Ep 782: How Smart Teams Stopped Prompting AI and Started Automating Workflows

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life.

When it comes to defining AI success at your organization, think, what's the metric that's most important? I think if you rewind maybe a year or a year and a half, a lot of people might look at things like, you utilization rate, or how many licenses of a certain AI system your company or your department has, or sharing prompt packs, right? But looking back at that old measurement of success now, it seems like it's a decade or more in the past. Because if I'm being honest, just because you have a, a seat for Copilot or ChatGPT or Claude or Gemini, doesn't necessarily mean that you are moving the bar. That's because I think we are well under the way of the transition from chatbot to autonomous agents in automating workflows without even having to be technical. So that's one of the big topics that we're gonna be tackling today on Everyday AI when we talk about how smart teams stopped prompting AI and instead are now starting to automate workflows, and we're gonna tell you exactly how it's done. I'm excited for today's show. I hope you are too. Welcome and what's going on? My name's Jordan Wilson. This is Everyday AI. This is your daily livestream podcast and free daily newsletter helping everyday business leaders like you and me not just keep up with all the changes in AI, cuz yeah, they happen daily, but I and our guests help you focus on what actually matters and how to move the needle to grow your company and your career. So if that's what you're trying to do, starts here with the unedited, unscripted livestream podcast. But to be the smartest person in AI, you might want to look at our newsletter, youreverydayai.com. We're going to be recapping the highlights from today's episode, and I already know it's going to be a good one. And we're going to give you all of the other AI news you need to know to keep up and stay ahead. So speaking of great resources, Section is one of them, a previous advertiser. And if you remember back a couple of years ago, we actually had the CEO on the show a little while ago as well. But today, This is going to be a good one. So please help me welcome livestream audience. We have Bobby Isaacson. There we go. We got him there. Hey, Bobby. He's the head of enterprise at Section. Bobby, thank you so much for joining the Everyday AI Show.

Yeah, Jordan, thanks for having me. Appreciate what you're doing for the space. Good to be here.

All right. So let's first— so everyone kind of understands what is Section and what is the work that you all do, because I know it's changed a lot even since, you know, we had Greg on the show a couple of years ago.

Yeah, yeah. So Section is an AI transformation partner for some of the world's best brands. Long story short, we help drive employee enablement, but also the change management and strategy so that AI is the transformative technology that many, most companies want it to be. Easier said than done.

Yeah, yeah. Ever, ever changing, right? And talk a little bit about what you do as head of enterprise. It sounds like maybe a little bit of everything, right? But what's your day-to-day look like before we, you know, tackle everyone else's workflows? Yeah.

Probably as you would expect in a 40-ish person company, a little bit of everything, but my primary focus is the go-to-market team. So sales, customer success, partnerships, really anything that touches enterprise customers and supporting them. Awesome.

So let's jump to the end and then let's rewind our way back through it. So, you know, when we talk about this shift going from, you know, hey, let's get everyone a seat of a certain AI and let's start using it, right? And now we fast forward and, you know, Some, I, I, I think the best people are barely even spending any time in there, right? Because they're just automating their workflows and they're just making the judgment calls in the backend. But describe this shift that you've seen over the last couple of years and, you know, kind of where we're at today when it comes to automating workflows. Yeah. Yeah.

Well, I think you hit the nail on the head, Jordan. The, the biggest shift is probably in terms of how users are using it and what good looks like in terms of moving from that prompting, querying, chatting with the LLMs to really building on top of it and automating different workflows. And I think one thing I just will mention, obviously agents are all the buzz right now and everybody wants to talk about agents. It's interesting that the organizations that we speak with, and I won't even, I won't name any of them by names, but oftentimes we get people that are asking for help with agents. They don't even understand what agents are, but it could be custom GPTs, it could be projects, and oftentimes it's agents. And I listened to your episode, I think it was yesterday, around what OpenAI is doing with workspace agents and really cool. But I think the main thing is if you are in the enterprise and you're using AI and you find yourself prompting every time and opening up a blank chat window, you're probably doing it wrong. You really need to take the time to think about how is work being done, what really drives results, and thinking about how AI can help scale that or automate it. And oftentimes that's going to be agents, but there's actually a lot of different ways that I can do that. That's not prompting. Yeah.

