A human with average intelligence plus AI is going to outperform AI by itself or the greatest human by themselves every day of the week.
SPEAKER_00Zach Engler is a certified chief AI officer and author who helps businesses navigate the future of work with 20 years simplifying complex tech. He empowers teams to use AI with clarity, confidence, and real results.
SPEAKER_01I've basically spent the last 20 years of my career taking highly complex subject matter and boiling it down.
SPEAKER_00The walk stage, I think you're saying of automations would be a key characteristic of that stage. What does that look like? How complicated is it for a company to do that? What does that process need to look like?
SPEAKER_01Once you've laid the groundwork and your teams have a solid understanding of prompting best practices, how to leverage a large language model and transcribing their meetings, then you can start looking at how do you do process mapping, how do you start looking at your employee workflows, and how do you start looking at your standard operating procedures to pick apart the areas where it makes sense to start automating different elements with AI tools.
SPEAKER_00What's the general thesis of turning on machines?
SPEAKER_01Turning on machines is a triple entendre, and it's this idea that these three stages are three critical elements that we're all going to face as AI gains that exponential capability we're talking about.
SPEAKER_00There's so many tools. Where do we start? How do you tell them to address those?
SPEAKER_01I always suggest, if anything else, just look to the big four Open AI, Anthropic Microsoft, and Google. Which one do you choose? I have to know them all, so when it comes to AI tools, I think sales teams could benefit from a tool called Perplexity. I think we are running up against a clock.
SPEAKER_00Welcome to Using AI at work. I'm your host, Chris Daigle. Each week we'll be learning how today's business owners, entrepreneurs, and ambitious professionals are getting more done with smart use of tomorrow's tech. Let's get started. Right now, every business leader is asking the same question. What are we going to do about AI? If this is you, ChiefAIOfficer.com has the answer. We give you a simple path forward where we provide executive and team training so your people know exactly how to safely use generative AI in their day-to-day. We also manage the deployment and implementation to make sure tools actually get adopted and deliver results. And we'll also guide company-wide transformation so AI becomes part of your operating system, not just another shiny object. The companies that act now will increase productivity, cut costs, and grow faster than their competitors. Those that wait will get left behind. So if you want to make AI work in your business, visit ChiefaiOfficer.com and see how we're helping companies of all sizes finally get results from AI. Greetings, uh listeners to Using AI at Podcast. Today we've got a very special episode with uh Chief AI Officer Zach Engler. Zach is not only a guest, but he's a friend. Everybody who comes on the show is a friend ultimately. However, uh Zach and I have been able to participate in the AI journey together over the past year that he's been involved in the chief AI officer community and he's done some amazing things, some of which we'll talk about today. So Zach brings with him to uh the AI space a successful and I think award-winning career in uh executive placement and recruitment in the AI uh environment, but more on the robotics side prior to his uh pursuit as a chief AI officer. Um but Zach, before we uh talk about the main topic today, which is Zach's new book, is there anything else that you think might help the audience better understand kind of your perspective and context?
SPEAKER_01Uh sure. Yeah, thank you so much for having me today, Chris. Uh pleasure as always to connect and super excited to talk about all things AI related. Uh and yeah, when it comes to my background, I think one of the biggest things that helps people understand my take on the whole AI uh at work type uh philosophies that I've developed is I've basically spent the last 20 years of my career taking highly complex subject matter and boiling it down so that either the people that I'm hiring for a role can understand what the company does and how we're doing it, or from a customer perspective, you know, helping customers digest, uh like in my time at Apple, highly complex concepts uh or technical challenges, uh, and then being able to digest that in a simple way that they can easily uh take advantage of quick wins. And so I've tried to take all of those lessons from my time at Apple, my time in training and development and my time in recruitment and boil that down so that uh now I'm working on explaining AI to you know huge corporations, small businesses, whoever's willing to listen.
SPEAKER_00Nice. So we're gonna talk about Zach's book today, turning on machines. I'm excited to learn more about the book. But before we do that, Zach, what what's the um what's the experience been like for you as a chief AI officer? Because we routinely work with companies and we explain it to them that it it is a it's a new role in the C-suite. It's not necessarily, it's not a tech thing, because uh any anybody in the tech space I think will agree. The way that that AI needs to be brought into a company, it's a people thing, it's not a technology thing necessarily. Um so what has the experience been like? Because now that you're working, now that you're a chief AI officer with a uh a significant concern, um, what is the what is the experience like for you and how are you interacting with clients and helping them kind of grasp why they need this?
