Welcome back to AI in action. I’m Eric host joined as always by Zack Terry Fast Slow Motion’s director of AI. And today we’re tackling one of the biggest challenges every leader faces. And that’s people. How do you find the right talent? How do you build a culture that thrives in this age of AI? To help us unpack that.

Man, I’m fired up. We got a special guest joining us today, Matt Dyar, chief operating officer of Fast Slow Motion. Matt has been driving culture at scale, and he’s seen firsthand how AI is reshaping both operations and strategy. Matt, so awesome to have you here, man. Thank you Eric. Happy to be here, man. Cool. Well, Matt, before we get into today’s topic, let’s talk about what’s hit AI this week.

What’s in the news? First of all, Anthropic made some headlines. Are Claude right? They did. Yeah. There’s, an interesting sort of capability that they’ve added to Claude, where it can identify quote unquote, abusive conversations, which what constitutes an abusive conversation. I think that’s an interesting question in and of itself, to say, how is a machine categorizing something as abusive and what are the implications of that downstream?

Right. If anything can be categorized as abusive, are we going to be cutting off access to these tools? I think the intent here is really around AI wellness. Obviously, we don’t want to give people tools to do things that they otherwise may not be able to do. And so, you know, you think of the classic example of someone who’s trying to build a bomb, and the AI is giving them the instructions on exactly how they can build and deploy that bomb.

Right. Obviously, we don’t want to be able to distribute that information. So things like that, I think is the angle behind this, right. Like how how do we prevent that information from from going out? I think the problem still remains though. It doesn’t really matter how many guardrails you put around this. There’s still a risk of prompt injection, of overriding those instructions.

We haven’t really solved that problem yet. And so this is kind of just, you know, the natural progression of trying to figure out how to solve a problem like that, how to prevent the the tool from actually getting responses that we don’t want it to give. And I think that the controversy here is, is, is people basically thinking about this, as are we thinking about a tool like Claude or an Lem as a sentient being, which of course it isn’t.

It’s it’s just a bunch of math and it’s a prediction model. Right? But, you know, when we start anthropomorphizing something like this, I think that’s where the controversy comes in, where we start thinking about these tools as human like, when in reality they’re not. And so I’ll be interested to kind of see kind of where this goes, obviously, you know, let’s be nice to our AI so that if it ever takes over, we’re not in the Terminator.

Right. But again, these are these are prediction models. They’re not they’re not sentient. They’re not thinking. They’re not self-improving. Right. And so, I think it’ll be interesting just to see, can we solve this problem? Is this a next step in the right direction towards being able to establish guardrails that work, or are we still going to be working on this, you know, six months to a year from.

Speaking of sentience, Elon, in the news, recently saying crack five has the potential to become sentient to be AGI. What are your thoughts on that? Well, I think any time we’re talking about a large language model and the word sentient in the same sentence, we’re probably wrong, because I don’t think we’re going to see large language models be the technology that takes us to AGI, and that’s probably an entire conversation and, and podcast that we can have over time.

But ultimately, these tools are prediction machines. They’re not sentient. They can’t think they are just mathematical, right? They’re they’re putting out a probabilistic response. So will the next Grok model be a better model than the last? I’m sure it will. Just like chat, GPT five was mostly a better model than the models before it, right? But these are really just incremental improvements.

These are mostly like getting the next iPhone every year. We’re not really seeing monumental changes in the capabilities of these models. We’re seeing small improvements. And I want those improvements. Right. Like I want the next iPhone. But I’m not going to be floored by the design or capabilities of it. They’re going to be improvements and things I already use and expect.

That’s what I think is probably the reality here. Of course, you never know, right? There could be something crazy that that pops out of Grok, but I don’t think, if I could pick any model to actually achieve something like AGI, I don’t think I’d want Grok to be the one. So. So what you’re saying is Elon was a bit hyperbolic.

This is I would I would probably argue towards that. Yes. Cool. Let’s talk about the business side. SoftBank making some moves. Lots of investment money nothing new. Of course there’s a lot of investment money going into these tools. But SoftBank is a part of this Stargate program which along with OpenAI, Oracle and MGM has more than $500 billion available to invest in AI infrastructure.

