A

I sit in meetings all day long as a leader of our organization where I'm hearing things that I didn't tell people to do. I didn't put it in place like, go do this, go pursue this. Because we're empowering AI forward professionals. They every day are challenging the traditional way of doing things and say, well, what about this? Why can't we just build an app for that? Why can't I build a skill for this? And so like, once that happens, then all of a sudden you start looking at people and be like, you're not even doing the role you were hired for anymore. Like, you're literally functioning in this, like, horrible whole new role that we almost have to like, rewrite your job description. Welcome to AI Answers, a special Q and A series from the Artificial Intelligence Show. I'm Paul Raitzer, founder and CEO of SmartRx and marketing AI institute. Every time we host our live virtual events and online classes, we get dozens of great questions from business leaders and practitioners who are navigating this fast moving world of AI. But we never have enough time to get to all of them. So we created the AI Answers series to address more of these questions and share real time insights into the topics and challenges professionals like you are facing. Whether you're just starting your AI journey or already putting it to work in your organization, these are the practical insights, use cases and strategies you need to grow smarter. Let's explore AI together. Welcome to episode 220 of the Artificial Intelligence Show. I'm your host, Paul Raetzer. This is a special episode, Mike. We have a special guest today. This is rare. We don't usually get the another guest with us, but so Micah Put is with me as always. And we are joined today by Taylor Rady, who is our director of research at SmartRx. Welcome, Taylor.

B

Thank you. Yeah, longtime listener, first time guest. So there we go.

A

So Taylor, Mike and I worked together for years at my agency. And then Taylor went off and built a wonderful agency of her own. And then we were lucky enough to have her rejoin us on the team in January of this year. And she's now heading up research. She works closely with Mike, our studio, creating a lot of the content, the curriculum, the research reports, the blueprints. So Taylor actually led the charge on our state of AI for business research, which we just published. Did that come out in May, guys? Is that when that came out? Okay, Y. So that came out in May. And so what we're doing with today's episode, this is part of our AI Answers series. So this is in addition to our weekly podcast. This is our 19th episode of the AI Answer Series. And so what we're going to do is a kind of a unique format. Taylor's going to actually walk us through 10 of the key findings from the State of AI for business research, which you can actually download at StateOfBusiness. AI, you can go and grab this report yourself, but she's going to walk us through key findings and then Mike is going to lead kind of a Q and A and discussion with Taylor and myself just around some reactions to those findings and some of the information we've been hearing from people who've been kind of reading and digesting the information. We've actually seen some cool spinoff content from this research. People have sort of taken it and run with it and done some of their own things. So that's going to be the format today. It's going to be kind of a top 10 things to take away. You're going to have access to the report immediately if you can go download it, throw it in NotebookLM, have a conversation with it if you want, and then we're going to go through the Q and A. So this episode is brought to us by AI for Business Bootcamp by SmartRx, which is coming to Columbus, Ohio on July 16th. This is a single day event from 8:30am to 5:30pm at the Hilton Columbus at Easton. And it's built for professionals and leaders who are ready to accelerate AI adoption and value creation. The day kicks off with a keynote that I'm going to do on the state of AI for business, kind of where we are, where it's going. Part of that will infuse some of the research we're going to talk about today. And then we're going to transition into two highly interactive workshops. Mike is going to kick off the day with AI Productivity Workshop that's going to be all about use cases, workflows, optimization of existing approaches, and then I'm going to focus on AI innovation. How do we accelerate growth and innovation through AI technology in the afternoon? So you're going to get, you know, real AI powered workflows, strategic frameworks that you can use to accelerate transformation, and you're going to leave with an immediately actionable plan for yourself and your teams. AI Academy members do get discounted pricing, so be sure to take advantage of that if you're an existing AI Academy member. And then there's also discounts available for teams of two or more and groups of 10 or more, which you actually had some conversations with people who are looking to bring 10 or more already, there's custom pricing of available so be sure to reach out to us. And then finally you can use Pod100 to take $100 off of your AI for Business boot camp ticket. So again, that is happening July 16th in Columbus. You can go to SmarterX AI and click on events to learn more about that. Okay. Taylor, you're going to be there too, right?

B

Yes, I will.

A

Yeah, you're going to be there all day as well. All right, so I'm going to turn it over to Taylor and she is going to kind of run us through all about the report and then go through the findings. Then Mike and I will jump in and we'll go through the Q and A.

B

All right, sounds good, Paul. So yeah, today we are talking about the 2026 State of AI for Business report released just last month by SmartRx. This is our sixth annual report, but for five years we actually surveyed the state of marketing AI. And this year we decided to expand it with the goal of capturing a much broader and more diverse picture of what's happening in the workplace. So as for respondents, you can see the full breakdown of methodology in the report, like Paul said, but a few highlights. Again, this is the largest respondent base in the survey's history. So 2,109 professionals took the survey and really that spans every function, every industry, every company size is represented in this year's data from small firms, up to billion dollar plus enterprises. Now, 48% of respondents are senior leaders, so CEOs, founders, presidents, C suite VPs and 80% are involved in AI purchasing decisions as decision makers themselves or as champions or influencers. So we're hearing a lot in this data about people who are responsible for AI adoption and for AI tech decisions in their companies. As far as geography, people always ask about that. Respondents predominantly live in the United States, the United Kingdom, Canada, Australia and Germany. About 82% of the respondents were in the United States. And then this was really cool to see the completion rate. 87% of respondents answered every single question and the vast majority answered most. So we have some really comprehensive and in depth data to go through today.

A

Taylor, the one thing I'll throw in there, and you might address this a little bit later on, but that completion rate is even more impressive when you realize there was qualitative answers as well that people actually took the time to fill in essay type questions, not just multiple choice, which really again, thank you to everyone who took the time to be a part of this. It's incredible.