SPEAKER_01Yeah, absolutely. In the in the corporate world and and in the business world in general, I think there's three key stages that every organization needs to work their way through in order to roll out AI effectively. And so whether it's internal or external, I'm helping every organization that I partner with understand the importance of starting at the crawl stage uh so that you can get a foundation laid across your organization. Then what does it look like to transition into the walk stage where you begin interconnecting departments with AI tools usage and starting to incorporate the use of agentic workflows and things like that? And we can double click into that if you want as well. But then the third stage, the run stage, where you're now looking at uh creating custom systems that are going to help your organization, you're creating custom tools. That's where unfortunately a lot of businesses think they need to start. They're they're told you need an AI plan, an AI strategy, um, but they don't realize there is a whole foundational layer you need to now set with your teams. The other piece that I think is important there that folks are missing is AI is not going to be an easy button, even if you did go develop that custom tool. The easy button is the fact that every single employee at their fingertips has the world's most intelligent system ready and waiting to help them out with their day-to-day work. And we've discovered three things that can help people instantly gain 10% productivity across the board in any sort of knowledge work. The the three things are number one, access to a large language model. Yeah. In in my work, I've seen a lot of companies just pump the brakes and say no AI tools whatsoever. And I think that's the wrong answer. I think you do need to have at least one known good AI large language model that your employees can work with. Uh number two is prompting techniques. And uh the philosophy that I try to teach here is something I borrowed from my time as a musician. So I actually went to school for music, uh, music performance. And I don't know if you ever tried to learn to play an instrument. Uh yeah. You know, have you played anything in your okay? So when you're trying to learn an instrument, you can yeah, you can go learn a bunch of songs, right? And it's fun, especially like if you're a guitar player, you can go learn a bunch of songs, but you have to memorize everything in the songs. If you learn how to play scales and you learn how the actual notes work together, you can create anything. You can create millions of different sounds and millions of different opportunities. And so I try to draw that analogy to prompting. You can go out and find prompt libraries with hundreds or thousands of uh great points that are tailored to specific scenarios, right? But then you got to go hunt and peck that library for the specific scenario you're working on. Whereas if you learn the prompt frameworks that I teach and are are something uh that are definitely accessible, it's it's public information, but I try to present it in a way that makes sense for a business use case. When you learn those frameworks, now you've got kind of this recipe book of core elements where you don't need to know the specific prompt. You just need to know how do I stack these frameworks when I'm building a go-to-market strategy, or how do I stack these other frameworks when I'm trying to evaluate my board report, or you know, so on and so forth. So I think learning how to understand the frameworks of prompting just makes you that much more dangerous. Plus, it's free. Like everything that I'm talking about right now is minimal cost for an organization. And I think that's a huge win. The last piece, the the three things that uh help organizations gain that 10% productivity right out of the gates is being able to transcribe your meetings and conversations. Um, this can get a little tricky if you work in the healthcare space or the legal space. You really have to be careful how you're transcribing and what you're transcribing. But that being said, that provides a huge uh amount of information that AI systems can then help on the back end when it comes to coordinating your different business initiatives, your different business objectives. So um, you know, most organizations, I think, at a certain level have Microsoft or Google platform rolled out, uh, which instantly gives them access to a large language model that's relatively safe to utilize. Uh, but picking up those prompting frameworks and and being able to uh target the appropriate meetings and conversations for transcription uh will will definitely unlock huge amounts of productivity just across the board. So once you lay that foundation, now you've established the walk stage and you're sorry, the crawl stage, and you're ready for the walk stage, where now you're looking at those agent deployments, you're looking at how departments can gain efficiency using AI tools to talk to each other. And that to me is really kind of where the rubber meets the road. And now you're looking at productivity gains of you know 40 to 80 percent in certain instances. Uh, when you look at software development, now you're talking potential for 100 to 150% productivity gains. And that's just the walk stage, right? Then when you start incorporating custom agentic workflows where AI systems are able to learn and adapt and improve over time, now you're getting into the 300 to 1000% productivity gain range, but you can't start there. You without everybody at your company having that foundational fundamental layer of understanding, because they need to know how to leverage the most out of the tools before they start rolling out the tools, uh, if that makes sense.
SPEAKER_00No, that's fantastic. The the concept of the scales makes perfect sense. You're absolutely right. So anybody that out there that's thought they've they've uh won the jackpot by downloading some prompt library with a thousand prompts, um helpful, but certainly not the ticket. Um that's great. So uh timeline for these stages, let's say you're working with a client. What's the timeline typically before you and they feel that they have like they're good to go in the crawl stage?
SPEAKER_01It it can vary depending on the size of the organization and how fast they want to deploy these tools. And so I try to break that down for companies as well to prioritize. One of the easiest ways to do that is um look at your business and and see where are my biggest cost centers and where are my biggest revenue drivers, and then figure out like where do I want to save the most time and cut down uh and and push people forward uh with their sales capacity, for example. So many of the executive level teams that I work with say that they were told they need an AI strategy and an AI plan. And the first question I ask is do you need an AI strategy for infrastructure or do you need an AI strategy for operations? Um and sometimes the shoulders go up and the and the eyes dart across the room because it's like, what's the difference between the two? Yeah. Right? And it's like, okay, so when it comes to infrastructure, those are AI systems that are, you know, home-built tools and resources that are gonna benefit the uh the way that the foundational elements of a company work. This could be, you know, an agent that analyzes your finances or uh does any sort of like predictive maintenance, things like that that are internally focused, I call, I like to call infrastructure so that it helps them understand like what they're building towards. On the flip side, operational AI is going to be things like the rollout of a large language model to every employee so that they can become more productive in their day-to-day uh communications or project work. Once you figure out which one of those two you want to pursue, then you can start building your plan and your timelines for okay, we want to cut costs in this department because that's our highest cost center. We want to increase revenue in by leveraging this department over here. Let's create a training plan for those teams to get the fundamentals laid out. And then from there we can start looking at how we're gonna interconnect into the middle, right? How are we gonna bring the rest of the company? How are we gonna bring the rest of the organization uh along for the ride as these kind of leading teams kick things off with early stage pilots?