And I think, you know, you brought up a good point because I think that, you know, maybe back in 2023, 2024, you know, yes, it was all about the prompting. And then you brought up the concept of, you know, kind of not starting over from scratch, right? And maybe that kind of brought us more into the, you know, context engineering era, you know, to make sure that, you know, some of your most important, you know, personal, departmental, uh, company sector context, you know, follows you from chat to chat. So you're not always starting over, right? Maybe even help us for, you know, maybe those of our audience that aren't, you know, always staying up to date with the latest and greatest, you know, what's the importance of not having to quote unquote start over, you know, when you are more in that chatbot era?

The importance of not having to start over when you're working with your AI, do you mean?

Yeah, exactly.

Yeah, yeah. I think, um, everybody has the same word, but context is king, or whatever you want to call it. I mean, this— these, these technologies, whatever platforms or LLMs you're using, I mean, they obviously have so much power, so much knowledge. A big part of that is honing them in to make sure that they're working effectively and intelligently for you. And so taking that time upfront to provide the context, whether it's in projects or GPTs or in memory or however you want to do that depending on how you're working, is so critical, right? Making sure that AI knows your goals, what your desired outcomes are, your style of working. I think that's become so much more important and the platforms thankfully have made that easier by building these projects and these custom GPTs and now agents. That you or I with our natural language can go in there and kind of like architect them in a way. But it's much more effective, faster to get the results that you want if you give the context in advance. Yeah.

And I think you, you hit on a lot of the key categories there, especially when it comes to not having to start over from scratch, right? Using projects and GPTs and, you know, each kind of the major platforms have their own versions of, you know, sorry, I don't want to leave anybody out.

Yeah.

Yeah, yeah, exactly. Exactly. You know, But what's, even for me, right? So when you say things like projects and GPTs, right? Like I just ingrained it in my brain for so long, like, right? Always going to a project, always going to a GPT, et cetera. But now I find myself like looking at that way of thinking and it's like, okay, why am I still the one going in and, you know, going into that project and, you know, taking this transcript and putting it in there? When it comes to the muscle memory of doing these things that were, you know, best practices for quote unquote context engineering, and now it's almost like they're antiquated. How do you start addressing that, you know, where sometimes it takes organizations many months to even understand, oh, we should be using projects, and now it's like, okay, well, if you're spending too much time there, you might be missing out on more of the agentic or, you know, automating the workflow side.

Yeah. Yeah. Well, I think this, this, my answer would hit on a few key things and, and starts with it starts with the top leaders. People need to demonstrate that behavior to others. It's hard 'cause we all are stuck in, you know, we're all busy, right? If, and I think a lot of people would argue that as powerful and as impactful as AI has been in a lot of organizations, it's just made us busier because expectations have risen. Maybe resources have gone down in other areas. And so a lot of people, you can't blame them for kind of being stuck in their ways of doing work. And it's hard to take a beat and think about, okay, how should I change? How should I do things? Things differently, things like giving AI more context, that doesn't come naturally for a lot of people. So I think a lot of it starts with at the top as leaders, and, and our CEO does a really good job of this. I try to implement as well, but you need to share the vision and demonstrate by example. I think that's one of the things that we see in the enterprise a lot, are leaders who actually not only walk the walk— or sorry, not only talk the talk, but actually walk the walk themselves. That goes a long way of people saying, hey, my CEO or my direct manager or leadership, whatever, they're actually doing it this way. I should probably do it. This thing isn't going away. This isn't a fad. And one actual example of how we've done that recently is, you know, I love being in this space. Our customers tell us a ton when we're on calls with them about what they're trying to do, what their challenges are, what they've already tried to do with AI. We get a lot of good information from them during the sales process. And if you look at traditional sales, what most, you know, enterprise salespeople have done, you follow up, you send a deck, and you ask for a next call to do a demo or something like that. But our CEO had this vision of, hey, we're getting so much good information from prospects, we should be using AI to put that into a brief and show we were listening and actually deliver value earlier on in the conversation. So we've actually been building out an agent that takes call transcripts and actually gives them a brief of, hey, here's where we think you should go with your AI transformation. It actually makes like thoughtful recommendation 'cause we've trained that agent with our frameworks, et cetera. So I challenge myself, I'm like, I'm gonna do the first 5 and show my team this is how you do it. And I, you know, I take time, I block time on my calendar to make sure that I actually do it. And then I ask the team, you know, we train the team on how to do it. We ask them, and then the next week after we release that, hey, at our next team meeting, my expectation is everybody's gonna show up with 2 briefs that you've generated using that agent. And I think, long way of answering your question, it starts at the top. You need to show people how it's done, give them the, the time, the framework of how to do that. 'Cause it's not always obvious for, for people and we're busy. Yeah.