B

Yeah, really incredible to see. We got thousands of. Yeah, write in responses. So that was. That was awesome. And then the last thing as far as the responses and the methodology is that the survey was largely shared and promoted through our owned channels for Smarter X and the Marketing AI Institute brands. So this audience likely does skew slightly more AI forward than the broader workforce. All research, as I know you guys talk about on the podcast, has some form of bias and that is maybe this one. But we're going to talk about how this more AI forward leaning audience actually makes some of the findings that much more surprising and interesting. So now I'll just run through the executive summary, which is Basically the top 10 key findings from the report, starting with the most notable we asked people what they think AI is going to do to jobs over the next three years and 71%, nearly three in four said AI will eliminate more jobs than it creates. Only 13% expect AI to be net job creating. The rest are unsure or expect things to roughly balance out. Second, AI is now essential to business success with near universal agreement. 74% of respondents say AI is critically important or very important to their success over the next 12 months, and 89% of CEOs and founders rate AI as critically or very important. Third, the biggest barriers to AI adoption aren't technical. They're human. A lack of education and training cited by 38% of respondents and a lack of awareness or understanding 35% remain the most common barriers to AI adoption. Those have been top answers in the past years as well. And interestingly, a lack of time, a new response option this year landed immediately among the top barriers. It was selected by 30% alongside fear or mistrust of AI at 29%. So between training, understanding, time, fear, basically the top AI adoption barriers share this common thread of Professionals are struggling to keep up with the pace of change, something we also saw quite a bit in our qualitative responses. Fourth, more than half of professionals have moved past experimenting with AI. We asked professionals how they would classify their understanding and adoption of AI today. And at SmartRx we break that down into five stages of curiosity, understanding, experimentation, integration, and transformation. So we would consider integration and transformation to be advanced users, and 53% of respondents say they're in that camp. They are in the integration or transformation phases of AI adoption, meaning that they have moved past testing tools into embedding AI in their workflows or reimagining how they work entirely. And only 12% consider themselves those more beginner stage AI users. Fifth, we also found organizations are falling behind their own employees. We also asked which stage of AI transformation best describes the organization understanding, piloting and scaling being the three major categories. And almost half of respondents, 47%, say their organization is still just piloting AI, and only one in four companies are scaling AI. Sixth, nearly half the workforce is not yet sold on AI. When asked how they personally feel about AI's impact on careers, business and society, 52% of respondents describe their overall sentiment toward AI as positive, but 48% are neutral, negative or unsure. Seventh, only 13% of organizations have the governance foundations to scale AI. So specifically we mean that 29%, only 29% have an AI roadmap, 39% have an AI council, 48% have generative AI policies, and 48% have AI ethics policies. So again, just 13% of respondents say they have all four of those in place, and a third have none of them at all. Eighth, AI training is increasing, but the majority of respondents still lack it. 32% say no training exists, 18% say it's still in development, and 3% aren't sure. So 53% effectively don't have access to corporate AI training. Ninth, as for tools and tech, ChatGPT dominates small firms. Copilot dominates the enterprise, not surprisingly. So. Overall, 59% of respondents say their organization provides them access to a ChatGPT license, but from there, platform preference definitely depends on company size. So 73% use ChatGPT at small firms. Those are up to 1 million in revenue. And 73% use Microsoft Copilot at large enterprises. Those are a billion plus. And lastly, CEOs and founders report being dramatically ahead of everyone else in their own personal AI adoption. So 65% of CEOs, founders, presidents are in the again those integration or transformation phases compared to 53% of directors and 48% of managers. So again, those are just the top 10 findings from this year's report. You can, you know, see this and much, much more in the report itself. And we'll now, you know, dive into a lot of those numbers and contextualize them and talk about what they really mean.

C

All right, Taylor, thank you for that awesome roundup of the key findings from the data. There is tons more in this report. It has been such a joy to read through all this awesome data. Super excited to talk with you and Paul about some of the findings here. But before we get into the findings, I want to ask, did you use AI at all to help create this report? And if so, how did you do that?

B

Yes, very much so. Yeah, we used AI a ton throughout this process. In fact, I did a whole talk at our AI for Writers summit a few weeks ago, going through the entire process and yeah, so for the bulk of the analysis, a lot of the calculations, the pivot tables, the cross cuts we used at the time it was Claude Opus 4.6 and then Opus 4.7. Once that dropped, we used Claude code a ton, especially in the verification stage. So proofreading the final PDF, fact checking numbers, making sure percentages were consistent. The report is just shy of 50 pages, so there were lots and lots of numbers and making sure that there weren't any inconsistencies that creeped in. We also used Gemini in Google Sheets a ton for building pivot tables that basically let me query the whole data set. We used NotebookLM and now we're actually using cloud design too, which we just started talking about using that for creating more charts and graphs, kind of repurposing content from the report.

A

So Taylor, let me throw a follow up question real quick. So since you joined us in January, one of the things we've had you focusing on is how is AI reinventing the role of research? And we're not talking about like research that goes into AI model development, we're talking about research that goes into content creation and information dissemination. You just shared some of the ways you're doing with what to you is remaining uniquely human. Like what, what about the report? Like where did you as the human bring the most value to the process? Or like what do you see is still very uniquely needing to be the human in the loop to do things like this?

B

I mean, I, I certainly think that asking the right questions is first and foremost. I mean obviously AI helps us to surface even more ways we can explore it, but I think we're the ones that are ultimately initiating the research. So we have some in questions we are inherently bringing to the process and making sure that we approach the research in the right way in order to uncover the answers that we're trying to look for now. Again, AI also helped a lot actually in the survey part of the process and making sure that. I was thinking through question wording and things like that, how those would ultimately shape the answers we would get. But I feel like that's a big piece of it also in some cases the how substantial, like that kind of correlation causation sort of a thing where you can read through and be like, well I, I think these are more correlation. I don't know that this is causing that And I think some of that piece of it, I certainly am digging into the numbers and trying to figure out, you know, what. What they ultimately mean. But a lot of the middle part is AI can help with a lot of it.

A

A lot of it comes down to that idea of taste and judgment that Mike, you and I talk on the podcast all the time. It's just knowing what to ask, knowing what to do with the answers, making the connections, the AI pushing the AI in different directions and, you know, asking those. And it is. It's becoming more and more. That's the role of the human. And a lot of that heavy lifting, repetitive stuff that Mike and I used to spend hundreds of hours doing when we would do this report every year. That's where the AI excels at. It's like, okay, let's the AI do it. We'll just verify the outputs and then we'll put the whole story together and figure out how to distribute that story and figure out how to connect it to each of the part. I love one of the things you're doing, and I'm sorry, if we get this little round, one of the things you're doing is being able to connect this to marketing team, the sales team, the CS team. Like, what do we do? How do we activate this research? And so that's the kind of stuff where we're still truly orchestrating everything, even though AI is playing a much bigger role in the development of the research we're doing 100%.

