Eric Housh 

Hey again, everyone, welcome back to the Fast Slow Motion Podcast where we share strategies, ideas and insights to help you build a business you love while enjoying your life. I’m your host as always, Eric Housh. And today we’ve got something a little bit different in store for you. If you’ve been listening for a while, you know on this show, we talk about building scalable systems. We talk about developing leaders. We talk about creating businesses that don’t consume you.

But there is another conversation happening inside Fast Slow Motion that’s worth bringing into this room. And that’s what we wanna do today. So joining me is our Director of AI at Fast Slow Motion, Zack Terry Terry. Zack Terry, great to have you on the show, man.

Zack Terry 

Eric, so good to be here. Thanks for having me.

Eric Housh 

So Zack Terry and I are no strangers to the podcast scene. We actually co-host another podcast called AI in Action. And today we want to talk about why that show exists, what we’ve been exploring there and how that content connects directly to the business leaders listening to this show. So think about this, if the Fast Slow Motion podcast, this podcast is all about building a business you love, AI in Action is about intelligently leveraging

AI and the AI tooling within that business. let’s get started at the top. Zack Terry, first of all, why did we think we needed another podcast focused specifically on AI?

Zack Terry 

I just don’t think there’s enough AI information out there today. And so of course we had to jump into this massive gulf of a lack of information in AI. No, I mean, obviously there’s a ton of information out there. think if we think about why did we start this, I think what immediately comes to mind is just exploring this whole new frontier, right? This is a great way for us to hold ourselves accountable.

Eric Housh 

You

Zack Terry 

and stay plugged into all of the latest happenings in AI. That’s what I think is one reason. But then the deeper reason here is in the work that we do and in the work that I do, we kept seeing a common thread among business leaders. And in general, it just felt like everyone was overwhelmed, overwhelmed by the headlines, overwhelmed by how much is happening and how quickly it’s happening.

overwhelmed by the pressure to do something with AI. think a lot of business leaders feel that and you may be trying to implement it. You may say, I’m not doing this right now, but I think a common thread is it’s really hard to understand how to practically apply AI in your business. And so that was just a big gap between hype and implementation. And obviously there’s a ton of hype. We see it in the news all the time. People are talking about the end of work as we know it.

And obviously there’s some extremes, right? I think our goal is to exist in the middle ground and to say, well, what’s realistic? What’s likely? What do we think is practical? Can AI be a benefit to your business? Absolutely. Is it going to change the world overnight for you? Probably not, right? So let’s start thinking about how to practically build this into the business. Look at what’s realistic, right? A lot of conversations that we hear were either way too technical or way too vague.

And I think it’s got to be somewhere in the middle. There is a technical aspect to this, but it doesn’t have to be diving into the weeds and understanding every aspect of how a transformer model works. It’s more about how do you apply this particular technology to your business to achieve results. So that’s really what we wanted to do is translate this AI hype more into real practical business execution. And specifically, because this is the playground that we work in, but how to do that.

within the Salesforce in the HubSpot ecosystems. These two platforms are absolute market leaders in the space. They were some of the first tools to come out with applied AI on top of their systems. And of course, we work with these tools and customers of these tools every single day. And so that was that was a goal for us as well, which is how can these platforms be used to help your business practically adopt AI? So.

We kind of distill that down. think it’s moving beyond AI awareness and hype to more practical application of AI that can actually help propel your business.

Eric Housh 

Yeah, there’s definitely a lot of information out there. Information coming out every day, information changing every day.

we had a conversation recently of a podcast host on this show. And just between the time that we recorded the show and, and, and, getting ready to publish it, things had changed drastically and it’s going to be that way. And that can just be overwhelming for business leaders. So again, the, idea behind AI and action is kind of just boil that down, give you some really focused content to let you know what’s real. What do you need to be thinking about? Is that one of the early themes that you and I got into really is how to get ready for AI. call it AI.

And we talked at that point in time around four pillar strategy, process, people, and then tech and data. Why is AI readiness such a big deal for companies?