And so they’ve deployed some of that capital to the Foxconn plant in Ohio for $375 million, with a focus on developing AI infrastructure in that plant. And so, yeah, this is this is just a part of that program. We’ll continue to see, I think a lot of acquisitions or maybe net new construction for data centers and things like that for this $500 billion, you know, fund that’s essentially been established by these companies.

Speaking of acquisitions, our friends over at Salesforce have have had some activity lately. What’s the, what’s the status of that two relatively new I won’t call it an acquisition. I will call them definitive agreements to acquire, because obviously all the paperwork has to get completed and the checks have to clear and everything has to be approved, to two companies.

This happened a little bit, longer ago. It’s not super recent, but there’s a company called way or Y cell Y AI, and this is a natural language to SQL query tool. So I can essentially describe what I want in this tool will turn it into a SQL query to go out and actually retrieve that information, which is really pretty great.

If you work with big data, you’re probably working with SQL in some way, and it can be difficult for a non-technical person to get involved with SQL and to go in and start working with those databases. But if you’re able to describe that and then have that turn into a query, that makes it a lot more usable. So I think the idea here for Salesforce is to be able to use this with tools like Data Claude.

There’s a lot of different query languages in Salesforce. Data Claude does use SQL. And then we have the SQL cycle. The Salesforce Object Query Language. There’s an Analytics Query Language as a SQL. So I don’t know if this will help with sort of those more proprietary languages. But this will definitely help with things like data, Claude, where we’re already seeing a ton of investment there because Salesforce is building sort of its entire Agentforce product on top of data Claude.

And so that will only become more important. So to to make something that could be difficult for non-technical users a little bit easier, allow them to query things with natural language. I’d love to see that come to the platform. So I’m excited about that one. And then we have a company called Rob Grillo, and they describe themselves as a generative AI operating system, mostly focused in the manufacturing space.

And so the reason Salesforce acquired this company seems to be specifically focused towards augmenting the capabilities of Agentforce and Slack. I have no idea kind of how that connection will actually manifest, because it’s an interesting sort of angle to take a company that’s focused on manufacturing and then apply that to, communication tool that you’re working with across your business.

So there must be some kind of something under the hood there that’s really promising in terms of that automation pipeline. So be interesting. Interested to see where that goes in terms of, hey, how do we apply this to slack and how do we apply this to Agentforce. So again, definitive agreements have been signed. I don’t know when those will actually be finalized, but I think, you know, anything with Salesforce, it can take, you know, a year or two years for for something to go from acquisition to actually seeing that in the platform.

So we probably have a little bit of time to wait to see what’s going to happen here. But I like the idea, especially with the natural language to SQL capability. It seems like acquisitions are just kind of risky because the technology’s moving so quickly. But the guys over Salesforce a lot smarter than I am, so we’ll trust that they know what they’re doing with.

That’s something like the AI space and how quickly it’s moving. I mean, you know, you kind of get that classic business problem of how do you maintain innovation, right? And you’re you’re doing that in-house or you’re acquiring companies to do that. So this is kind of one of the primary growth strategies of mature enterprise based companies to go out and acquire companies that are small, they’re innovative, they’re agile and nimble.

And instead of having to develop that in-house using what has become a lot of probably bureaucracy and red tape, you’re able to take advantage of that innovation by purchasing it and then layering it into your platform. Families like MIT coming out with a study this week saying 95% of AI projects are failing, failing to deliver return. What do you make of that?

I think overall, this is not terribly surprising. And there’s there’s a lot of reasons behind that. But there is already a metric that suggests something like 70 to 80% of general I.T projects fail. Just general. It’s not even AI specific. So it’s not that big of a jump to move from 80% to 95% when you consider that this is a very new technology.

Also, this technology is complex. It’s not easy. I think that it is really easy to go and chat with one of these frontier models, OpenAI, you know, ChatGPT you open the chat window and you chat with it is an entirely different story to layer in those capabilities into an enterprise in a way that works with your business, it works with your process, and it’s done in a secure fashion.

There’s a lot of complexity there, and you really have to think about it from an entire holistic architectural point of view. It’s not one of those things where you’re just going to plug this in, and you’re immediately going to start seeing the benefit. You really do have to think, how does this apply to the entire architecture? How does this apply to our software development lifecycle?