B

And that's why it's also so fun, because, yeah, I mean, the exciting part of the job is not building the pivot tables. The exciting part is asking questions and getting to explore the data. And so I get to just do so much more of that. So, yeah, that's incredible.

C

All right, so let's get into some of these findings here and what they mean. And again, you know, we got tons and tons of questions from folks during both the webinar we did during this, where we answered during the launch webinar, a fair amount of questions, but also tons that we couldn't get to. So that's where we've kind of sourced a lot of these from, either by combining a bunch of similar questions or just taking some straight from our audience. So the big one here is really the headline finding, right, where 71% of people think more jobs will be eliminated by AI than created by AI in the next three years. I'm curious, Taylor, if you see. Do you see in the responses any difference in the sentiment here between things like Big or small companies, industry, Any other cohort within the data that thinks differently about this or is it pretty uniform how pessimistic people are about this?

B

Yeah, so that I think was the most interesting thing about that finding. Not just how high that number is, but how uniform it was. It was remarkably consistent through really every way that you wanted to kind of splice the data. I mean, we looked at, you know, company size, we looked at industry overall, roughly 7 functions, roughly 7 in 10 respondents expect AI to eliminate more jobs than it creates. So I pulled some of the numbers here just for us to kind of talk about. But as far as how many believe more jobs will be eliminated than created company size, it ranges from 69% at large companies to 74% at enterprises. So again, not a huge swing there. Small and midsize fall right in the middle at 71% industry. Again, it really just ranges from media and entertainment. 69% believe AI is going to cut jobs. Finance, 73%. And again, really not a meaningful difference by function as well. So, yeah, that was, I think what was really interesting. Now what we have seen is this number go up pretty significantly year over year. That was probably the most notable. But yeah, the breakdowns themselves, it's pretty much seven out of 10 people, no matter which way you slice it.

C

I mean, that was just like I knew we were headed that direction, but it still was very. My jaw dropped when I saw that, how high that percentage was. So with this though, so an interesting kind of tension or disconnect here is that 71% overall think more jobs will be eliminated. However, we also asked if people were concerned about their own job, about AI's impact on their own role, and just 20% said say they were seriously concerned about AI's impact on their own work. I'm curious what you make of that disconnect.

B

Yeah, so this was again just to kind of talk about the process here. We used a cloud project to produce the report. And so I have that project still and I dropped the report in there. Plus I have all of our original data. So I was able to like, for this I can jump back in and ask these follow up questions and drill into the data a lot more quickly. And so I looked specifically out of curiosity at the stage of AI adoption and how concerned they are about their job. So basically, are more advanced users more confident? Is that potentially where that disconnect is coming from as they're not super concerned about themselves because they feel pretty safe? And that is generally true. That basically when you look at who's the most concerned and who's the least concerned? You definitely see that as individuals become more advanced in their AI usage, more and more they say they're not concerned at all. So I think that, that, you know, two things can be true at once. I mean, certainly you could have people who are kind of blissfully ignorant of how vulnerable they could be to AI disruption and maybe with more beginner level, that's where they are advanced users. I mean again, you clearly see as they become more advanced in their usage, they are more likely to say that they are feel safe. So it might be that they're confident that they're safe, but they think that maybe their peers are not. And that's where the, the job disruption is going to come from, is from everybody else.

A

I have yet to meet someone who uses AI as much as the three of us do that doesn't worry about other people's jobs who don't use it like that. That was as soon as I saw this data that was my assumption is like, well, these are AI forward people who get it. They realize that their peers who aren't using AI yet in an advanced way may be at far greater risk and they're realizing, well, I can do the work of like three people right now. Like we're just not going to need as many people in marketing or as many people in sales. And so that was always how I looked at this data, was like, well yeah, it's a more AI forward group. The bias is going to be other people's jobs are going to be at risk. But I'm doing everything I'm supposed to do, so I'm feeling pretty good about my prospects right now.

B

Yeah.

C

You know, also one issue here, Paul, you and I have discussed this is you start to see it really starkly I think when it comes to any business that hires or works with third party service providers, like I've said three times in the last week due to experiments we're running. Wow. We're probably not going to need to hire a freelancer or a contractor or an agency for that specific thing. Doesn't mean we wouldn't work with them in plenty of other contexts. But even beyond the hiring picture, it's like there are just things that we do not need to spend money to hire someone to do because of what the capabilities, especially the more advanced ones enable.

A

Well, yeah, and as you get more into the agentic stuff, I'm sure your mind's working this way, Mike, all the time. Taylor, you do with the things you were Both working on, you're just like, I'll just spin up an agent tonight while I'm like at dinner and like I'll just do that thing I would have hired somebody for otherwise.

C

Right, right.

B

And Paul, to your point about the people who are most, I guess, aware of AI's capabilities. So interestingly one of the things with functions finance and software engineering and IT were the ones most concerned about their jobs.

A

Okay.

B

Of all the different functions, finance, 26% concerned, 20% very concerned. And software engineering, 30% concern, 15% very concerned. So I just think, interesting because when you think about, again, you guys talk about coding a lot on here, it's like those are the people who are really seeing what is possible. And I guess as they're seeing, you know, what AI can do, they're more and more thinking they're feeling less secure in where they are.

A

Yeah, yeah, that's pretty wild too because if you're at 20% overall, but then those two specific roles or industries you're, you're closing in at close to 50% concerned or very concerned. That's a, those are outliers. Like, that's a pretty significant break.

B

Yeah.

C

So I'm curious about some of this data about the top barriers to adoption of AI. So we had talked about, Taylor, you'd kind of mentioned some of the top barriers here being lack of education and training, lack of awareness or understanding, lack of time and fear or mistrust of AI. Those round out the top four. But we got a great question specific to how barriers change depending on where an organization is at. Someone asked, you know, what are the biggest barriers that companies specifically face as they go from piloting to scaling AI? And do you maybe want to comment on some effective ways to think about or overcome those barriers?