Zack Terry 

think this is an evolving conversation. AI readiness is different today than it was a year ago. But what it really comes down to is I think a lot of companies, we hear how important it is to be at the forefront of AI and we jump straight to the tools before actually defining the outcome that we’re trying to achieve. And so that’s one of the big things with AI readiness, which is we’ve got to have an understanding of where we’re trying to go. We need a map.

We got to put a map in place before we can get to a destination. And so we can use tools to help us get there. Maybe we’re using GPS and we’re not using a paper map. At least I hope we are. Right. And so, you know, use using the tool to get from point A to point B that that’s really kind of what AI is, but having the tool doesn’t decide the destination for you. Right. And so that’s really important just to kind of take a step back and say, what bottlenecks do we have in the business? And even beyond that, do we understand?

our business processes. Have we articulated them? Are they written down? If someone were to join the company today, are we relying on a single person and the knowledge that they have in their head to onboard that employee or have we established documented processes? And so just thinking about that is a really good baseline, right? That’s one really important part of it, which is we’ve got to understand who we are as a business and what we do and what those processes are. Because AI is ultimately going to work.

on top of that foundation. And it’s probably going to have a hard time figuring out what to do if we’re not able to articulate what it should actually do. So that’s a really big part of that, which is stepping back and understanding, what do we actually want to do here? Not just let’s put AI into the business, but let’s understand where we’re trying to go. And then we can look at the tools that are available just like we would with any other technology. This is not specific to AI. This is the same advice that we would have provided to any business looking to do

Let’s say a CRM implementation or a migration or a transformation. It’s all about, what is the end state? Where are we trying to go? And let’s use that as a map to figure out how to get there. Right. So ultimately it comes down to broken processes will just be accelerated with AI. Fragmented data will lead to inconsistent results. And so in addition to the processes being defined, you also have to have a good data foundation. And that data foundation is.

really more about the context than anything. And so I think that the conversation a year ago was how do I clean my data in a way that it’s really well versed and conducive to having performant AI. And that’s still true. We need good data, but it’s really more about the context. And as an example of that, I could sit here and actually dictate to an AI. And that’s not going to be clean formatted sentence structure with paragraphs. It’s just going to be a stream of consciousness, right?

but you can get really good results because the context that’s provided is so robust. And so taking that idea and spreading that out across the business and just thinking about, where do we have context siloed throughout the business? Are our different departments talking to each other or are they completely fragmented? And if they’re fragmented, how do we connect that together so that we get the full story and we give generative AI everything it needs to actually perform well? So those are kind of the big highlights. And then we also talked about a framework that we call the step model.

And the step model is thinking about what AI will do for your business on sort of a curve, right? And so it starts with support at the most basic level, just doing things like everyday tasks, helping you write emails, drafting website copy, know, basic things, right? Things we’ve all probably tried and used AI for. Moving on to think, which is where we’re actually using AI as a strategic thinking partner.

And so going beyond just writing emails and generating texts, but using it as a sounding board to help identify potential solutions, really kind of give you some more context and information. And then when we move beyond, think we got to engage. And this is where AI is actually directly embedded into daily workflows. A human should still be in the loop and ultimately be evaluating what the output of those workflows are, but doing things like categorizing incoming customer interactions and

understanding what the unstructured wall of text means and moving that into structured data, having AI do something like that for us. And then finally, when we get to the P that’s perform and that is truly more agentic AI. That is where we have an autonomous system that has a full set of instructions and a process that it understands and it is able to go out and actually execute things on our behalf. think about something like.

A SDR or a lead nurturing agent that is acting on triggers. And when I get a lead in the system, it understands that I have that lead. It pulls the context from the lead. It generates an email. It sends that outbound email. It’s ready to receive responses. When a response comes in, maybe it’s getting a question about the products and services that we’re offering. And it understands to go and retrieve that information from our knowledge base and then send a reply email back. so that’s where the whole workflow is all happening with just that AI agent. So that’s where it’s really performing.

on our behalf. so thinking about that from the context of AI readiness, I think it’s helpful to look at this as a gradual progression. You’re not going to start with perform on day one, right? It’s going to be a gradual progression. And that’s both in terms of what’s realistic to implement in your business. There’s just time constraints and resource constraints, but then it’s also implementing that into an overall strategy. So understanding how am I going to get our team to adopt these features?