SPEAKER_00Love that. All right. So tell walk me through what this um the walk stage where we start to introduce, uh I think you were saying uh automations would be uh a key characteristic of that stage. What does that look like? Um, how complicated is it for a company to do that? What does that process need to look like?
SPEAKER_01Yeah. So when you're once you've laid the groundwork and your teams have a solid understanding of prompting best practices, how to leverage a large language model and transcribing their meetings, then you can start looking at what are uh how do you do process mapping? How do you start looking at your employee workflows, and how do you start looking at your standard operating procedures to pick apart the areas where it makes sense to start automating different uh elements with AI tools? And this, you know, usually can be anywhere from a three to four month project. Um and I do try to encourage a sense of urgency around this because as we march forward in time here, I think we we are running up against a clock. Every organization at the forefront of this technology has had the last nine months to start exploring and rolling out AI agents into their workflows. Those agents don't sleep, those agents don't eat, those agents don't have families, they don't take vacations, right? And the faster you can perfect how those agents interconnect and work throughout your departments, um, as they scale and get more intelligent over the coming months, you're just your company's gonna have an increased velocity and it's gonna be able to ride that wave of technological advancement that uh companies like OpenAI, Anthropic, Microsoft, Google are pushing the envelope on. If you if you take if you take too much time and you sit back and wait to even lay that base layer and then start moving towards this idea of having agents do work, you're you're essentially falling behind at an exponential rate, right? Like if the if the technology is advancing at an exponential rate and you're not adopting it, you're falling behind at an exponential rate. And I think that gets more and more dangerous the closer we get to 2027.
SPEAKER_00So, you know, that that's an interesting concept because I've been thinking about this a lot, and and companies say, well, we we haven't decided yet, right? And to me, standing still isn't you're you're you're you're losing ground. Combined with the fact that you look around and maybe your competition hasn't like rolled out their uh AI impact or anything yet, but you don't see what's happening under the surface. They could be working with a Zach, right? And they could be training their staff up and they could be building these things out and testing them before they roll them out. So if you're waiting, all of a sudden, not only are you falling behind because the technology continues to advance and get easier, but your competition that started three months ago starts to roll things out. The opportunity for you as a business who's been hesitant uh to catch up, I just don't, I just don't see it happening.
SPEAKER_01Exactly. And I I I think that idea that maybe risk uh plays a part in in the stalling of that decision-making process. I think uh fear of making a mistake or picking the wrong tool is uh another key element that causes companies to pause on that decision-making process. And you know, to your point, this is one of those areas that I think my working at a startup really helped me understand. Like sometimes you just gotta build an imperfect system and iterate on the back end. And AI, I think, is a perfect example of that for any major business at this point. You just gotta find a small pocket of your organization that's trusted uh to leverage these tools and get them going, get off zero, uh, right? Get get somebody in your company moving forward, exploring these tools and looking for creative ways to then roll it out to the rest of the teams in a more managed, structured way. But it again, it's it's cursory stuff, Chris, that I think many organizations miss and and overcomplicate the the conversation before they even dive in. Um so I think that's one area where I our our company and our team at C4 really helps break that kind of bubble of illusions and hype and doom and really pair it back to what is going to move the needle on your biggest problem right now, or what is gonna move the needle on a handful of really small problems uh that you can then make cover ground more quickly once you get those out of your way.
SPEAKER_00I love it. You mentioned the the the risk and the the tool confusion. Um, what would you suggest? I I know you said, hey guys, you you got to get started. You don't have a choice. Um, but what would you suggest? How do you how do you suggest a uh a listener address the the hey, we'd like to get started, but we don't know enough about the risks to to really get serious about it? And there's so many tools, where do we start? How do you tell them to address those?
SPEAKER_01Thankfully, a lot of the big players in this space have created robust platforms that are secure uh that you can get started with. And so I always suggest, if anything else, just look to the big four OpenAI, Anthropic, Microsoft, and Google. They all have fantastic platforms that uh are relatively inexpensive, that with the proper uh knowledge can be leveraged in a secure and uh protected fashion. And whether that's opening up Copilot and uh leveraging the tools that Microsoft provides to get started, uh that's that's one approach. You can also seek support from consultants like us on how do I how do I then take this and apply it to business use cases so that my learning journey can serve me a twofold win where I'm not only learning how to use these tools, but I'm getting work done while I do it, you know? And a lot of the trainings, I I did a seven-hour training with a regional liquor distributor a couple of weeks ago. We we spent seven hours talking about these tools, working through different use cases. And by the end of it, they walked away not only with a whole new understanding of how they can set up their governance, manage risk, they were able to then walk away with a couple different agents that help them with their day-to-day work. Um and that's something that I think is really meaningful and really helpful to consider when you're looking at these problems of how can how can we do work along the learning journey, right? And now you're working smarter and not harder for for rolling out these these new tools.