B

Yeah, so I mean we, we didn't ask specifically about, you know, company wide barriers, but basically I would argue that it's basically the same as, you know, the barriers to AI adoption. We're talking about generally spanning adoption across the organization. And one of the things that we looked at was, yeah, how those barriers break out based on those organization wide stages of understanding, piloting and scaling. And you know, what we found was a really clear pattern that, you know, when you are at the understanding the earliest stages of company wide AI adoption, awareness and education are by and far the biggest barriers you face, which makes sense. They basically don't know what they don't know. They realize they have a lot to understand and to learn. Piloting, they flip. But it becomes education, awareness and then Time are the top three barriers that they face. And then when you hit scaling, education drops down. Time is by far the biggest barrier to AI adoption. 42% upskilling organizations say that time is their biggest barrier. So it's just interesting to see overall the pattern and I mean you can see it makes sense I guess in terms of the overall progression that you have to start with education and training, you have to first onboard your team to build that AI literacy. But then from there I think where organizations start to fall is they don't protect the time that it takes to actually apply AI because this isn't just like a piece of software that you onboard people one time. This is something that requires real change management and rethinking workflows. And so I think organizations are struggling to carve out the time to do that and to think about the application pieces. So I think those are some things we're thinking about and I do think it's really helpful to think about it in the context of that overall, I guess life cycle of your organization, which stage you're at, what you might need, what solutions might be best for you at that moment in time.

C

Yeah, I, the lack of time just jumps out to me so often just in the talks I give the people we teach and speak to and counsel and everyone is trying to say I'm going to carve out time, I'm going to block off my calendar for AI every week or whatever. And you know, sometimes it works but I, I just really also counsel people. It's time to just start figuring out how you can be building and iterating as you go. Like what is the next thing on your to do list? Great, don't care what it is, focus on applying AI to that. Because if you do not achieve these standard out of the gate productivity gains using AI, it's going to be really hard to do more complicated stuff. You're just never going to have the time unless you're already figuring out how AI can transform your work. I mean AI is the solution to this problem.

B

It feels like, yeah, absolutely, yeah. One of the other things that came up was so one of the qualitative, the open ended questions we asked was what is your biggest struggle with AI right now? And so it was open ended write in responses. But then we used AI to basically find keyword matching and identify the top themes. The top thing by far, pace of change, finding time to learn, and you can see anecdotally when you read them as well, is it's finding time to keep up Finding time to learn. There's so much is happening so fast and so you can really hear it come through that. That is the thing that people, even advanced users, they feel like they're not doing enough, they're not learning fast enough, they're not adapting fast enough. And so to your point, finding ways to carve out that time, even in small ways, I think is, is essential.

A

A really good example of that just from yesterday for me was. So we're recording this on June 10th. This was the day after Fable 5 came out. Claude. Fable 5. I'm in meetings all day. Tuesday, June 9th, I'm getting text messages from people. Did you try it yet? I'm like, I don't even. I mean I've just been in meeting for two hours and like the world changed and like everyone's experimenting with Fable 5 and like, oh my gosh, it's a leap forwarded model, which it is. And like, what does it mean? And so I find myself, I get home that night and I'm just like scanning X and I'm trying to like get all the perspectives, I'm trying to understand the moment, like, what does this fundamentally change for us? And I need, I need time. Like I literally this afternoon after we record this, I have a clawed planning block for four hours on my calendar just to try and wrap my head around everything anthropic has released in the last 30 days and what it means to our organization. And it's just like. So I would say we are squarely in the scaling phase. And I feel the time restraint every day. Like I just can't do enough to keep up.

C

Yeah, that's so true. Yesterday was a very good illustration of this. Okay, so Taylor, the report found you had mentioned that only 13% of organizations have all four foundational pieces of the AI governance in place. So we kind of define those as AI councils, AI roadmaps, generative AI policies, and AI ethics policies. We asked about each of them individually, found that only 13% have all four. Can you maybe unpack this data a little more? Are there any commonalities between the organizations that have all these? Maybe dive into this a little deeper for us?

B

Yeah, for sure. So overall, organizations are most likely to have policies. So a generative AI policy or an AI ethics policy. And they are least likely to have an AI roadmap, which in itself is kind of. And then AI council falls in the middle, which is interesting because the policies are the most reactive and they're telling people what not to do more. So whereas the roadmap is really having priorities and a vision and kind of a strategy in place. So I thought that was interesting. But when you look at who is most and least likely to have this kind of governance, not a ton of meaningful difference by industry and even by size and revenue. I mean, it makes sense. Enterprises are slightly more likely to have governance in place than small businesses, but there isn't a huge difference there. What makes a huge difference is again, their stage of AI adoption. So the stage of AI deployment, understanding, piloting, scaling, that is a really strong predictor of governance. And organizations that are scaling AI are 8.6 times more likely than understanding stage organizations to have all four of those foundations in place. I actually pulled a few numbers specific to on average, how many of just any of those four pillars that they have. So at understanding, on average, they have 0.9 pillars, piloting 1.7 and scaling 2.3. So you can see a really clear linear. You know, the further you are, the more of these pieces you have in place. We also asked a separate question about momentum. So that was a new question of basically not just how far you are, but, you know, that kind of trough of disillusionment. Have you somehow stalled out? Do you have really inconsistent or siloed progress? And again, we saw a really clear correlation here. So if you have really stalled progress, you maybe only have one piece in place, whereas if you are steady or accelerating your progress of AI adoption, you have 1.8 or 2.1. So you have a lot more pieces in place now. I think the question then is it's kind of chicken and egg, like, are these organizations just really getting a lot of things done? They're rolling out AI, they're putting all these, their counsel in place and their policies. Or is governance the thing that allows these teams to have the kind of guidance and the guardrails to really experiment and expand their AI use? So I'm not sure on that.

C

But, you know, that is really interesting because that literally plays into the next question someone asked is like, how are you thinking about the best way to put governance in place while not slowing the company down? It sounds like that those might not be as at odds as someone might think. It sounds like, is that kind of what you're getting at?