So it starts with baby steps, right? It starts with small things, improving incremental value over time. And then as you start to layer in these capabilities, these processes get more mature, and then you’re able to stack AI capabilities on top of each other until you’re really starting to see performance as opposed to just supporting activities.

Eric Housh 

And I think just to kind of put an exclamation point on that, Zai, AI hasn’t really fundamentally changed anything about running a great business, right? It’s you got to have consistent process. You got to have data you can trust. If we take it back to the scalable business framework, right? It’s the process that your people can consistently execute. It’s the visibility of the data that matters to your organization. That data then feeding context to these agents and to these AI tools.

to help you better execute and help you streamline. So again, can’t bolt AI onto dysfunction, right? You have to have clarity first. You have to build on that clarity. One of the things I think we’ve been really intentional about with AI in action and here at Fast Slow Motion is the hype cycle.

Right? These AI tools, you know, they come out, there’s a thousand articles written. I’m personally subscribed to four or five AI newsletters. There’s always a good deal of hype and froth built in there. We’ve covered product releases, pricing shifts, technical updates, but we’ve always tried to do that through a business lens. For the audience listening out there, why do we do that? Why is that important?

Zack Terry 

think we touched on it a little bit, but the fact of the matter is this is moving at a pace that I don’t think is precedented in any other sort of technological paradigm shift, right? So just that fact lends itself to us paying attention to it and distilling it and understanding what it really means. In addition to that, there’s so much money changing hands.

And so this this is just such a big change across so many industries. The incentives around investment are huge and the incentive for an AI company to perform is also really big. Right. And so I think it’s just important to dive into. OK, what is a five hundred billion dollar investment really mean? Right. What does that mean for my business? Like.

If I’m going to implement AI, does that mean I get a better model? Does that mean that this is going to work faster for me? And so I think sometimes it’s hard to understand exactly how that translates. And so our goal there is really just to, for ourselves, be aware of what’s going on. And so it’s a great way for us to stay on top of things. But then I think we, also just want to distill that down.

and try to get behind what’s really going on in terms of what that means for the business. so our goal is to really interpret those headlines rather than just give a stream of updates. And so, yes, we want to give you the updates, but then we also want to talk about what does that mean and, you know, focus on really kind of explaining the changes and what they mean for operators. So rather than saying, hey, here’s what launched. It’s more about this is how this could impact your CRM, your workflows, your team.

So really trying to get around a lot of the noise and I think do our best just to filter that signal a little bit while still making sure we’re talking about the latest current events. And so it’s really just about clarity before you get to an execution step.

Eric Housh 

Yeah, absolutely. Cutting through that noise, giving our business leaders in the audience just that perspective of what’s important right now and what do they really need to be thinking about versus what may be a little bit more of the hype cycle. One thing we talk a lot about, Zack Terry, we actually hit on this, I think in our last episode is data. Data hub, data 360, clean data.

 We hit on this earlier in this episode, but let’s stay there for just a minute. Why is data coming up so much in these AI conversations?

Zack Terry 

Data is super important because ultimately AI is feeding off of your data. So if you’re building an application of AI on top of your business, what that AI is going to have access to and reflect back to you is the data foundation that you’ve built, right? And so tools like Data Hub in HubSpot and Data360 or what used to be called Data Cloud in Salesforce, these are…

really foundational business tools that allow you to unify very large and disparate sources of data into a single source of truth. And the benefit of that when it comes to AI is context. It’s all about context. Context is king. So if I’ve got a contact in my CRM, but that particular person, that individual is interacting with us, not just in marketing, in sales,

over email, we’re doing tasks as a salesperson, they might be placing orders on an e-commerce website, we might have some information about them in the ERP, maybe we have some invoices that are stored somewhere else. There’s obviously just a ton of business data that you could have about a single individual. Well, without having something that unifies all that together, you are either having to make concessions and say, it’s okay, I don’t have all the context, which means you’re not gonna have as good of a response.