How are we going to deploy it? How is it going to be secure? And so I think what’s happening is, is a lot of businesses are starting to realize that it is a lot more complex than they thought. Right. And I’m speaking specifically to companies that are building their own AI. They’re building their own capabilities into their architecture.

This is, frankly, I think one of the big benefits to a tool like Agentforce or Breeze when you talk about HubSpot layering AI capabilities into a platform, or Salesforce layering AI capabilities into their platform, you are not having to what Salesforce would call DIY your AI. They’re giving you the tools to use it and scale that across the enterprise.

They’re already thinking about things like security. You have the ability to control things on that security model. So I think I think there’s kind of a different way to frame this. If you’re looking at buying tools as opposed to building tools, I would be surprised if the buyers of tools were not more successful than those that are building their tools.

I don’t know where the details are exactly within sort of that research that MIT has, has put out. But it’s not surprising it projects fail all the time. This is a new technology. It’s a complex technology. So I think overall, if you’re interested in still doing this, but but this sort of statistic is a little bit scary.

Yeah I agree it is a little bit scary. But I think the same things that we’ve been recommending still apply, which is start small, identify those potential high impact use cases, focus on some wins for, you know, things that maybe your employees just really don’t enjoy doing right. Look for areas where AI has a high probability of success.

And don’t fear that hype cycle, right? Start small and start scaling just over time as you see success and you start to count on those wins, right? So basically don’t over invest at the beginning. Do what we’ve always done, which is iterative software development, figure out what’s working and what’s not working, and don’t try to do it all at once.

Absolutely. All right. Well let’s let’s get to the main topic today. And that is how do you find the right talent in the age of AI? So, Matt, when it comes to hiring today, where does I fit and where does it just not fit? So I think a lot of people listening may maybe expecting me to give a silver bullet answer right here.

You do this, this one AI thing, and it takes care of all of your hiring. And you get and you get a perfect organization. And the reality is we just don’t think that’s the case. So I doesn’t change the fundamentals of hiring. I think when you’re when you’re hiring for culture, you are hiring, the same team members that are going to be, impactful in an AI driven sort of environment.

So at a 30,000ft view, let me back up and say, what are you trying to accomplish with hiring in an AI? In an AI age? And I think it is. You are trying to create a nimble organization that can adapt with all of the change around us. Okay. And and the job with AI is, you know, by the time you talk about it, it’s outdated.

But what we’re going to kind of unpack today are some timeless principles that I think whether you’re listening and I’m just going to say it, whether you’re listening to this podcast in 2025 or 2035, I really hope what we’re talking about applies. You like that, Zach Dobbs Yeah, I do, I do, yeah, yeah. Because we know it’s going to be a whole different set of tools by that.

I guess it is. But but principles can be timeless and and at fast. So motion man, we love a framework. We love our frameworks. And we came across Patrick Lynchian, his book, The Ideal Team Player a number of years ago. And man, it just really, really made sense. It allows us to be really intentional with our hiring processes.

And for any organization out there that is, you find yourself hiring people that are just taking jobs, that are jobs. We would really encourage you to get intentional with your hiring processes. Get intentional with the characteristics that you’re looking for. And we’re going to unpack three of them today. Three characteristics. We’ve we’ve built our entire hiring model around these three principles.

And I’m going to give you a fourth, which I think is very impactful in the age of AI. But this concept of of hiring an ideal team player, you need these three characteristics, not just one, not just to what you’re what you’re trying to do is look for someone who embodies these three characteristics. And they are people who are who display humility, people who are really, really hungry, and people who are smart, meaning interpersonally smart.

Now let’s let’s unpack those just really, really quick, when we talk about finding team members who are humble, we’re talking about people, who are more interested in the team than they are interested in themselves. And a natural byproduct of humility is coach ability. And, man, I cannot stress this enough. In this day and age trying to create an adapting organization, you have to have coachable team members and and looking for humility is one way to do it.

A coachable team member is going to be one of those people that is just willing to learn. I they’re open, they’re a dry sponge, if you will, and that’s really what you want. You want somebody who doesn’t have it all figured out. You want somebody who who is is humble enough to take coaching and, and adapt with these new tools.

The second principle is hunger. You need strong work ethic, and people who display hunger and their work. I really think you have to align well with the mission of the organization first. But when you when you hire someone who aligns with the mission of your organization, they naturally work harder. And really, what we’re looking for with hunger are people who are who are willing to, just be gritty and go out there and work through these tools because these tools take take a good bit of effort.