B

I mean, yeah, that's kind of what we saw. Again, it's a little bit of, is it correlation? Is it causation? But you could certainly make a case. And I know that anecdotally you've heard that, you know, once you have these policies in place, People feel like they have the freedom to know what they can and can't do. And also certainly I think having an AI council, I know you guys have talked about this a lot as a really important piece to get all the other policies in place, to get training in place, to start prioritizing use cases and things like that. So I certainly could see how people think of governance as stifling progress. But there seems to be a correlation here that it might actually help unlock more momentum and more AI adoption across the organization.

C

Yeah, it's also worth remembering just the very big differences between company sizes and their needs. Right. Like, I've spoken over the past week with a couple different enterprises. It's not a question of these enterprise employees at say, like a big healthcare or manufacturing concern racing ahead and doing all this crazy experimentation, then getting curtailed by guardrails. Like, they're literally not allowed to go do things unless they know what they're able to do and how they're able to do it. So I definitely could see that not having as much of an effect. It can actually accelerate progress.

B

Yeah, absolutely.

C

So we asked a question this year about which AI topics people wanted the most training on, and 51%, one of the top responses said that they wanted training on, quote, using AI agents in my work. Now, this is not surprising. Agents are a huge deal. Do you have any recommendations for best practices to consider or at least things to think about? I'm curious, kind of both of you, what you would say to this around agent governance? That's still very early, but agents create a whole mess of other problems that simple chat AI chatbots do not.

B

Yeah, I mean, yeah, I mean, we're definitely figuring this out ourselves. So I feel like this is like something we are learning in real time. I mean, I think the thing I think of is that part of the messy things about agents is that there's such a lack of consensus around the definition because it really can be kind of a spectrum. It's like you're talking about degrees of autonomy and access to tools. So I feel like at least one piece of it can be. I mean, again, having governments in place is really important of any kind to make sure that people are not just going out and connecting things. But I feel like you can have degrees about. Like, a deep research report is a pretty safe place to start experimenting and understanding what agentic AI can look like. But it's just going out to the Internet. It's not connecting to, you know, a lot of tools and things like that. And so I Feel like you can start to train your team and put get comfortable with what kinds of governance and what kind of questions you need to answer before you start opening up access to more autonomous agents and more access to like different tools and systems in your company.

C

Yeah. Paul, I'm curious. I know we've talked about some agent horror stories the last several weeks on the podcast. Like, how are you thinking about governance of agents as a leader in the

A

simplest form just has to be addressed in the AI policies. So, you know, we've talked for years about generative AI policies and guiding the use of AI assistants and chatbots. But to Taylor's point, with agents, they really become valuable when they have the context of your knowledge base and they're connected to different data sources. And so that requires governance. It's like what is allowed to be connected where and what are we allowed to do with that information? So one of the things with Fable 5 that came up is the retention policy of the data that Anthropic changed as part of it, they're going to now keep even if you're an enterprise customer, they're going to keep your data for 30 days. Well, that changes the dynamics. That's a, that's a no go for a lot of companies that have assumed Anthropic wasn't keeping their data because that was the terms of use up until June 9th. So it's a very dynamic environment. But you have to establish these policies and then you're probably going to need to be doing at least a third 30 day, every 30 day audit of. Has anything changed technologically with the labs that we're giving access to our data? That should adjust the way we're running our policies. So yeah, I don't know. It's other than a multidisciplinary, like it legal, the departments that you're functioning within, like marketing, sales, like those leaders are going to have to work together to continually evolve those policies. And then there's the whole change management and impact on talent and you're turning agents loose that all of a sudden are doing the work of people and they're like not understanding, wait a second, this thing's doing what I was doing. What does this mean to me? There's a lot of layers to it, but the basics of governance is define policies and then have a system in place to regularly audit and update them and then communicate those updates to your people and why you're making the updates you're making.

C

Okay, before we dive into the second half of the questions, we've got for you all, a quick announcement. This episode is also brought to us by AI Academy by Smarter X and AI Academy helps individuals and businesses accelerate their AI literacy and transformation through personalized learning journeys and an AI powered learning platform. We add educational content to AI Academy literally every week, so you will always stay up to date with the latest AI trends and technologies. This particular episode is brought to us specifically by our AI for Industries collection which features eight core series and certificates designed to jumpstart AI understanding and adoption. And in this roster so far we have AI for professional services, AI for healthcare, AI for software and technology, AI for insurance, AI for financial services, for retail, for manufacturing, and our latest installment, which is quite new is AI for Education, which Taylor, I believe you were the instructor of. So thank you for doing that. So these series are an ideal launch pad for organizations that want to level up their teams and accelerate AI adoption and impact. We have individual and business account plans available. You can also buy single courses and series for one time fees. So go ahead and go to Academy SmartRx AI to learn more. You can use the code POD100 for a hundred dollars off any individual plan. Okay, let's dive into some more questions here. So I'm curious Taylor, have you heard organizations talking more about how roles are changing or been seeing more of this? Because really at a broad level. And Paul, I'd love to get your take on this too. Like when one person can do something end to end with AI's help, kind of what Paul just alluded to with agents. Do we expect traditional role definitions to change?

B

Yeah, I mean, I would think so. And yeah, I know Paul's been thinking about this quite a bit. I mean one way I think about this, even just without even changing a job title, 26% of the survey respondents classified their adoption of AI as transformative as, you know, rethinking their role in a fundamentally different way. So even without changing their job title, I mean the way that they are doing their job by their own definition has completely changed. So that has already changed right now before you even touch like an org chart. Another thing that I think is interesting, one of the things that we expanded this year as far as the roles that we asked, 22% of people who own adoption and integration of AI technology were dedicated AI leadership. So we're also again seeing an emergence of new roles, Chief AI officer, head of AI and things like that. So that's another way in which roles and org charts are definitely changing and evolving because of AI.