SPEAKER_00Yeah. So what tools do you recommend? Like, like what are your preferences for uh app quick application? Um, you mentioned the opportunity for transcription and that sort of thing. Like, what are some of the tools that you usually recommend to clients?
SPEAKER_01I think some of the first tools to take a look at after you check off one of the big four and you've got tools uh can be I I have to know them all. So I'm yeah, you know, signed up and subscribed to all four. And um, I I think at the end of the day, at an enterprise level, most organizations have Microsoft already deployed, Microsoft 365. So they've got Copilot sitting there, and it's just sitting there for a vast majority of their teams. Let's get them started using it. Let's look at those day-to-day use cases. Um, and then beyond that, I think each team has different needs uh when it comes to AI tools. I think sales teams could benefit uh from a tool called perplexity. It's fantastic at research, it's fantastic at citing its sources so that if hallucination is a problem where the AI is trying to make stuff up on you, you can actually go click into the resources it cites and validate is this real? Is this something that is going to be relevant to my use case or is this something that's a distraction or a made-up uh website, for example? I think there's other tools uh like voice AI systems as well that can benefit from uh at ex at an executive level, can help leaders digest information much more quickly. So when you look at Google's notebook LM, for example, it's an amazing tool where you can dump in tons of information like your board report, your balance sheets, your uh company website. And have the AI digest that all and build a podcast that you can interact with around those concepts. Uh, you know, you think of a CEO trying to prepare for a board meeting, you could spend hours doing research, hours getting familiar with all the different ongoings of the company. But how cool would it be if you could have a self-tailored podcast that coaches you how to have that discussion, that gives you those sticky board members' points of view that are going to push you on certain subjects. You can download that in a podcast and listen to it before you go to bed and wake up the next morning ready to rock and roll. Like those are the types of simple things that add a ton of value and cost almost nothing, you know. But if you haven't had the time to practice or try it out, it's not prevalent. It's not something you would maybe initially think to try. You're gonna maybe look at the tools as like an email helper instead, you know. So the it's things like that that I want to help people move past and and roll out in unique and novel ways.
SPEAKER_00Great advice. All of this has been um, I'm glad we went down here before we started talking about the book. Now let's shift gears. Uh for the listeners, the reason that we're doing this episode, I love talking to Zach. He's obviously you can tell he's got a lot of uh insight into how it's being used in real life day-to-day businesses. Um, but he shared with me that he was about to release a book called Turning on Machines. Um, I think that based on what you've already heard in this conversation, like you know that he knows what he's talking about. So, what's the general thesis of turning on machines?
SPEAKER_01Turning on machines is a triple entendre. And it's this idea that these three stages are three critical elements that we're all going to face as AI gains that exponential uh capability we're talking about. So turning on machines can mean activation, right? So the first part of the book is all about the rise of AI, and I break down a historical view of what AI or how machines have been leveraged and how innovation has cropped up throughout different stages of humanity's evolution. So we look at stone tools, for example, and we figure out you know, what happened when humanity discovered stone tools? How were they leveraged? How did they impact the way we work, the way we live, the way we uh use them in our day-to-day life, and draw from those lessons to then apply to AI. So one example here would be in there's archaeological evidence that stone tools were actually perfected by sometimes novice users who didn't necessarily know all the traditional ways to carve and shape a stone, and they would by accident discover new capabilities or new forms of chipping stone that was more effective. And I draw that conclusion or analogy to AI there's a ton of really phenomenal PhD researchers developing these systems, but now the tools are so easy to access, the average person can go out and mess around and use different types of pattern matching to come up with unique ways to use the tools for sales, for marketing, for operations, for HR, right? There's there's tons of unique ways to use these tools that uh engineers wouldn't necessarily think to apply.
SPEAKER_00Yes.
SPEAKER_01The second entendre of turning on machines is this idea of being turned on or rebelled against or revolt. Um, and so that's the second part of the book is what types of sweeping change are we going to see as a result of AI proliferating itself through the workplace? Uh, what types of pushback are people likely to give as a result of that change? Um, what types of friction is that going to cause? And here I like to call out uh this concept I've come up with called the trifurcation of work, where I see work getting split not once, not twice, but three times by AI. And each of the different ways that it's split is the first way being the use of AI only in your day-to-day work tasks, the second way being AI plus a human working on tasks, and the third way being a uh human only, right? Areas where you don't want AI involved for various reasons, or areas where people don't want AI involved. And then the last part of the book is turning on, and this is what usually raises eyebrows, turning on in a sense of romantic involvement or or being attracted to um AI. And I firmly believe that we're entering into an era where there's evidence already uh that people are falling in love with their AI chatbots, people are having deep, meaningful conversations with these tools and systems, and they aren't even that good yet. Like there's this is the worst this technology is ever going to be. Um, and yet there's still, you know, I read a story the other day about a woman who bought two wedding rings, one for each of the AI partners that she's developed. Um, there's a man in the book we talk about a man in South Korea who built himself a robot wife and wanted to legally marry her. Yeah. Like again, these aren't these are like rudimentary levels of technology. Imagine 10 years from now when you cannot tell the difference between an AI and a person, right? Then we start adding whole different layers of complexity to this whole situation. And so in the book, uh, we cover those three major areas and give the you the reader tools to think about how that's going to impact their life.