And I'm glad you brought up this top-down because speaking of that, right, I did, you know, talk to, you know, Greg a couple, like 2 years ago on the podcast. And I didn't even, like, I should have went back and relistened to it before this, but I'm like, so many things stuck out from what he said 2 years ago. And I distinctly remember him saying that, you know, when, you know, talking with the board, he would always tell the board, and he, you know, he told this story on, on that show, he would say, hey, before you come to me, you know, you should be using these GPTs or something like that first, right? What are some of those best practices, you know, for that top-down approach? Because I know some CEOs live and breathe it. Some don't, but they still want everyone else to. How do you address that top-down, you know, if that is the approach an organization wants to take? Because there's different ways you can do it, right? But maybe if the CEO's on the fence, right, how do you address that?

First of all, if the CEO's on the fence, if the CEO's not doing this themselves, I would say the very first thing I would say, get outta the way and let somebody in who's actually going to, to live and breathe this. Seriously, like, if you're a leader of an organization and if you're not walking that walk, whether it's because you're at a certain point of your career or however you feel, if you're not doing this yourself, make room for someone who will. You're doing a disservice for all of your people that are underneath you, that whose lives and careers depend on you. This is coming whether you like it or not, and I don't think that you should be at the helm of any organization if you're not truly embodying this. And, and I mean, it, you, you bring up a good point, Jordan. We see it all the time. We see leaders, oftentimes CEOs, but other leaders as well, say, oh, this is what we're gonna do. Oh, we're an AI-native company. And we've literally talked to CEOs who have said that. And then you get them behind closed doors and their favorite use case of, you know, using ChatGPT is researching restaurants to go to dinner that weekend. And it's, it's, it's not going to work. So then how do you make that then trickle down to the rest of the organization in a meaningful way? It's really, really hard. So I think the first thing is leaders need to actually walk that walk and, and, and it can't, they can't just talk the talk. And be specific. It's not enough to say, oh, we're gonna be an AI-powered company. If you're not using AI in 6 or 12 months, you won't be here anymore. Those are empty threats, you know? And I think it's an organization's responsibility to yes, you can cast that vision, but you need to help enable your people to get there. And, and one way that we contrast this is, is, um, most organizations are obviously saying this is important. This is number one, if not top 3 things on most CEOs' mind right now is how do I brace my organization and prepare for the, the AI, uh, era, which is already here. And most companies have a policy as it relates to AI. Um, but when you think about like a policy, a policy most is first reaction is most employees is, oh, that's what I shouldn't do with AI. And it almost discourages use. So we oftentimes ask organizations, do you have an AI manifesto? You know, most of them, no, what's that? It's different. Meaning an AI policy says what you can't do. A manifesto says, hey, here's what we want you to do. Here's what we encourage you to do. Here's what we expect you to do. And it's more specific and it probably even hinges on here's how some of our operating principles are gonna change. But it specifically should touch on where do we want to get to as an organization? What do we want to be doing with AI? What are our goals and how ambitious do, do we want to be? Do we think we're gonna need to be there in 6 months, 12 months, et cetera? But that's a big dichotomy of like, have you actually laid out the vision for what you want people doing versus are you just trying to crack the whip and, and force people into doing something without actually enabling them? So I think that's the really, the big thing is you need to put the structure in place. cast a clear vision where it's easy for employees to see what they should be doing and then enable them. So the bottoms-up piece is give them the ability to learn. This is a big change in people's lives and people's careers, and you shouldn't leave it for them to go figure that out. Enable them, train them on role-specific, how to use this, what use cases they should be doing, etc.

Speaking of enabling and training, my personal take is that process looked a lot different, you know, in the chatbot era to get people from zero to chatbot as an example. but now as, you know, autonomous, uh, you know, agents or automating workflows is now the new, you know, de facto. What does that set of challenges look like, you know, in your experience, you know, with Section working with, you know, all of these enterprises? Like, what do those challenges look like? And what are, you know, some of the most common ways that you're seeing organizations either overcoming those common challenges or failing?