You don’t just pop it open and get perfect results. Sometimes you you have to be pretty arduous with these tools. You got to not only test different tools, but you got to get to know these tools and let the tools get to know you. And a really hungry team member, looks forward to these challenges. And, and they don’t have problems adopting new tools.

And then third, you look for smart team members. And we’re not we’re not talking intelligence here. Intelligence doesn’t hurt. But what we are talking about is those interpersonal skills that people have the ability to read a room. And what happens when you can read a room and you have high IQ? You become an influencer within your organization.

And this is this is really, really powerful. You know, you find champions within your organization who are humble, who are hungry. And now all of a sudden, they they’re highly influential within within the organization. And you just have a winning combo. Now somebody listening may say, that sounds awesome. That sounds great. Humble, hungry, smart. I have no clue what you’re talking about.

Or how do we hire for these concepts? And I would say a couple things. One, Patrick Lynch, you only has an amazing book just called The Ideal Team Player. It’s an easy read. I recommend that you consume it. Have your H.R. Manager consume it, have your C-suite consume it. But then I think in the last week or so, the table group, which is Patrick Lynch’s organization, has come out with a new assessment called the Ideal team player assessment.

And we can link that down in the notes. But I would say read the book, adopt the assessment. It may be something you want to start panning out to your candidates during the interview process, and then get to know interview questions that unpack humility, hunger, and an emotional intelligence. And we can link that as well. There’s a lot of great resources online that give you about a half a dozen questions around each one of these, and I would say start adopting those into your your hiring practices as soon as possible.

Now, I told you I was going to give you a fourth characteristic. This one isn’t in the ideal team player, but something that has just become really, really important. And that is the art of curiosity. Now, a lot of times when you find somebody who meets those three characteristics, they are just naturally curious. But I would say verify, trust but verify.

And one of my favorite things to do in the interview process when we’re getting to know candidates is flip the script, flip the script on a candidate, and say, what questions do you have about the company, the role, or myself? Now the right answer is, oh my gosh, I’ve got a whole list of pages that I would love to ask you.

And usually for the right candidate, you’ve got to cut them off at some point, say, okay, you’ve asked 3 or 4 questions. We we got to move on. But be very, very cautious of the individual says, no, I’m good. I don’t I don’t have any questions. I’m perfectly fine. Unfortunately, a lot of times that’s a telltale sign that they’re very transactional.

They’re looking for a job, and they’re probably not aligned with your organization. But, boy, if you get somebody who’s curious and they are going to be natural tinkerers and the age of AI humble, hungry, smart and curious, it makes a lot of sense, right? The tools are going to change. The tools are going to evolve. It’s less about that.

It’s more about, hey, let’s hire the right people. That given the fact that this technology is moving quickly, can apply it in interesting ways, are curious about how it can change your job, streamline their job, help them connect more with their team, help them work more efficiently and better. Absolutely. Let’s let’s hit the second part of this and we find these people.

We’ve got systems now where we can evaluate them as we’re having conversations in the hiring process. We bring them on board. We we put them on, as a part of a team. How do you empower that team to build culture? And what does I have to do with that? Man? And power is the word. That is the job of leadership and the age of AI.

You have to empower your people. And we believe you kind of. I’ll keep it really, really simple. You do this and one of two ways you do it officially and then you do it unofficially. So officially you need a policy, right? If you have no other takeaways from this episode today, it’s if you don’t have an AI policy, you need to develop one.

Now, I could probably run a whole episode on how to develop an AI policy, but what I would actually encourage you to do is use AI to develop your AI policy. Have a conversation with Gemini or ChatGPT or Grok five. When it comes out. Grok 5th May just do it all for you. From what we from what we understand.

But talk with AI, tell it a little bit about your business, about your people and what you’re trying to accomplish. Talk about your data. Talk about your security. Ask it to help you develop an AI policy. So I think, first and foremost, officially, you have to put that in place in order to then empower your team. And then there is the unofficial empowering of the team.

And this can be done in so many ways. But but if we really just boil it down to brass tacks, you have to be talking about AI. It has to be part of your daily conversation. In a previous episode, you heard John Burdette talk about fast, slow motion has really started. This AI-first sort of strategy. And if you haven’t listened to the episode, go back and listen to that episode with John.

