that reminds me, Paul, I'm sure you saw Aaron Levy's, wrote this really neat blog post, I guess it was, talking about, it also got a lot of airtime on Twitter about too many people in the enterprise, so I'm answering your question, Andy, to bring it back to your question, they're thinking about ROI and they're too focused on,
Paul Roetzer (20:44)
Yeah, we talked about it.
Adam Brotman (21:00)
the R and not realizing how little the investment really is in the I. it's like, first of all, the I is $25 a month per head, maybe $100, depending on how you use it and depending on what kind of plan you're on and what functions are mostly using it. it's not like your typical SaaS software expense. The I is actually more investment of mindshare, of company time to get educated, to do the right governance.
Like that's the eye, but they, but that's not how most enterprises are still thinking about it. They're still thinking about like, how can I like reduce expenses or get more from existing people? And, you know, and, and I think that Paul is also right when he mentioned technology, like there's still this sort of muscle memory in every organization to just, let's just give this to tech and ask tech to handle it for us. And I think, so the state of the union is
Forum3 (21:34)
.
Adam Brotman (21:56)
I do feel like it's this really unevenly distributed situation where CEOs are more aware of the power of AI. Their, their teams are using it. More of them have like either authorized or unauthorized sort of workflows that they're using way more than a year ago. Like you go now and you talk to a company and Paul, don't know if you've had this experience and they're like, they're all sorta using it or not or whatever, but it's like,
Forum3 (22:02)
.
Adam Brotman (22:24)
for level one workflow usage.
And they're probably using, without naming any names or calling anybody out too much, they're using not probably the right platform for what they should be doing. And they're not using it correctly, even if they have the right platform. They're on auto mode. I can't tell you, just side note, how many times I talk to clients and they're like, I did this. I tried to do your more advanced workflow.
And I got this crappy answer. And I'm like, what model do you use? they're like, GPT 5.2. And I said, no. which model, which version of 3.2? And they're like, no, I always just use auto. And you see these insanely bad sometimes answers that come out of auto when they're trying to do a more advanced workflow. I think the state, to sum it up, I think the state of the union is this really
compared to a year ago in enterprise, this very chaotic almost moment where people understand the power of it. They're starting to use it more than they were before. But they're not taking the time to get educated through the kind of classes that you can get through Paul's company, for example. They're not getting themselves proficient enough. And so I actually think there's a bit of a
Forum3 (23:23)
.
Adam Brotman (23:44)
almost like cognitive dissonance happening in the enterprise right now that it will get cleared up. And I'll say one last thing on that, which is that, as you were describing the state of the industry, I had this image in my head of this almost muddy situation for a of enterprises, because they're like, is it about productivity? Is it about growth and innovation? Is it about
Forum3 (23:56)
.
Adam Brotman (24:04)
Is it about my learning how to use these tools or is it about agents? I keep hearing about these agents. And the answer to this is yes to all of that. And that's a hard, I think that's making change management really hard in the enterprise, and that long-winded way of answering, like, I think when you have all those things going on, plus the fear of losing your job and whatever, I think it makes change management really hard.
Paul Roetzer (24:27)
One thing I just real quick, I'll add to that that I was thinking as you were saying it is I feel like 2024 and 2025 was, oh, we better go buy some licenses for a gen AI platform and start doing something because they're asking about it on earnings calls and my peers are doing it. And like, we got to do something. It doesn't matter what the budget is. Just go spend the money $25 ahead and go get the thing. And then as 2025 starts kind of winding down.
The ROI question starts coming. like, what are we really getting out of these GNI platforms? We spent $2 million last year on chat GBT licenses or copilot. Like what did we get out of that? Is it really working? And, and it's almost the wrong question to be asking because the answer might be, we don't know. It's not really clear what the ROI is when the underlying question is, did you train anybody how to use it? Like, okay, yes, you spent 2 million on tools and yes, you gave them to your employees and utilization rate is low, but.