A

Yeah, I would say just like from our perspective. At SmartRx, we're very intentional, obviously about being an AI forward organization. There's a clear vision from leadership of where we're going and the expectation that we are going to build a smarter, more efficient, more innovative organization from the ground up. And so some of that is intentional, literally in the job descriptions. I mean, Mike and Taylor can attest that within job descriptions there is requirements around your work with, with AI, your work with agents, your plan, you know, being involved in the planning and production and utilization of agents in your job. And so there's things that are very intentional. We obviously provide AI Academy training. Like our team has full access to AI Academy. We talk about this stuff all day long. We have an AI tech channel within Zoom that we share things on all day long. So there's very intentional things, but then there's just the organic role changing. Taylor and Mike are two great examples of this. Like, I sit in meetings all day long as a leader of our organization where I'm hearing things that I didn't tell people to do. I didn't put it in place like, go do this, go pursue this. Because we're empowering AI Forward professionals. They every day are challenging the traditional way of doing things and saying, well, what about this? Why can't we just build an app for that? Why can't I build a skill for this? And so like once that happens, then all of a sudden you start looking at people like you're not even doing the role you were hired for anymore. Like, you're literally functioning in this like whole new role that we almost have to like rewrite your job description. And I do think that a lot of the role changing is going to be organic in that way. You're going to train AI Forward people and they're going to be the ones that figure out what the next evolution of their role looks like. And it might be an entirely different title, or it might just be that the role of a content marketer is different, the role of a researcher is different, the role of a salesperson is different. And I don't know. And so I kind of actually like that. That's how we're approaching it, where you empower people to think differently about what they're doing and come to the table with like entirely new ideas and processes.

C

Yeah, that's such an exciting part of the day. To day as well, is like, it's not even just saying, okay, you're empowering a marketer or salesperson or whoever to go use AI. It's in any specific business if you spend enough time in the real world, do you realize, like, there's a hundred things that come up every day that are not at all a part of your job description or are unique to the business? You know, like, we have so many quirks, for instance, of our learning management system that we have to solve for that. There's not really. That's not specified in every single person's role. But now that people know AI can be their partner in solving these things, they're just adding so many skills to their toolkit. It's pretty impressive. All right, so next step. I'm curious when we. So we asked how people classify their own understanding and adoption of AI, and we also asked how they ranked their confidence in evaluating AI technology and Taylor CEOs, presidents, founders, far and away, most likely, out of all the different roles to say they were more advanced and mature in their AI adoption. When we did our launch webinar, we had several people asking essentially, like, okay, that's a really interesting stat. Are CEOs really this far ahead on AI, or are they overconfident in their abilities? How are you reading this data point?

B

Yeah, I mean, I definitely think there can be a few different things happening here. So overall we saw, yeah, as far as confidence in evaluating AI technology, for example, that was very clearly based on, I guess, like in the org chart. So the CEO, the founder, they were the most confident, then vp, then director, then manager, then down to like a specialist, which is probably a function of how much of a decision maker they are, how involved they are. Obviously, being more senior, more advanced in your career, some of that is going to kind of come with the territory. Also, our audience skews AI forward, so they might be more AI forward CEOs. We also had in that category of, like, CEOs and founders, if you are a very small business or even like a solopreneur, an entrepreneur, they would fall into that camp. So if you are in charge of everything, then you are also probably very heavily involved in AI. But I imagine that there are probably also, at the same time, a lot of CEOs and presidents of large companies that are not very AI forward and are maybe more confident in their understanding of AI than their team might agree with. So I think there's a lot of different ways that you could read into this, but I'd say those are a few different things worth keeping in mind.

C

Yeah, Paul, I'm curious about your read on that. You talk to a lot of CEOs.

A

I think that what Taylor said is accurate. My assumption here is that the majority of the CEOs executives that are taking this survey are listeners of the podcast. They're people who follow what we do, which means they're AI forward by nature, highly curious, probably experimenting a lot. And then I would say the reality, and I don't know if we have this data, Taylor, but we have a very strong contingent of larger businesses that participated in the survey. I would guess it's not a lot of CEOs of billion dollar plus companies. So the majority of the CEOs I would assume are probably more in the SMB market. And to Taylor's point, they have to be like if, if you're a CEO of a mid sized business, like be a hundred to 250 people, whatever, and you're not AI forward yet. Like the reality is your company's probably not doing so great or doesn't have great prospects for the near future. So I just think it's a combination of likely, it's more SMB CEOs that are probably taking the survey and that they are likely AI forward people by nature and so they're going to have that kind of confidence level in their understanding and adoption.

C

Someone asked, how are you handling institutional resistance to AI? And they gave some really interesting context here. They said, I'm the AI subject matter expert for the company I work for and I've been trying to increase adoption across departments, but I come across roadblocks either from poor communication around the topic of AI between managers and their direct reports or personal ethical concerns that people are using to qualify, outright refusal to use the technology. Taylor, how do you look at this topic? Do we have any kind of data or metrics around this?

B

So yeah, I think there's a lot of ways that you could think about this and a lot of potential. You know, like this person cited a number of reasons behind the resistance. One of the things that I dug into was who owns AI adoption within the company and how that might correlate to the momentum. So does ownership by certain people lead to better outcomes, more widespread adoption and more success there? And not surprisingly, overall dedicated AI leadership was the most likely to have better AI adoption and kind of momentum without the the company. And also then following behind that was cross functional AI councils. So if you have someone dedicated to it, that is likely going to lead to better, more consistent progress and momentum within the organization. Next up is executive leadership. So specifically, if we're saying positive momentum, 64% cite positive momentum. If they have dedicated AI leadership, 60% cite positive momentum. If they HAVE executive leadership owning AI at the bottom, technology leadership. So if it owns AI adoption, only 47% say that they have positive momentum. They are much more likely to say that they have this really inconsistent or siloed progress. So I think that really speaks to what Paul talked about in his opening letter of the report. That if this is something that is just handed off to, you know, legal or IT to solve, it doesn't really get rolled out in the right way. Whereas if there's a head of IT achieve head of AI, a chief AI officer or, you know, the CEO is owning it, you have that kind of top down authority to actually get things done.