SPEAKER_00So um all three interesting topics as well. And on that last one, you know, uh there's a lot of talk now about um the people leveraging the models for mental health, um, stress, you know, like like like working through life problems and things like that. So I totally see it. So you mentioned that, you know, I like this idea of naive users of the models being able to do things that an engineer wouldn't ever think of. And I it's brilliant, and I totally agree. I mean, that's what I am. I'm I'm not an engineer, I'm not a tech person. Most of the listeners aren't, right? But they're still getting world-class results from the models, right? So this is um an interesting concept. So, based on what you've seen, and and you know, uh, what do they say? History uh may not repeat itself, but it certainly rhymes, right? Um, what uh the trends that you saw in your research about new tools and their impact on the existing paradigm, what kind of parallels are you drawing from what we've seen in the past to kind of where we're going with AI or where you think we're going?
SPEAKER_01Yeah, absolutely. I think one of my favorite ones as a result of the research was uh looking back at the invention of the printing press. When Gutenberg's printing press proliferated throughout Europe and Asia, um, you saw this amazing proliferation of information and knowledge, which was fantastic, right? People could uh learn more, learn more quickly, they could discover new subjects and information much more quickly, they could communicate ideas and uh and and push change more quickly. However, uh, a lot of falsehoods and lies and propaganda were also proliferated, right? Uh, at the same time. And that feels kind of familiar to what's happening today, right? When you look at social media and the rise of how people are sharing uh the different forms of communication we leverage today, it's getting harder and harder to tell what's real. And then when you add in the fact that you can create an AI avatar that's hard, you know, very hard to distinguish now between a human and an AI avatar, um, and or on social media, how do you tell when it's a real person or a bot? Kind of the similar concept. And so one of the ways that they addressed that back in the 1600 and 1700s was let's create journalistic standards, right? Let's create trusted sources and trusted methods for how we showcase our work, how how we show that we're standing on valid and concrete examples. And I think there's a parallel there that uh maybe that maybe there's a million-dollar idea out there for somebody to take hold of of how do we build those new trusted verified sources. Um and you're seeing some early signs of this. I know Sam Altman has some ideas around cryptocurrency and and tokenization of information being used. Um, but I think even more simple, like look at LinkedIn. LinkedIn uh has a feature that allows you to upload your ID, right? And you can validate I'm I'm a real person. It's a pretty straightforward process. It only takes a couple minutes. But once you get that check mark, when you go on LinkedIn, you can know that somebody actually had to like validate their identification as opposed to some other social platforms where you just pay a fee and that that's your check mark. I think that's a little easier to circumvent even even though it's one extra step. You can still circumvent it. So that tends to be why I use LinkedIn as my my go-to social media, because I can tell, yeah, you know, this person is either real or they had to do a really good job forging an ID. Uh yeah.
SPEAKER_00So, okay, that's um certainly I think everybody's heard of the deep fake examples, and they've probably heard politicians, you know, and and humorous memes and things like that. So definitely something that we can all relate to. With the second concept, um, turning on machines, when I talk to clients, we go on site, there's always that kind of wink, week, nudge, nudge, like, hey, when's Terminator coming? Right? Like that, that is a concern. And being a professional that understands, I mean, you've been using the tools as I have, and you see that like it's advancing. It was great when it first came out, but I would not go back to where we were with what we've got now. So like it's inevitable. Yeah. Right. So, so how are you like what were some of the the I guess the the novel ideas that you developed in in the research for that section of the book?