Forum3 (25:18)
It is love.
Paul Roetzer (25:20)
Did you actually teach them what use cases to apply them to or prioritize based on the potential return? And in most enterprises, the answer is no, we didn't do any of that.
Forum3 (25:26)
at the close of the enterprise.
Adam Brotman (25:30)
tell me if I got this right, Paul, the 24 and 25 people had to just get off the snide and be like, let's buy some licenses. Let's be able to answer questions to the board and the CEO that we're doing something. But, now everyone's sort of in this spot where they haven't been trained. They haven't slowed down to actually understand how to set themselves up and go through the right change management, go through the right learning curve.
Paul Roetzer (25:41)
doing something exactly.
Adam Brotman (25:53)
so they can actually use it.
Paul Roetzer (25:55)
Yeah, and think one of the tricks is ROI became, well, we can just get rid of some people. So like the shortest answer from a publicly traded company or a VC backed company or a PE owned company is, what's the ROI for all this money you're spending on generative AI? Well, we don't think we need as many people, so let's reduce the workforce 10%. Now, even if you don't admit that's why you did it, it now at least answers the ROI question. So we did the AI stuff because everybody was doing it, we had to.
Forum3 (25:55)
So,
Paul Roetzer (26:23)
Now we got to show some return. Well, the easiest way to show return when you haven't actually done the change management work is just get rid of some people and then tell Wall Street you were able to reduce headcount. It's like, great. Okay. Stock price is up 4%. And like, now we bought ourselves another three months to figure out the actual ROI of this thing.
Forum3 (26:39)
so let's, I mean, let's talk some more about, we, let's, we started to go down the path, high level operating plan of 2026 state of the union and the enterprise. got into agents. We've now jumped into change management. I want to give them Paul, is there anything else that you want to highlight from 2026?
Paul Roetzer (26:58)
At a high level, think the most important thing from a C-suite perspective is you can't build your organizational staffing plans, budgeting strategy based on what today's models are capable of. Because one, your team probably doesn't actually fully understand even what today's models are capable of, as we've already talked about. You're probably utilizing 10 % of the capabilities as a company right now.
But there's a lot of advancements being made. The models are getting smarter, more generally capable very quickly. Every three to six months, there's not a full reset, but definitely I got to revisit what's going on. And so as a CEO, when you're making plans for this year, and certainly when you're looking out, you know, two to three years, you've got to have some mechanism to be staying at the frontier of what these models are going to be capable of in six to 12 months. As a CEO myself,
We're in a scale up mode where we're going to be hiring. And I am actively trying to assess every role and saying, do I need this role in 12 months before I make the hire? And I think a lot of CEOs are going to do that with their existing teams. They're going to, but think about it. Does your HR team know to do that? Like is your HR team trained on what the models are capable of? What agents are going to be able to do? If the answer is no, then how in the world are you going to properly staff the company? Stuff like that. That's, so I think.
Where the models are truly and where they're gonna go is like fundamental knowledge that most leadership teams don't have.
Forum3 (28:26)
so I guess I'll turn it to you, Adam, like, like in terms of translating the 2026 AI operating model into CEO execution, like what's your guidance for
our virtual fictional CEO summit. What do you have to say to them about what they should be their execution for the enterprise?
Adam Brotman (28:47)
I would say, mean, building on what Paul just said about, you need a way, like Paul, Paul just described almost a paradox.