Or you’re going and you’re making disparate calls out to these systems and trying to figure out that context in real time. Well, when you have something that unifies it all together, it’s all in one place. And so you’re not having to create all of these extra methods to go out and get that information as it’s needed. It’s just kind of there, right? And so it’s able to provide that additional context, go and retrieve that information to provide those better results. So without that single source of truth, it’s really difficult to operate.

both as a person that’s working at a business and trying to get information, but also as an AI agent. If an AI doesn’t have access to that context, the response is going to be more limited. So I think ultimately AI performance is going to be highly dependent on the context that it has to work with. Good data is important, but the primary goal is providing the AI with enough context to understand.

what question is being asked and how to resolve that question and how that works within the larger business context and the business process. So I think what we see is that businesses tend to underestimate how much dirty data costs them. And this is true before AI, but it’s especially true now because if you want to invest in some kind of AI solution, generative AI specifically, and you’re going to plug that into your data, well, you might find that that investment doesn’t bring you much of an ROI.

because the data source that it’s grounded on isn’t very good. Ultimately, it’s doing its job. It’s going out and finding that information, but that information may not be correct or accurate, or you may just be missing some context. And so having something like Data Hub or Data 360 to be able to unify all of that data and provide all of that context right where it’s needed is a fantastic asset to creating effective AI solutions. ultimately, it comes down to clean data.

Good context, that’s going to increase trust and adoption because you’re gonna get better responses, better quality of responses from your AI tooling. And then of course, if we wanna talk about AI maturity, we talked about that a little bit on a previous episode with the latest Salesforce release and how they’re focusing so much on AI maturity, but we really can’t have a mature AI system or a mature automation engine without trusted data and good context.

Eric Housh 

Yeah, and that’s something, you we talk about the scalable business framework here at FastSlow Motion all the time. Again.

consistent process, automation is streamlined in that process, visibility of the data that matters. AI doesn’t change any of that, right? If you apply AI correctly, it empowers that model, it strengthens it, it accelerates it in your ability to scale. So it’s funny, Zack Terry, you know, we’ve been having this AI conversation for several months now, and it didn’t seem like the word context was really as front and center as it has been really over the past four or five months.

I think as businesses mature with their AI journey, they’re starting to understand that, these tools are only as smart as the fuel that I’m giving them in the form of really great data. So that conversation around context is an extremely important one that we have with all of our clients.

Zack Terry 

Yeah, and I think if you go back and look at the conversation a year ago, maybe a little bit more than a year ago, people were talking about this new role called prompt engineer. And this was kind of all the rage for a while, right? Well, that concept has essentially completely left. There’s no prompt engineer role. Now it’s context engineer. And what really happened is these models just got a lot better. And so…

Eric Housh 

Yep.

Zack Terry 

the importance of providing the perfectly worded prompt just kind of went away. And now it’s more about providing the right context. And then these models can kind of figure out what you’re trying to achieve. But it is all about providing that context. So now we see context engineer coming up as the replacement to prompt engineer. And the idea there is understanding all of your systems and all of your data sources and figuring out

on a targeted basis, depending on what type of AI tool you’re building or what type of agent you’re building, how do I understand the data and the processes that the AI needs to do that job? And then how do I weave it together in a way that the agent has all of the context it needs to actually perform and do that job well?

Eric Housh 

Zack Terry, let’s spend a little bit of time talking about pricing models because this is something that I think HubSpot and Salesforce are still sort of figuring out how to price the value that AI can add to an enterprise. We have this concept out there of trust layers. So now we have this data that could be potentially sensitive data in our business systems. We’re wanting the LLMs to help us operate on that, but we wanna protect that data.

So that layer of trust has to be there. And then one that’s coming up a lot more recently, the agentic enterprise. So for those leaders that are listening out there that, you know, maybe thinking about these different things, let’s give them some grounding. What do they need to understand about these concepts?