SPEAKER_01Yeah. So on that front, when it comes to like the pushback that we'll likely see, um there's already been several notable examples in in the real world of like the Hollywood writer strike last year, did a was a really good example of people standing up and saying, no, you're not gonna just take all of my life's work and upload it into a large language model and then recreate scripts or uh recreate movies with my likeness. Um, and I think you're gonna see more and more of that proliferate into the business world once employees start realizing, you know, that that same kind of approach could happen to them. The the longshore strikes that also happened uh uh in in on the East Coast were another great example of people fearing that the automation was gonna essentially wipe out their livelihoods. Um and so you saw these huge strikes of people sitting down and saying, No, I'm not going to let you automate away my livelihood, even though it may be inevitable uh and it may be more efficient. Uh, you can't just, you know, discard people. And I think in my research, what I found was that uh there's there's this well-known chess chess player named Gary Kasparov, and he he was the first person to be beat by AI uh in a chess game. And over time he developed this concept called Kasparov's Law that says a human with average intelligence plus AI is going to outperform AI by itself or the greatest human by themselves every day of the week. Uh, and he's proven this over and over again. And I think it's this concept that businesses should also adopt when we go back to this idea of the traffication of work or whatever it may be. Your teams are going to be way more productive if they're paired up with AI effectively and given tools and resources to use it effectively than if you just try to roll out automation by itself or if you expect everybody at the company to do the heavy lifting on their own. Um, so I think making that uh analogy and being able to just see that roll out in the real world is what opens that mindset to business leaders to say, you know, it may look good on a spreadsheet and it may be good for a quarter's worth of gains, but in the long run, we're actually going to be doing ourselves a disservice by automating away. And we've seen companies like uh Shopify or uh Microsoft rolling back some of the uh initial automation, automate everything approach that they had taken and saying, you know, actually we do need to have a human in the loop for some of these things.
SPEAKER_00Yeah, no doubt. So for the companies that say, yeah, yeah, we want to keep our people, but we are interested in the automation side of things. Um and it was I thought it was an interesting analogy because a big part of the Hollywood strike was people started seeing the trend that was happening, that AI can produce what's taking me, you know, my my entire career to be able to do, whether it's as the actor or as the uh the technician behind the scenes, it's producing the film itself, right? This idea of employees having that same concern, I had never thought about it that way. That no, I I do my job in a very particular way very well. And if I've documented everything and I've created a knowledge clone and I've AI fied all the SOPs from my role, like my contribution or my perceived uh ice insulation from any type of disruption in the company, it it may no longer exist. If I am too much of a princess, they might just say, listen, we've got everything we need. Yeah. Uh have a good career.
SPEAKER_01So that's that's like this absolutely. That's this huge reason for pushback. And and uh this is a this is a big reason why teams don't want to adopt these tools. So my one of the things that I've come up with on that front is this idea that you're gonna become a machine shepherd, right? Uh, which it's like a weird phrase, but yeah. So you you go out and you you learn, hey, uh here's my knowledge base, here's the things that I've done. I'm gonna coach up this AI to help me with these tasks. And all of a sudden you're now 10% or sorry, 10 times more productive in your work. You're now going to guide and shepherd and lead these semi-autonomous agents into doing work. And where it maybe took 10 people to do a job, now one person is going to do that job with the help of agents, right? Doing that work. That person has now become a machine shepherd. Another good example of becoming a machine shepherd would be autonomous vehicles, right? You have dozens of truck drivers out there doing routes. All of a sudden, autonomous trucks come along. And one of those really great truck drivers took the time to learn how the AI systems work, and then they go and they monitor the 10 or 12 trucks that are driving autonomously when they run into problems or situations they can't handle, they remote in and they help the trucks out, right? This is that concept of a shepherd coming to life that's gonna change over the course of the next decade. And I talk about what that change looks like and how your role will shift in the book. But this concept of, okay, if I'm gonna work in a domain that's gonna be disrupted by AI, I need to learn how to become a shepherd. Then you have a set of tools and resources at your disposal to evolve and grow your career with AI. That's not for everybody, though, Chris. And I'm all for the fact, like there's gonna be a group of people out there that just don't want AI involved in their lives. And that's okay. This book is still for them too, though. Like, even though it's about AI, I think it's good to educate yourself on the change that's gonna happen. And there's a set of roles out there that people could go do that are likely not gonna be as impacted by AI or robotics over the course of the next two decades. I think jobs like um dentists, for example, are one where like I don't want any sort of robot with sharp tools around my head personally. Like I just, even if they're really great and really affordable, like I'm sorry, that's just not a thing I'm gonna allow. Um, and I think there's a lot of other people out there that would feel that way too. Um, so I think certain jobs where there's human-to-human contact, right? Yeah, healthcare, uh surgery, they may be using AI to support some of their decision making or things like that. But at the end of the day, it's still a person digging in your mouth or it's still a person giving you a massage or whatever it may be. One of the other categories that I think is really interesting in the business world is this idea of experiential roles, right? So it's gonna be pretty hard for an AI to be a rock climbing expedition guide or a canoe expedition guide or a spee lunking guide or a surfing teacher or uh, you know, whatever it may be. Areas where there's an experience, uh, I think are gonna be really hard for AI and robotics to touch. So that could be another uh way to kind of protect yourself and insulate yourself. If you don't want to uh work with AI, you can kind of circumvent it and work around it. And then there's two other ones I won't dive into deeply, but I'll just say like roles and that require certification. So like I just redid my basement. I needed to have an electrician come in and certify, like a certified electrician come in and validate like all the wiring was done properly. Nobody's gonna hand that stuff over to an AI or a robot anytime soon. Um, or or other roles in government, like uh judges and and lawyers, I think, uh will also be required legally to have a person in that role and capacity. So you can kind of look at the world in those two lights of do I want to work with AI and become a machine shepherd, or do I want to go down this path of protecting myself in a new career? I'm gonna pivot to an area where it's gonna be really hard for AI and robots to do this stuff.