And you need a way out of that paradox. The paradox is you're not even currently fully up to speed, either as a CEO or as a C-suite or as an organization, about what these models are capable of. And they're about to get better in three months from now. And so what do you do? How do you break out of that? You need a way to both stay abreast but make sure the organization is sort of surfing that right. We've always said, and I say this to CEOs with you, Andy, all the time, like,
You need to literally or figuratively create a function that is an AI transformation function. it's weird. I lived through this when we created the digital function at Starbucks. And there was no such thing as a chief digital officer when we did it. Now it's just normal. But at the time, it was a radical thing to be a chief digital officer. And I used to almost be embarrassed to say I was CDO because there was no such thing. And
And I think I'm not saying we always tell CEOs is you don't need to go out and create a chief AI officer right now. A of people can't afford that. They don't even know what that means and how they go find somebody. So we always tell people like get a couple of people in your organization and deputize them and give them. It's funny, Paul. We always tell everyone like the first thing on their to do list. They have to listen to every second of Paul and Mike's podcast every week. Like that's a non-negotiable. I'm serious. Like somebody like
That's not easy, by the way. know, Paul, you know that. If you're busy, that means you're making a point to be like, got to get the hour and 24 minutes in and stop and think and tune your LinkedIn and your X or your social media algorithms or YouTube or whatever to stay abreast of it. Right there, that's a commitment by somebody in the organization, number one. And number two, those people that you call them a task force, you call them an AI leader, you can call them a chief AI officer.
Paul Roetzer (30:22)
Hmm.
Adam Brotman (30:41)
They need to have the seal of the CEO to really be a change agent in the organization. And what I just said is not as easier said than done. CEOs, they'll say they get it. And Andy and I, I'm not being snarky. I'm not throwing stones. CEOs will be like, on a scale of nine to 10, I'm a nine after talking to you guys and taking your boot camp and whatever. And then somewhere in their organization, it'll get sidelined. It'll get pushed.
back to IT, they just aren't doing it. And at some point it'll happen, just like it happened in the digital transformation era of 10, 15 years ago. that's my advice is Paul's right, you have to stay abreast of it. It's what Ethan Molyth said. Ethan Molyth told us, tell your clients, tell those CEOs they need to tomorrow start what's called an AI R &D lab. They don't even have seat licenses.
tell them to start a lab, they're going to laugh us out of the office. And he's like, well, then you're not being strong enough. You need to tell them to do it. Because if they're not experimenting, they don't have some group that's pushing every function to experiment and facilitate that, then they're not moving fast enough. And I think he was right. And that's still our advice, even though it's a hard pill to swallow for lot of CEOs. don't know, Paul, what do you think about that?
Paul Roetzer (31:57)
Yeah. Centers of excellence is a common thing that we're seeing pop up. And the other related theme, I think to what you're saying is the idea where the CEO is the chief AI officer. I actually had somebody that say that to me because we were thinking about hiring a chief AI officer with SmarterX and somebody was like, well, aren't you the chief AI officer? I was like, hmm, that's an interesting question. And I guess technically, yeah, it's like, I don't have the ability to build automations and like build
Adam Brotman (32:01)
Thank Thank
Paul Roetzer (32:26)
truly build the tools outside of what I can do within the GNI platforms. So I'm not the technical person, but I think the theory here is correct that it actually is probably the business leader who should be the AI officer. And so I think in some ways the CEOs who see themselves as the chief AI officer may be in the best situation to drive true transformation and innovation.
Forum3 (32:39)
some ways. ⁓
Yeah,
I'd have a couple of things to add that one is Paul, I think that's correct directionally, which is we believe that AI first companies start and end with AI first leaders. And that typically is the CEO. And so that's why this whole episode and focusing on this, the 50 CEOs in the audience is a really important one. And yet the technology is just like, you know, the agents, the,
longer horizon, more memory, the, just the, the, unless you're in the bubble of dealing with this stuff, the way in which work is getting done differently than the way for the last 30 years, the last 50 years it's been done. you know, for money, money of those CEOs, it's kind of like, yeah, I use AI to write an email. and so the gap is there, like it's the
People are awakened that it's super important, but the fear has grown across the enterprise in the organization and training is insufficient. Adam and I do change management support with enterprises and consulting. That's insufficient. both of, we're doing the good fight. And even Malik is right. probably should have been, we would have been right to say you should be starting an AI R &D lab today.
Paul Roetzer (33:56)
Yeah.
Forum3 (34:03)
but we would have gotten laughed out of the room. Yeah, yeah.