But John, very quickly, very early on with our leadership team, developed this AI-first mentality, and that has just been pushed down throughout the organization as quickly as possible. It’s not a trickle. It’s a it’s a it’s a push that AI-first mentality hard down, throughout the organization. And so as a leader you need to be using it.

You need to be talking about it. You you need to be visible, in your utilization of AI. And then, man, there’s so many different ways that you can empower the team. One thing that we did very early on was we had an AI prompt competition. How are all the ways that you’re using AI and you have to give permission to your team in that policy.

You got to let them know you want them to be playing with AI tools, specifically which AI tools you want them to be utilizing. But once you lay out that official policy, you kind of open the floodgates a little bit. You kind of you you got to let go of the reins a little bit and let your team play with these tools and figure out the boundaries and how it can help them in their and their daily life.

And then and then you go, hey, Eric, man, that problem that you’re working on. I was talking with Zach the other day and he wrote an awesome prompt, why don’t you all get together and talk about that? The work he’s doing may be able to help you. So you you’ve got to start connecting people within your organization who are using who are using the tools.

A couple other things that come to mind. One thing that, John Burdett had the foresight very early on was find a champion inside of your organization. That’s pretty obvious. Zach is our champion, and he’s doing an amazing job, pushing AI utilization out throughout our organization. But it doesn’t even have to be an official title change.

A lot of times inside of a company, just find either a person or a group of people so you can kind of deem a champion. Maybe you put together a weekly call, but you empower them to kind of, be the eyes and ears of your organization and let them kind of carry the torch forward. And then third, I think training is very important.

Now, this may be a little tricky if you don’t have that champion, but, but, but once you have the policy, once you have a champion or a group of champions, really start empowering them to train the team. Find some opportunities, whether it’s weekly, bi weekly, even monthly. Just get your team together and talk about I share. I talk about the state of AI.

If you’re able to do that, you will find your people empowered. You will find your people utilizing AI, and you will see that culture flywheel start to grow. And we’ve seen, a lot of examples internally, like, in the company training, recently, Michael Smith, our director of people, he talked about how we take the assessments that we, that we ask people to complete in the hiring process, and we’re able to put those into an AI tool.

And the AI tool will surface, insights, analyze those patterns and personality dynamics of leaders and their team members and their, their, their cohorts, their and, and help them better relate to each other, help them better connect to each other. So, Matt, how is that work? Like, without replacing human judgment, how is the tool augmenting that process?

Man, it’s it’s helping us dive through these assessments quickly. Like, can you do it without AI? Yes, but a lot of these assessments that we’ve, we’ve done on our team are like 50 pages long, like it’s a ton of information, which is, which is it’s amazing information. But when you’ve got a question like, how are these three people going to communicate well with each other or a team of eight?

How how is this team going to, be cohesive? How are they going to come together? How do they how do they understand quickly the idiosyncrasies of each other? You know, sometimes that’s 3 or 400 pages of information that without AI, you you’ve got to plug away and go through it. And we’ve just found, when you when you capture information about your team, which we really love, behavioral insights and motives.

It’s a DISC-based approach. We just had a podcast that had Dr. Abelson sit on it, where Michael Smith and him talked about, talked about all kinds of, behavioral assessments and how to use utilize them inside of your organization. So go check that out. But once you have the data, that tool just man cuts down the time frame and actually helps us find insights that we might have missed with the naked eye.

So I does a really good job when you load these assessments and tell it the pain point. It’s really able to come back quickly. We’re we’re talking seconds here. Not about hours like doing it doing it manual. We’re talking seconds. It’s able to bring that information back. And and I’d say there’s two areas where we’ve really seen this impactful when new team members join the organization and they get plugged into a group, we quickly empower the leader to understand how to lead this person properly.

And then two, it doesn’t happen often, but every now and then, you know, two team members may, they just rub each other the wrong way. You know, there’s friction between them. They’re, they’re at odds. And you need to get those people back into a good place. A lot of times, what we’ll do is take both of those assessments, feed it into the system, talk about the pain point, and then coach the leader on how to mediate those team members so that we get back to a good place.