Yeah. Yeah. Well, first of all, we would agree with your, with your vision, Jordan, like that has changed and, and we've seen that ourselves within Section. You know, 2 years ago we were doing this primarily through courses and videos and we were recording, you know, things like how to build a custom GPT or, you know, prompting fundamentals and people loved it. Uh, the challenge was people loved it, they were getting a lot of value, but we saw, hey, this doesn't scale. This isn't gonna be future ready because The power of AI is it's unlike any other technology. Take Salesforce or Workday or any of these systems as predecessors of, of big technology. Amazing technology, but you could kind of train people on the specific capabilities. And like, I'll use sales just 'cause that's what I'm most familiar with. I could train a seller to how do you create a new opportunity in Salesforce? How do you move it in stages? How do you update your forecast? Like those were trained capabilities that were relatively static. AI is so different for, for literally every different role. The key is how do you actually get someone in finance to use AI to close the books, not just summarize documents they didn't want to read or shorten, you know, emails they didn't want to take the time. And you can't train that through traditional methods. You can't create a training course or a workshop, A, because it never gets the personalization level. But as you well know, by the time you create a video and edit it and put out a course, The technology's changed. So we've just seen that's been such a massive shift is the, what's so important. We, we use AI to train on AI because only AI can keep pace and get to know what is Jordan, what are his responsibilities, how familiar is he already with AI? So therefore, what kind of training can we give him that's really gonna move the needle so that he's actually applying AI to his most valuable use cases?

AI moves too fast to follow, but you're expected to keep up. Otherwise, your career or company might lag behind while AI-native competitors leap ahead. But you don't have 10 hours a day to understand it all. That's what I do for you. But after 700+ episodes of Everyday AI, the most common questions I get is, where do I start? That's why we created the Start Here series, an ongoing podcast series of more than a dozen episodes you can listen to in order. It covers the AI basics for beginners and sharpens the skills of AI champions pushing their companies forward. In the ongoing series, we explain complex trends in simple language that you can turn into action. There's 3 ways to jump in. Number 1, go scroll back to the first one in episode 691. Number 2, tap the link in your show notes at any time for the Start Here series, or you can just go to starthereseries.com, which also gives you free access to our inner circle community where you can connect with other business leaders doing the same. The Start Here series will slow down the pace of AI so you can get ahead.

Yeah, and using AI to, to learn AI, right? Like, I think that's such a great point. It's something I bring up all the time. You know, people will, you know, ask me a question and I'll say, hey, I'll, I'll answer it, but this is probably something you might be better off asking an AI, especially when it has all the context about, you know, your role, your organization, your chat history, things that, you know, me as human can't understand. But, you know, one thing that's kind of related to that is, you know, as we are in this kind of, you know, ever-evolving, you know, not saying it's a new era, but I— it is, right? The autonomous agent, the automating workflows being the new norm. Have you seen that shift? The big problems, right? Shift from maybe originally they were, you know, technical, traditional digital transformation, and now are they just more it's hard for people to give up agency, right? Whereas before it's like, oh, well, only I was the one that knew the answer to that question. 'Cause you know, I'm Bill, I work in IT and I've been here for 30 years and the institutional knowledge is in my brain. But when that context is shared across an organization, that changes both what is possible and the bottlenecks, right? So what does that look like now?

Yeah, I would say yes and no, to be honest, Jordan. And, and our perspective comes from working, we work with some of the, the biggest and best brands out there ranging from, you know, Nike, Anheuser-Busch, Adobe, Autodesk, you name it. Um, so technology, uh, manufacturing, finance. I'd say pretty consistently now we're seeing fewer and fewer people that are saying, I'm not going to use AI, like they're, they're totally anti. Um, there's a desire to use it, but we do this data quite frequently, and I'm sure you see it. Less than 5% of the workforce is actually what we would consider proficient or expert at using AI for their job. The majority of the enterprise employees are still at what we would call either novice or experimenting. And the challenge with that is they may be using it quite frequently, as often as daily, but if you actually peel back the onion in terms of how they're using it, it's pretty superficial stuff. Either, you know, the amount of, uh, prompts that are one-offs, right, which kind of shows us, hey, that you're using it as a Google replacement. Um, uh, you know, people using it to write better emails and summarize documents, that's kind of the, the primary use still within the enterprise. So people here, I think agents and things have been marketed well and people want to do them, but people, most people still don't understand them, how to do them, right? So that's a lot of what we're seeing in the enterprise is we need to help people understand what these greater capabilities are, whether they're agents or otherwise. And then it still does come back to like, you still need to prompt to create most agents. You know, again, I watched one of your episodes recently You're using prompts to build these, um, OpenAI Workspace agents. So we do need to make sure that people have the foundation of AI and then help them think about in context. 'Cause if you try to teach everybody everything there is about to know about agents, they're gonna get overwhelmed and sometimes discouraged pretty quickly. So I think the key thing is helping them understand what are the problems that this can solve for you and what opportunity can it unlock. Yeah.