SPEAKER_00You know, that's uh one of the things I just took away from that was there's people that they they might hear that you know AI is changing the workplace and they think, oh man, I've got all this uh all this, all my 10,000 hours in what I do. And now that's not gonna matter anymore. I've got to go learn AI. And the way that you positioned it was because of your 10,000 hours, yes, you've got to learn something like look at things through a new lens, but that's gonna make you a better uh machine shepherd, right? So I think that's very encouraging for exactly um any knowledge worker for sure that is concerned about AI's impact on their role over hell just the next 12 months. Um so let's talk about the the third section of the book.
SPEAKER_01And I'll just add yeah. Could I just add to that, Chris? Uh sorry to interrupt. Please the faster you adopt these tools, if you want to stay in the knowledge worker lane, the more ahead of everybody else you will be, right?
SPEAKER_00Yep. For sure. And I guess, you know, we talked about this concept of businesses waiting on AI, and waiting is actually falling behind, it's not standing still. Same thing applies for the employee. Right. Yeah, for sure. Um so now let's talk about the the third section a little bit more. Um, help me better understand this trifurcation concept.
SPEAKER_01Okay, so when you when you look at a process at a at a job uh or or a workflow within how an employee does their day-to-day work, as you start mapping out from start to finish how that work gets done, there's going to be areas where it makes sense to incorporate AI. Let's say areas where there's highly repeatable digital tasks, areas where there's a lot of data analysis or ingestion, um, where it makes sense to just have an AI go do that work and support that work 100% hands off, just doing it automatically. That's kind of that first layer of the trifurcation. Uh, mapping those out, identifying them, building a tool or a system to take it on. The second element of the trifurcation is this idea of humans working with a machine. Oh, yeah. What's that category?
SPEAKER_00What would you call that category?
SPEAKER_01Uh AI only. AI only AI only work. Okay. And then you've got that that middle ground where it's human plus machine, or some people call it human in the loop. And this idea that there's going to be a semi-autonomous task the AI can do, but a human's going to validate the work that's done in some shape or form. They're either going to be a part of the process along the way as the AI builds something, or they're going to be the quality checker at the end of the uh process and validate that what was produced is accurate and high quality.
SPEAKER_00Okay. And then what's that? I I guess trifurcation would indicate three. What's the third human only?
SPEAKER_01So the human only, exactly. So you've got AI only, human plus AI, and then human only. And this idea that, you know, if you look at um like a hairstylist role, for example, uh, a hairstylist is probably gonna be one of those other roles that's really safe, right? You got sharp objects around somebody's head. That's probably gonna be a human-only task, right? Of of actually cutting somebody's hair. But uh and so you're you're gonna protect that. You're gonna make sure that AI is not involved in that in any way. Um, even though AI may schedule the haircut, even though AI may process the payments and you know, track the overall performance of the team, things like that. At the end of the day, it's still gonna be a person with a set of scissors doing the actual work.
SPEAKER_00So, how would you recommend? Is this something, is this an activity I can do? Do I need um is there a framework that I should use to identify, okay, which parts of my role or my department or my entire business fall into one of these three categories?
SPEAKER_01Yeah, that's a really good question. And I think it's unique to every business because the way that their work gets done, the way they provide their service, the way they provide their product is tailored to their customers and their clients. So you have to map out in the AI only. Section, for example, what are those parts of the business where it is okay if a mistake occurs, right? As a result of the AI performing its task. In the realm of finance, that becomes much more high stakes than in the realm of customer care, for example, right? If an AI struggles with a problem with customer care, it can route them to a person. If it gets the balance sheet wrong for your entire organization, that's a much bigger problem and the stakes are a lot higher. So going through and mapping out where's the ROI and how much of a negative impact would the AI have if it did something incorrectly is a great first step to figure out where we can let AI do its thing and where do we have to have a human in the loop to validate once it's done.
SPEAKER_00This is fantastic. So the book is coming out, but I guess by the time this will be released, the book will be out. Usually when we have speakers on here, uh I always say, oh, where can we find out more information about, you know, if people want to follow what you're doing? But in particular, I would like listeners to focus on the book. Um I think at the beginning of this episode, we had a chance to hear Zach's perspective on things. And listen, I talked to a lot of AI experts and AI thought leaders and stuff like that. And I heard some ways of thinking about bringing AI into the business that I don't hear other people talking about, which is always encouraging and an indication that this person is not just talking about it, but they're actually doing it. So these ideas, if you are an AI professional, you're an AI enthusiast, and you want to bring these into your business or bring them into your network, or you're delivering these services for other companies, the topics that are discussed in this book are things that you're gonna need to have a good answer. They're they're gonna be common, um, at least in my experience, these are the things that come up. So I, if nothing else, I would suggest that this would just reading this book is gonna better prepare you as that AI champion in your own business industry, department, uh whatever, to have quick answers that uh are obviously well thought out and address those friction points perhaps with a company adopting it, or the boss saying, Hey, yeah, let's do it, or the board approving your direction with the company, or a client saying, I don't get it, why shouldn't you use that kind of thing? So Well, Zach, any closing remarks?