Paul Roetzer (34:04)
Staffed by whom? Like that's the problem though is like,
Adam Brotman (34:06)
Right.
Paul Roetzer (34:07)
yeah,
it sounds great in practice. was telling software companies like five years ago, they needed an AI, AGI horizons team, which is like, Hey, we'll get there. Like you need to have a team that's like figuring out what is your business model when we get there. And it's like, okay, that sounds cool. Who's on that team? Like who do we even, and now we're back to the understanding problem of like, there's so little literacy around truly what is possible. Like we have,
Adam Brotman (34:08)
That's right.
Forum3 (34:16)
Yeah.
Exactly.
Paul Roetzer (34:33)
We have basic literacy of, I can go use ChatGPT and I know you give it some prompts and I've built an image and I've done a couple things. Like, yeah, for some basic literacy. It's like, okay, what about the advanced stuff? Like how good are you at doing deep research projects and using Notebook LM and building custom GPTs? like, if someone says, what's the ROI of it? It's like, oh, I just saved 90 hours last week. Like that's the ROI. If you can't answer those things, your literacy and your competency is actually quite low. And so I would argue that most CEOs...
have very low true AI literacy if we benchmark it against things like know how to build custom GPTs, competent with deep research and notebook. Like if those are your things, then my guess is we're very low.
Forum3 (35:14)
We'll put a number on it, How low?
Paul Roetzer (35:16)
Oh, it publicly traded CEOs in terms of like competency. would say basic literacy is, you know, maybe you're looking at like that 50 to 75 % range where like, yeah, like they go in, they use it few times a week. Like they're playing around with it. The advanced stuff like Adam and I were talking about where you spent your break doing some project that either created value that would have taken you a couple hundred hours or did a thing you would have otherwise not had time to do.
Forum3 (35:19)
Yeah.
Paul Roetzer (35:43)
and could save themselves a couple hundred hours without having to call up IT, less than 5%, I would probably pretty confidently say.
Forum3 (35:48)
Yeah,
think less than 5%. Adam, anything that you wanna, we're coming to a close here. Any final question or comment that you wanna throw out?
Adam Brotman (35:58)
I guess my last kind of comment slash question for Paul, I'd to get Paul your thoughts of this, is like, you've got this combination of things that are complicated and moving very fast that are contributing to CEOs and other leaders being at that 5 % when, you know,
Paul Roetzer (36:26)
Mm-hmm.
Adam Brotman (36:26)
when it's like our collective mission between your company and our company and our teams are like, want to get them to that. We want to get these folks to that next level. I just want to end by asking you about the concept of agents. haven't talked. I mean, it's a big topic. was like a huge topic. Do you feel like if you could have
teams focus on something? Would it be just getting educated and don't worry about agents and don't worry about, you know, that topic? Or do you do you feel like no, you can't there's no shortcut, you gotta walk and chew gum at the same time. And is agents like a really important topic for business leaders to get their head out around in January of 2026?
Paul Roetzer (37:10)
I think the agent things for most people is just going to happen organically. They're going to be built into the systems. You can already go in and turn on agent mode in chat GPT. think, you know, Google Gemini is starting to infuse agents right into like Gmail and calendars and things like that. So you're going to have these like base level agents, which just is going to become secondary to people, I think to use. It's like, yeah, whatever you want to call it. I just turned on this workflow that does these seven steps for me to achieve this goal. Like it's basically what an agent is. It's just like.
Forum3 (37:18)
.
Paul Roetzer (37:37)
But there's varying degrees of autonomy of these agents. The thing I would say is we generally guide people to just get really, really good at your platform of choice. So if it's Microsoft Co-Pilot, great. If it's Salesforce you live in because you're a sales executive, it's fine. Google Gemini, chat GPT, like those are going to be the primary one that most people would use.