So onboarding and mediation are two places where I and and these assessments are wonderful to utilize inside of the company. And you said something there. And I want to hammer that point home. We talk about the human in the loop, and the fact that, yeah, this analysis, these insights generated by AI are extremely helpful can really shorten that cycle versus us having to read through it.

But ultimately, the leaders have to interpret that data. The leaders have to coach. We have to act on that data. Asking AI to do that obviously not possible right now. But I don’t think we want to. Right. That’s the soul of the company. The soul of the culture is, hey, we have this data now. Let’s talk to our people.

Let’s figure out how to make them better. Make them work better together. Yeah. I mean, ultimately, the user is responsible for the output, right? So if you take something that you get from AI and you don’t vet it and that goes out into the world and it’s wrong, well, that’s my fault. It’s not the A’s fault, right? That’s my fault for for trusting in and not verifying and not vetting it.

Right. So I do think that’s really important. And the human in the loop concept is, is important for, for those type of examples. But especially when we start talking about a genetic AI where A’s doing things on its own, that’s where it’s also really important to have humans in the right place to be checking those throughout the process, as opposed to allowing it access to your entire system and kind of, you know, giving it the ability to just run wild and do whatever it needs.

Very important to build out kind of that architecture and keep those humans in the right places. So, I think in terms of, hey, how can we build a culture that is conducive to using AI and experimenting and being curious and using it in the right and safe kind of way? We we touched on some of these things, but this idea behind sharing props and sharing ideas, I think it’s much bigger than just, hey, here’s like some magic words, a recipe to go and plug into AI.

It’s about thinking about those ideas in ways you might not have thought of before. So, for example, Michael Johnson on our team, he uses AI in a way that I never would have thought about if he hadn’t shared that prompt. He uses it to simulate focus groups, which I think is super interesting. Yeah, that’s just such an interesting concept, right?

It’s so different than just asking a question back and forth and receiving an answer. But I would have never thought about that if he hadn’t shared this prompt that he’s using and how he’s getting the tool to work in that way. And so it’s more than just a set of instructions. It’s thinking about ideas and sharing ideas and really fostering that, culture of collaboration and sharing.

And then the other thing that we do on a regular basis, we’ve already done a couple of these this year where we do this outside of AI at the company, which I think is really important. We do these show and tell. Like, what have your team been working on? What did they accomplish? What are some cool use cases or maybe some out of the box solutions we’ve created?

And we’ve been doing that with AI specifically. And yes, it has to do with Agentforce and HubSpot breeze and things that we’re building for customers, but also it’s how are you using AI in your day to day job? And so we’re sort of explaining and showing off, hey, this is like a workflow that I’ve been using with these AI tools, and this is how I’ve done it.

And these are the results that I’ve gotten. And so I think sharing that across the company is a really great way to just inspire this culture of experimentation and, sparking some ideas that maybe you didn’t have if we weren’t sharing that across the company. We also do these regular briefings where we’re talking just about the state of AI.

You know, what’s going on in the news, what tools are available? Are we using the right tools for the job that we want it to accomplish? And then of course, having having that AI champion. Matt, I know you said you said that I’m the champion. I would even turn around and say we have multiple champions at the company, right?

Every everyone is is really sort of thinking through how to use these tools. And, I would say, you know, Eric and Matt and Chad and John, I mean, everyone is really championing this from the leadership teams or across the board, right? So, I do think it’s it’s it’s key to have alignment across the leadership team and across the company.

One person can’t be the sole champion. And so I think I think the idea here is really to align on how you want to communicate this and what your process is going to be in terms of establishing a culture that is curious and willing to experiment. Amazing. Zach. Matt, it’s been wonderful today. Matt. As we’re wrapping it up, any final thoughts?

Stay curious. Stay curious. Matt. That’s it’s perfect. It’s perfect. So we’ve talked about today, finding the right people is all about finding the right people. I the tools they’re there. But really hiring for adaptability, hiring those people who are humble, hungry, smart, and above all, curious and empowering them to thrive. We’ve got a lot of tools.

We’ve mentioned a lot in the show notes. We’re going to link them down there. If you’re if you’ve got questions about a, the things we’re talking about, maybe resonating with you, but you’re still feeling kind of lost. We’d love to have a conversation with you. Just head over to fastslowmotion.com/ai. That’s the quickest way to connect with us, and we look forward to the conversation.