A

Couple of notes real quick, Mike. I would say it's critical to get through all these obstacles that you have to lead with empathy. You have to understand where employees are at and that some of them don't like AI, Some of them are afraid it's going to take their jobs, some of them just don't have the time to learn it. But they want to. Like nobody wants to be irrelevant or obsolete in their career. So like you have to be able to meet people where they are. You have to have from a leadership perspective, a clear vision for the future of work and what that means to your business. Like if you are mandating AI literacy and competency over the next 12 months, say that if it's going to affect people's future employment at a company, say that, like give them the reason why you're enabling all these things through, you know, actionable guidance on technology use through AI education and training systems. And then think of it as a communications and change management program. This is not a technology thing. Like technology is the core that's going to enable it, but this is all about how you communicate it to the different people within your organization and then how you empower them through change management and how you're transparent about everything in that process. So the obstacles can't just be overcome with one thing or another. It is truly a comprehensive approach to transformation. And that's where most companies we talk to are really struggling, is they just treated it as a technology thing to be solved, that we're just going to buy some licenses and give people the Genai app tools and it'll all work itself out. And that is not the way to go.

C

Amen to that. Especially that communication piece of it. Paul, Like I'm biased just because we're, you know, communicators, writers, but I've just seen so many organizations struggle with that. It's like before you do anything Else, technology wise, bring someone in, have them talk to your organization like, or get your communications in order. It can go such a long way.

A

That's why we talked about up front. I think, like, marketing is often the tip of the spear when it comes to adoption within organizations. As your marketing and comms people, they've got to be able to work with HR and leadership to communicate this stuff to people internally. It's an internal communications plan. And if you're a big enterprise, you have an internal comms team, they better have high degrees of AI literacy. Like, you can't have the communications people who control the message, not understand the moment themselves. And so that's where we just, we see this lack of, like, we use this term a lot in the podcast. Lack of situational awareness of the true story scope of what has to happen. And I think everybody just wants the quick fix and it's just not the reality.

C

Yeah. One last thing I would just add that I see a lot is that people fail to appreciate that the narrative environment is very different around this than many other things. Like, if I go home at the end of the day and I haven't heard during work about our benefits plan, I'm not reading headlines about benefits plans. Like, I'm not getting inundated with hype and misinformation and doom and gloom about benefits plans. If you're not saying stuff about AI, your employees, your teams, they are hearing about it from somewhere else, and it's probably not great. Yeah, exactly. Right. So it's like, really important to understand. Okay, next question. How are you reconciling the challenge of teaching AI skills with folks who enter the conversation at such. With such wildly different diverse experiences and backgrounds? So I think Taylor's kind of comes from the fact that this data shows people are all over the place in terms of AI understanding, adoption, sophistication. We see this, all of us see this every day as speakers, instructors. It's crazy. You can go into a room of 100 people and it's almost. Sometimes, Taylor, I feel like we talk about this a lot. It's like they're in a hundred different places. Like how with AI, how do you reconcile all that?

B

Yeah, I mean, I. I think getting a baseline of where people are to your point is really important to figure out exactly who you're talking to and at what level you're talking to them. Yeah, I don't know. It's hard. I mean, it's something we talk about internally. It's like, how do you teach this when people are in such wildly different places. I mean, one thing that comes to mind because we just did our AI for Education course is the ability to provide, I guess, broader, more personalized learning and education to different people and meet them where they are, wherever they are in their AI journey and help them get to that next step is more possible than ever. So that's certainly, I think a piece of it is not having a one size fits all kind of training plan, but finding those ways where you can identify your more beginner, intermediate and advanced users, your power users, and start to meet people with different types of content and training that matches where they are currently in their understanding and use of AI.

C

Yeah, it's interesting. I also kind of flip that too. And we've talked a lot about, I think about like what doesn't change or what would. What would everyone need to learn regardless of level. Right. It's like regardless of how good or bad I am with AI, I already, I know for a fact I'm going to need to find use cases. How can I teach people to do that? You know, stuff like that. It's not the whole ball game, but you kind of either look at the hyper personalized side or look at the evergreen side in tandem. I feel like.

B

Yeah, the timeless stuff for sure.

A

Yeah, yeah.

C

Okay, for this next question, Paul, I'm going to throw this one at you. As I know we've talked about this more and more on the weekly podcast we someone asked. We are starting to get pressure from a handful of employees at this person's organization to measure AI usage and its impact on the environment. I don't know where this organization is based or what industry they're in, but they said the then they are under pressure to purchase certificates to offset our impact. The question is, are you hearing more concerns about AI's environmental impact and how are companies addressing it?

A

We get asked questions about it all the time. I mean, anytime I go do a keynote somewhere, it's one of the first questions I'll get asked whether it's water usage or energy usage, powering the data centers. Most people don't like the answer, which is there's just not a hell of a lot you can do at the moment. I mean the most for most people that are users of the technology, which is going to be mandated as part of your job. Like I've said this and I don't mean it to sound cold in any way at all. I'm like completely empathetic to this. It's why we have Karen Howe as a keynote at Macon this year is like, we're trying to actually raise awareness about the environmental impact. What I will tell people though, is if you don't use AI and you work in basically any industry, in any knowledge work role, you won't have a job in three to five years. So, like, not using. It's literally like saying, I refuse to use a computer because, like, computers negatively impact the environment. Okay, that's fine. Like, you can take that position, but you won't have a job that requires the use of a computer. And so I feel like there. I can't come up with a single knowledge work role where AI won't be infused into what you're doing either. In the hardware you use, the software you use, whatever it may be, it's just going to be there. So not using it because of your feelings toward the environment. It is a viable personal choice to make, but it is not a good career choice to make. And so then the question becomes, okay, but what can we do? So there are probably things that an organizational level, if you work in a bigger enterprise, there's probably things that the IT team can do specifically probably it to think about the most efficient usage of the technology. And then I think down to the individual level. The higher your AI literacy, the more you understand about the different models you can choose to use, the better you are at prompting, the better you are providing proper context. So fewer prompts are needed to get the output you need. Like, that is really the best way to minimize the impact on the environment at an individual level is to get really, really good at using it. So, like, as a really practical example, if I go into Anthropic Claude today and I have a project I want to work on, I can choose from any of what in episode 218, I listed them any of like 12 models. Like, they're all in the dropdown. If I know that Sonnet 4.6 or whatever the Sonnet current Sonnet version is will do the job, then I'm going to do that. I'm going to use Sonnet 4.6. I'm not going to try Fable 5 to help me write a strategic brief, like, I don't need it. So I'm going to use far fewer tokens. And then in the prompt I give it, I'm going to be really thoughtful upfront of like, what exactly I want. Here's all the context you need so that the AI can do its job as quickly as possible. That's it. Like to me that at a personal level, like, I've Used the analogy. I think I heard it from somebody else. It's like turning the lights off when you leave a room. Like, we all want to save electricity, but the reality is we all use electricity every day. So just don't waste it. Don't leave the water running in the sink. Like there's little. That's basically where we're at with AI is be a, be someone who understands how to use the tools as efficiently as possible so that you minimize the impact you're having on the environment. You maximize the value of the output you're creating.