Get really good at it, use it every day. It's like anything else. Like you've got to keep the muscle memory building. Like you've got to be in there trying it for different things for yourself, for your business, but you need to have go-to reasons to try it. Like Mike and I talk a lot about this internally. When a new model comes out. So let's say you're big Chachapi to enterprise user and you've got three or four use cases as a CEO that you're saving 10 hours a week on and you're loving the thing. And then a new model comes up.
or new capability comes out as part of Chat GPT. And you're like, I don't even know what the use cases are for this. Like, I don't even know how to test that thing. And so that's what you run up against is the ability to even know, does the new innovation matter to you? Like, is it something relevant? But I think at a high level, if you just focus on getting really good at your Gen.ai platform of choice, you understand the feature set within it.
Forum3 (38:33)
.
Paul Roetzer (38:50)
You've got some go-to use cases and prompts you can use to do the different things that enables like deep research and standard deep or like thinking ⁓ version of it. You're going to get more value than most of your peers. You don't have to know all the latest startups and all the new models and you know, how to use quad code. Cause everybody's talking about this week. Like there's a good chance it's just not that relevant to you, but what is relevant is knowing what the models are capable of.
Forum3 (38:59)
.
.
Paul Roetzer (39:15)
having your go-to use cases that you know you can save time or drive innovation or growth with, and then empower your team to go find those same things for themselves. If you just do those things as a leader, I think you're gonna get pretty far.
Forum3 (39:28)
Great. So with that, let's do, I want to do one round robin. We're going to start with Adam. Adam, give a sentence, max two sentences of what you want CEOs to remember from this conversation in this episode. What stood out for you?
Adam Brotman (39:43)
I would pull up the quote from Paul at beginning of this episode when he was standing
figuratively speaking in front of the 50 Fortune 1000 CEOs in Las Vegas and he said Stop thinking of this as a tech Implementation and an ROI exercise and start thinking about this as a people change management future of work exercise
Forum3 (40:07)
awesome. Paul, how about you? What single sentence or two would you want CEOs to remember from this conversation?
Paul Roetzer (40:13)
Yeah, the thing I said in my Maycon keynote this year that I'll just read right here is optimization is 10 % thinking, innovation is 10x thinking. And if I'm the CEO, I would rather 10x thinking. I want to be looking at Gen.ai as an innovation and growth platform, not a reduce my people and do things faster platform.
Forum3 (40:31)
I'll answer the same question for me. AI first companies start and end with AI first leaders. I think I said that actually. So, so I'll, I'll quote myself. I mean, it's really for me. And I actually thought Paul, your answer, which is, I actually think that companies do need AI officers yesterday. they need, and they need councils, that are looking out at the horizon because the gap of
Paul Roetzer (40:38)
Hahaha ⁓
Forum3 (40:56)
the technological advancement and the human issues. What we're bucketing is change management, which is just human fear in the face of change. And at times, unwanted change and change that affects me as an employee. I think that the passive aggressive nature of humans in organizations and the political nature of that in humans, it's just gonna grow this year.
and I think it really, it's the job of the, not only is the job of the CEO to lead with AI and show that it can be used productively for fantastic results in the business, but it's also to deal with the, those emotional laggards of the human, the human condition, which are totally justifiable and understandable that we're going through this massive change moment in the world. And, and I think that's scary to people. So.
That the job of the AI officer is also to deal with those human, those human pieces. So onward into 26. Paul with that, ⁓ I want to thank the audience, but Paul, thank you for coming on. You've been a great friend of the firm. I'm booking it now. We're going to do this again next year. ⁓ So, and we'll continue to listen. thanks Paul with that. Thanks to the audience for listening to AI first with Adam and Andy.
Paul Roetzer (42:00)
All right, let's do it.
Forum3 (42:09)
For more resources on how to become AI first, you can visit our website, www.form3.com, download case studies, research briefing, executive summaries, and join our email list. We also invite you to join our AI first community, a curated hub and network for leaders turning AI hype into action. We truly believe you can't over invest in AI learning. Onward.