And, and I'm glad you brought up this concept of, you know, the actual, I, I've heard the term different, right? The elite users, the, you know, true AI champions within the organizations, right? Those that are really pushing the frontier for at least AI use within a company. So if we just say they're the elite users, right? How do you start closing that gap between everyone else versus those that are actually pushing the technology and using, you know, kind of the latest era of technology? I mean, is it an L&D problem? Is it a training problem? Is it a use case? Like, you know, how do you, you know, as a CEO or as a, you know, head of, you know, business, right? Or sales, like whatever leader is listening to this and they see that, they're like, yeah, we definitely have our elite users and then we have everyone else. How do you like, what are the best ways that you've seen to close that gap?

Yeah, yeah. And it's funny because for whatever reason, maybe they're the noisiest or whatever, but a lot of the companies, they know that they have pockets of advanced usage. And a lot of times companies come to us and say, oh, well, we need this, this, this because our employees are doing that. And then when you actually ask them, it's like, okay, that's like one, again, 1 to 5% of your workforce. The bigger opportunity actually lies in bringing the rest along. And Jordan, as you can imagine, it's probably, you know, it's not one thing. It's not like, oh, let's just train everybody. I think you mentioned a lot of it is helping people understand this contextually. What are their use cases going to be? I go back to people in finance. Can we actually help them understand what's going to help them close the books faster every month or every quarter? For my team, it's what are the repeatable things that you find yourself doing a lot? Building proposals, we have a, we have an ROI calculator and not everybody on my team loves math and spreadsheets and things along those lines. So like we built out, it's just a project on Claude where they can say, okay, this company has this many employees and the average salary is this, and it can build out an ROI summary for them. And so it's helping to understand, okay, what are things that are really going to be valuable to people and understanding the use cases more so than anything. And then like we talked about change management, now my expectation is that we've created these resources, people know how to use them. When we're sending out proposals, they should always include an ROI calculator. And I will help you with it if, if you need help, you know, creating it the first time. But it's enabling people, it's setting that change management strategy and then leaders holding their teams accountable to, to do, doing things in a different way.

Yeah. So we've talked a lot and tackled this from a lot of different angles in today's conversation, Bobby. As we wrap up, you know, cuz I'm, I'm sure that this is gonna resonate with a lot of our people listening, you know, whether they're a business owner, whether they're a new employee, you know, I think this is one of the biggest issues people are facing right now. But what's your one most important piece of advice, uh, to help groups, to help organizations, to help enterprises kind of make that transition from moving to prompting to actually using autonomous workflows?

Mm-hmm. Um, can I give kind of two-ish?

Please, yes. Yeah, you can break the rules, you know.

Yeah. Um, the first is just like encourage and foster and create this, um, environment of curiosity. Um, you're probably not gonna get this right the first time as an individual user. but encourage people, hire people, uh, that are curious because like you said, there's so much to be learned here. And if you invest the time in using AI, you can usually get to what you want to do. And oftentimes the AI will help you get there. I think you, as you probably know, asking the AI, hey, how do I write a better prompt? Or how could I do this? Or what am I doing wrong? Like you can get a lot of coaching from the AI. So I think a lot of it is is hire and foster the sense of curiosity, because it's going to be really hard whether you're working at a company of 10 or 10,000. You're not always going to be able to sit down and have a one-on-one and coach everybody on exactly how to do these things. Your people have to have that sense of curiosity, and probably, I guess, an environment of knowing that, like, it doesn't have to be perfect and you can try different things and break things. That's the first thing. And I think then the second thing is provide structured time to allow them to do this. One of the biggest challenges, and we work with some of the most reputable and successful companies in the world, a lot of people have been doing certain things a certain way for 10, 20, 30 years. It's hard to tell 'em you have to do things 'cause they're like, wait a second, I've been really successful. I'm hitting my number, or I'm doing this. And to tell them you have to do something a different way, especially when they're busy, is hard. So give them time to experiment, to try different things. Show them other examples of how it's working, but give them that creative freedom and structure and time to, to explore it themselves.

All right. I think that's great advice. I love those two things. Yeah. Being curious and giving people the space to, to be curious, I think is a great way to wrap up today's show. So Bobby, thank you so much for taking time out of your day to join Everyday AI. We really appreciate it.

Yeah, I appreciate you having me. This was great. Thanks, Jordan. All right.

And if you miss anything, y'all, don't worry. We're going to be recapping it in today's newsletter. So make sure you go to youreverydayai.com. Thank you for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks, y'all.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit youreverydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.