Zack Terry 

A lot to unpack in this conversation. So with pricing, think the market as a whole is still figuring this out. Not just Salesforce HubSpot, but I mean the entire market that’s based on generative AI and the tools that are provided are still figuring this out. I mean, we see frontier models changing pricing all the time. I think what’s common is we’re seeing a lot more per token costs that’s associated, which is

how many tokens go in and how many tokens come out and pricing it that way. So a lot of those frontier models do that. What we’re seeing with companies like Salesforce and HubSpot is they’re shifting from more of what used to be called like a conversation model, which is I initiate a conversation with an agent and I pay a flat rate for that conversation. That has pros and cons, but now we’re seeing that shift more to an outcome based or a usage based model. And so for example, instead of paying a flat,

$2 for a conversation for a Salesforce service agent. You’re now paying for what the agent is actually accomplishing. And so if somebody comes in and starts chatting with an agent, you don’t actually incur cost unless that agent takes action. Now that action could be querying some information from the database. It could be retrieving information from knowledge. It could be actually making an update to records inside the CRM. It could be calling out to an external system, you know, whatever that is.

But ultimately, the agent has to do something in order for pricing to start being accrued. And so we’re really just seeing that move to this full on variable cost type of a model with a combination of things like licensing. And so some Salesforce customers might want to use AI a lot more than others. And so they have an option to pay for what’s called unmetered usage, which is a licensed

based tool where you pay, just like you’re used to, a per user per month fee in order for those users to essentially have unlimited access to internal AI usage. There’s some caveats to that. We don’t have to get into that here. It’s all kind of based on UI. Am I interacting directly with that agent? But there’s optionality is really what I’m getting at. And so depending on which platform you’re using, you may choose to just pay as you go.

You may choose to buy a pool of credits upfront at a discount. You may choose to use something like an unmetered license to just pay a regular seat based price per month in order to have unlimited usage. So there’s a lot of different ways you can go about it when we’re talking specifically about CRMs like Salesforce and HubSpot. So ultimately I think this used to be kind of a black box, but we’re seeing a lot more maturity around

the ability to monitor your consumption, to establish thresholds for consumption, to put some warnings and other stop gaps in place to ensure that if you go over a certain threshold, if you’re on HubSpot, you can just stop. You can just say, I’m not gonna go any farther than that this month, or you can auto purchase additional credits. With Salesforce, it’s a little bit different. There’s not a stop gap necessarily, but you do have proactive reporting, and so it’s really important to spend regular time every week.

going through and looking at your consumption and making sure that you’re changing your forecast to say, I using more than I expected? And what can I expect over the next one, two, three months? Just doing a lot more proactive reporting around what you can expect, right? So I do think the fact that we’re seeing the additional reporting is really great. We’re able to dive in and slice and dice this information at a finer detail than we used to. We can do some custom tagging in Salesforce now to show us.

Exactly who’s using AI and when they’re using it and how often they’re using it and help us plan and budget for, additional usage down the road. When we talk about trust layers and governance, these are going to be different depending on the platform you’re using. But ultimately this comes down to enterprise adoption of AI and having certain compliance requirements and assurances around what companies are doing with that data. And so when you work with a company like Salesforce or HubSpot and you’re using their AI tooling,

They’ve established agreements with these frontier model providers to make sure that your business data is not being trained with Salesforce. You have what’s called data masking. Now they’re changing a few things around data masking. So, I would say if you’re listening to this, make sure that you are checking on the latest updates there. Some of those products are being moved around or retired in some instances, but the idea with that is you’re able to have AI identify sensitive information in a prompt.

And in some cases, it can mask that data when it goes to the LLM and then demask it when it comes back. Again, there are caveats there. So make sure that you’re looking at your specific use case to determine if that applies, because in some cases it does not apply. But that’s the idea behind it is making sure that we’re basically trying to be as safe as possible around our data and who can use that data and access that data. And if they can train models on that data, generally, if you’re in the enterprise world, there are protections against that.

even if you’re using a frontier model on its own and you’re on an enterprise plan, typically the agreement is that that data is not going to be used to train models. All of this though is still pointing towards having a human in loop as essential. And so we always want to make sure that any AI output that’s getting generated is not being inherently trusted, that we are always evaluating those outputs. We have a human who is an expert that is able to verify and if something is wrong,

can go and modify and change. Because ultimately, we are still owning these responses, right? We can’t blame AI for coming up with something that isn’t true or unverifiable. And so we still want to make sure that we have a human in the loop that we’re verifying and that we are not taking it at face value, so to speak. those are kind the main points. And then the other piece there is when we talk about the word agentic, that gets thrown around a lot. I think the word agent is all over the place.