SPEAKER_01And there's one Yeah. I was gonna say there's one key element of the book that I neglected to mention, and that's um that the book is built with a series of self-reflection questions at the end of each chapter. Uh I've at this stage uh in my career, I've discovered that adults don't like to be told what to think. They like to discover and learn things for themselves. And what in my time at Apple, that was actually one of the key elements that we learned for how to how to teach adults effectively is this idea of depth of knowledge questioning, where you ask a set of questions where even though you know the answer, you're gonna guide the learner to a particular conclusion or help them discover the conclusion that makes the most sense for them and their particular situation. So I at the end of each chapter, I asked the series of self-reflection questions to put the control back in the reader's hands and say, how is this gonna impact me? How is this gonna impact my life? How is this gonna impact my work as it's relevant to me? There is so much that's outside of your sphere of control in the way AI is changing the world, the way AI is changing work. I want to put that control back into your hands and say, hey, what is in your sphere of control that you can take a hold of and start taking action on today? Um, and I think also maybe today, I don't know if you feel this way, but I certainly do. There's a lot of people saying how you should think or what you should think. Um, in my research for the book, I've I've probably read over 50 or 60 different books on AI, and a lot of them are written by incredibly intelligent PhDs who have done the research, who have put in the time, but they still tell you why they're right and why you should believe what they have to say. And although they're probably right, like they they're they're on the right track, um, this idea that you need to form an opinion for yourself, and it's so much easier when somebody asks you a question rather than telling you what to think. Um, the other piece of that, taking that approach, is you'll be 12% more likely to remember what you actually read in the book and apply it to your day-to-day than you would if you were just being talked at. So I I think um it there's a fair amount of research I did into this idea of how do we get the average person off zero when it comes to the use of AI? How do we get them thinking about these tools? And my big hope with this book, Chris, is that if we can raise the awareness level of everybody to some degree, we'll collectively be more likely to make better decisions about how AI is proliferated through society, about what we're willing to accept, right? Um, because again, there's only so much in your sphere of control. Um in the book, I talk about this uh philosopher Albert Camus. He was an absurdist. Uh, if if you're not familiar, go check out absurdism. Um, and it's this idea that even in a gigantic universe where you are just a spec and ultimately can't control anything, there's beauty in the fact that you do get to control what's around you and how you handle and deal with things. And this concept that you can do that in any way that you want to tackle, you can look at the world and just marvel at the fact that it's at your fingertips. What are you going to do with it? Right. So that's kind of the book uh meaning and intent at the end of the day is how do we raise that awareness level and how do we put some control back in your hands?
SPEAKER_00I love it. And congratulations on the release of the book. It's quite an accomplishment for sure. So tell people where they can go and get more information, uh, buy the book, uh, learn more about the content that you're putting out as far as your c your uh reflections on AI.
SPEAKER_01Yeah, if you visit zakengler.com, uh it's just my name spelled out, ZachIngler.com, or if you go to turningonmachines.com, uh, I've got both those URLs synced up to more information about myself and the book. Uh you can pre-order it now. Uh we're working on finalizing the launch date and getting the actual orders of the book uh logistics finalized. And uh by the time this show airs, it should be set up on zakingler.com to purchase the book directly. Um and then and then from there, I'll be sharing more as we get closer to the launch date on my social feeds as well. So if you haven't connected with me on LinkedIn, like I'm I'm an open book at this point, Chris. And I think that was one thing that really attracted me to the the chief AI officer community, which I know we don't always dive into on the show, but like at the end of the day, that is truly made a huge impact on my life. And I want to say thank you to you personally. Uh if it were for this podcast, I I would not be here talking to you right now, and I would not be a chief AI officer. Um, a lot of this stuff is a direct result of the work that you and this amazing community have developed. And I think there's something to be said for coming together with a group of like-minded individuals.
SPEAKER_00Yep, for sure. Thank you for that. So, everybody, Z A C Zach Ingler. Yeah, if you're going to zachingler.com. But listen, check the show notes because we're gonna have the links to all of this stuff. Um, I'm a big fan of the positions that uh Zach takes on AI. I'm a big fan of the performance that he gets for his clients. So um we're uh we're glad to have you on here, Zach. And um, we will be uh I'll be waiting for the release of the book so that I can get my copy for sure. I'll probably be giving a few of these out to uh peers and clients as well. So again, congratulations. Thank you so much, everybody. We'll see you on the next episode of Using AI at Work. Go Use AI. Thanks, Chris. Thanks for tuning in to Using AI at Work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor, Chief AI Officer, for empowering businesses with AI education and training. Visit their website for free AI readiness assessment and AI strategy guide to help you get started using AI at work. That's www.chiefaiofficer.com. Follow us on Twitter at the handle UsingAI at work and visit www.usingai at work.com for free resources to help you harness AI in your role.