B

I love that.

C

Okay, Paul, another question I want to throw your way first. What industries do you believe had the greatest opportunity for growth in the age of AI? Do you see anything in your experience or in your research that's most likely to decline?

A

This is one where I think it's very personal. Your answer to this is going to be very personal. I think every industry, you can look around and say there's ability to build a smarter version of a company. The name SmartRx, this is literally where it came from, came from is I, I, years ago, I was like, you could just build a smarter version of every department, every company. And so the X became the variable of what industry you want to do, what company type do you want to do? And so that's what it was meant to be. And so if you're an attorney or if you're in HR or if you're in sales, or if you're in manufacturing or like whatever it is you do, whatever industry you're in, there's an opportunity there for dramatic growth, to build an AI native version of a company in that industry, or to join a company that's trying to build an AI native version. And there's enormous total addressable markets in every industry to just infuse AI in responsible ways to that industry. So I don't know that I look around and pick like one specific one. We're obviously very bullish on media companies because first and foremost we are a media company. We create content and build an audience, events, education. Like I'm very, very bullish. We're trying to reimagine what research looks like, trying to make it more real time, more dynamic and like consulting services, advisory service. Like I'm very bullish on all those things, but that's, that's what we do. It's what I know. So I would say whatever you do, there's, there's growth potential in it.

C

Okay, Taylor, I'm curious about your perspective on this one. We've got a couple final questions here. Do you think that AI will progress past the barriers identified in the report? And in your opinion, what do you think is the biggest driver to enable that to happen?

B

So the thing that's interesting here, as far as AI progressing past the barriers, I mean, a lot of the barriers we identified weren't specific to things with the technology itself, the models itself, that needs to change or be solved. I mean, when we asked which are the top barriers to AI adoption, people could choose up to three. 15% chose budget, 10% chose tech failing to meet expectations, 9% lack of the right data. So that was really toward the bottom of things that felt like they were somehow the technology was inadequate. The top barriers were lack of education and training, lack of awareness or understanding, lack of time, fear, or mistrust of AI. So 29% of people picked that. That's really the only one that feels like it implies maybe a failure of the tech itself. Like, you could be fearful of hallucinations or a lack of trust in the outputs. But really, I mean, the barriers we're talking about are specific to people and building awareness, education, trust, you know, change management. I mean, that is really, I think, more of what needs to happen more so than the technology itself somehow needing to be different. So I think it kind of goes back to a lot of the things that we're talking about of communication and empathy and training and being there for your people at this moment in time, more so than, you know, Fable 5 solving all of our problems.

C

All right, so our final question here. I'd like to get both your perspectives. Maybe Taylor, you first, and then, Paul, you can close us out here. But from the report, what was the most surprising insight to you this year and why?

B

So there was another question that we asked this year about their company's top outcome. What was their company's top outcome with AI? We've asked this in the past. Every year it has been productivity. Everybody wants productivity. And I just had this feeling that I wanted to change the question because, I mean, you guys talk about this all the time. That, like, we see the writing on the wall with jobs, with. With headcount, with flat headcount. And so we changed the options for this question this year to specifically define productivity as increasing output with the same resources. And then we added decreasing operating costs either through reducing headcount or overhead. Now, my assumption was that, okay, productivity is not going to be the leader anymore. We're going to see some people saying that they're trying to reduce operating Costs with AI, that feels like it's at least somewhat directionally true. And I was totally wrong. Productivity was again at the top. Only 4% selected, reducing operating costs. It was one of the lowest selected options of anything that we put on the survey. So that alone was interesting. But then you also put it into the context of 71% believe that AI is going to cut jobs. So to me, it felt like this weird disconnect of, okay, so you do believe that jobs are going to be cut. That seems overwhelmingly true, or at least overwhelmingly believed to be true. But nobody really thinks it's going to happen at their company, and that's not what their intent and their goal is, but they think somebody else's goal is to do that. So there was just this weird disconnect of looking at these multiple questions together. And then you look at 20% are not concerned about their job, but 71% believe jobs. So there just feels like there's this weird moment where there's this. I don't know if it's cognitive dissonance or if it's just you. You believe that your company is different from how everybody else is going to act. And I thought that was a really interesting finding.

C

Paul, how about you?

A

You know, the jump in jobs, I think, is obviously a standout data. The perception of the negative impact on jobs. I'm not surprised by that, though. Like, I was actually surprised how low it was last year at 53%. So I wouldn't say that's surprising to me. The thing that, you know, every year we watch this data and I wait for these leaps, like in, you know, number of people that now have AI education and training, or the number of people that, you know, think that the impact on jobs is going to be negative. The one that jumped out to me is, I think the first time we did this was this year where we actually looked at the four factors of governance, and only 13% have all four of them. That's one to keep an eye on because I think that is so fundamental to properly scaling AI, treating it as a true transformation initiative, not just a technology project. And so I'll be really curious in future years to start watching that combined data point to see how many organizations really are doing all the necessary components to do AI in a responsible way.

C

Awesome. Well, just a quick reminder for everyone as we wrap up here. Go to StateOfBusiness AI to check out the full report. Tons, tons more in there. Taylor, incredible job on this this year. It's such a, such a privilege to work a little bit with you on this and just see the end result and course. Thank you for answering all these awesome questions from our audience. Paul, thank you for doing the same. Really appreciate it.

A

Yeah, thanks, Mike. And I think we may be seeing Taylor again on the podcast sometime soon, so for sure.

B

Happy to come back.

A

All right, thank you. Thanks for listening to AI answers. To keep learning, visit SmarterX AI, where you'll find on demand courses, upcoming classes, and practical resources to guide your AI journey. And if you've got a question for a future episode, we'd love to hear it. That's it for now. Continue exploring and keep asking great questions about AI.