And I would hazard to guess that most use of that word doesn’t actually apply to agent. So just saying, Hey, I’m chatting with an AI agent. Well, is it, is it actually agentic? it performing autonomously on your behalf? Does it plan? Does it actually come up with a plan and then execute that plan? If it’s not doing that, it’s probably not agentic. But the main thing there is when we talk about a system doing something autonomously.

It can’t be chaos. It has to be ordered. We have to come up with the right process. And so I think the difference there is AI agents are great when we have ambiguity around the outcome. But if we have a deterministic process, an AI agent is probably not the right tool for the job. And so it’s all about figuring out, does AI actually need to be used here? Or is this better suited to something like custom code or a declarative automation tool like Flow or HubSpot workflows?

So that’s a really important question to ask when it comes to, should I use AI or not? Let’s think about how deterministic the process is. Let’s think about how verifiable the process needs to be. And that can help in sort of a decision tree to determine if AI is the right tool for the job.

Eric Housh 

Yeah, I think it’s always, you know, considering things through that risk lens, right? From pricing, if you don’t understand the pricing and how these things are set up, you could be exposed from a cost perspective. If you don’t keep a human in the loop, you know, you could be exposed from a chaos perspective or, you know, having things run sort of rampant and out of control. And obviously, if you haven’t thought through…

 the sensitivity of your data, then you could have this proprietary data out there training these systems and being exposed when it shouldn’t be. I think that’s without going into a full on sales pitch, I think a lot of the value that companies like ours and we try to provide is helping people think through that risk and how do we really apply AI to increase that leverage without falling into some of those pitfalls.

Zack Terry 

Yeah, and frankly, Eric, that is, I think, one of the biggest value propositions of using something like Agent Force or HubSpot Breeze. You are getting some built-in security because these companies are building tools that are designed to be secure. Whereas if you’re just going straight to a frontier model and you’re having to build all of that scaffolding yourself,

There’s a lot that you have to think about and a lot of engineering that you have to do to even approximate the same level of enterprise security that you’re going to have with these tools that are building AI on top of what is already a secure platform. Right. And so that’s just another thing that I think is really important to consider. You could probably get some cost savings by going out and using a frontier model just on its own.

But are you going to get the privacy considerations? Are you going to get the agreements around enterprise data and how these models are using that data? Probably not without jumping through some serious hoops. So I do think that that is a major value proposition, especially if you’re already on one of these platforms and you’re thinking about implementing AI. The fact that these companies are building it with enterprise security in mind, I think should lend a little bit more peace of mind.

Granted, we are delving into extremely bleeding edge technology. So there’s always risks involved, even when you’re talking about enterprise scale businesses and tools. But using those tools with well-established companies that have a long history of serving enterprise customers, I think is a major value proposition when you’re thinking about implementing this bleeding edge tech.

Eric Housh 

Yeah, it’s a great point. And the fact that there are literally hundreds of thousands of other businesses out there using the Salesforce and HubSpot tech, weighing in with those guys, giving them those sorts of ideas on how those tools can be better. That rising tide is gonna lift all boats. Zack Terry, as we draw a little bit closer to our time for close here.

The idea with this episode of the Fast Slow Motion podcast was to introduce the AI in Action podcast and talk a little bit more about our discussion topics over there. Most people that listen to podcasts don’t have infinite time to catch all the episodes they want. So let’s make it really easy for our listeners out there. FSM now, we have these two podcasts. Who should really listen to what?

Zack Terry 

So the Fast Slow Motion podcast, that’s all about leadership, scale, and strategy. AI in action is more about tactical AI execution, examples, discourse on the latest AI events and happenings in the news. So one helps you build the right business. The other helps you learn about and apply AI to that business. So I think together, they are a fantastic complement to each other, but they serve two distinct purposes.

So if you’re interested in both of them, I think you’re going to get both philosophy and practice in terms of building a scalable business as well as thinking about how to apply AI to that business.

Eric Housh 

And what we always try to do because we try to stay respectful of the limited time that our business leader audience has is, know, there’s transcripts, there’s show notes, there’s resources linked on each of these. So if you don’t have time to consume the full episode, you can get the Cliff Notes versions and the key takeaways just by visiting fastslowmotion.com and their respective podcast pages for these shows.

Zack Terry, we like to close the FSM podcast with the same question for all of our guests, but we’re going to flip it up a little bit this time. For leaders hesitant about AI, what’s one piece of really impactful advice you’d give them?

Zack Terry 

Well, I’m not going give you one. I’m going to give you a few. Yeah, there’s a few things to consider. The first thing I would say is just take a deep breath. We’re all figuring this out as we go. We like to use the analogy that we’re building the plane as we’re flying it. And that’s true for all of us, right? Nobody is just astronomically ahead of somebody else. We’re all figuring this out together. That said, if you’re thinking about implementing AI,

Eric Housh 

Yeah, love it. Yeah, yeah, break it down.

Zack Terry 

my encouragement would be to start small. Don’t try to skip all the way to the performing step right out of the gate. Start with support and think with execution. Look at one to two quick wins and think about how you can apply AI to that process, but make sure you understand the process first and then measure it and determine if it adds value. So in the same vein, don’t try to do too much at once. What you want to achieve with AI is probably a much bigger project than one to two quick wins. And that’s okay.

Because you can stack those wins on top of each other. What you learn with those quick wins will help inform the larger wins that you want to achieve down the road. So think intentionally about how AI can assist. Understand what AI is good at and what it’s not good at. Understand when a process is deterministic and requires an exact plan in order to achieve something. If A, then B.

That’s probably not a great use case for AI. It can be. You can plug AI into that process. But when determinism comes into play, be thinking about how more standard automation can fit in. And where there’s ambiguity in the process, think about how generative AI may be able to assess. And then finally, and this is a big one, don’t just assume how AI should be implemented at your business. Pull your teams first.

talk to your employees, understand how they are already using AI because most of them are probably already doing this in some way. So spend some time thinking through how they’re actually using it today. And that can help inform.

some potential high impact areas that you could use to implement AI securely within your business, as opposed to having people using it outside of the confines of the systems and tools that you have. So definitely spend some time with your employees, understanding how they’re using it. And if they’re not using it, understand why. Is there a reason that they haven’t started experimenting with it? Have they?

not really come to understand what parts of their particular job AI might be able to help them with. And that can also help just as much as understanding how other employees are using it.

Eric Housh 

It’s a phrase we use commonly when we talk about technological transformation, which is crawl, walk, run, right? You don’t have to boil the ocean. You don’t have to solve everything that’s wrong with your business. You just have to start somewhere. You have to give your people permission to experiment. You have to have the right processes and systems and guardrails set up to make sure that the technology is going to work for you. On AI in action on that show, we are about to start diving into real.

business use cases across industries and breaking down what’s actually working, what’s failed, what’s producing ROI. We’ve gotten now scores of businesses that we’ve helped with this AI sort of transformation and figuring out how AI fits into their different business models. So if you want to take that leap from theory to proof, head over to fastslomotion.com. Under the resources tab at the very top, you can find that AI in action podcast.

That’s also on all the major podcast platforms and on YouTube. Go check out a couple of episodes. We’d really love to know what you think about that show and the content. Drop us a comment, leave us a review, give us the thumbs up or thumbs down. And definitely if there’s any topics that we haven’t hit on yet that you’d like to see us dig into more deeply, let us know there. Zack Terry, it’s always a good time hanging out with you, man. Thanks for joining me today.

Zack Terry 

Great to be here, Eric. Thank you.

Eric Housh 

All right, see you guys next time.