Dean Carter:
There is a lot of AI washing, I think. If you're designing something to replace humans, I think that's gonna scare people. There are things about that HR business partner job that are so human that you need, now there's 60% of it maybe that isn't, and you can free them up to do a hundred percent of the time the work they've always wanted to do. But that, I think there's a lot, you can't just tag AI on something.
Steve Smith:
Hey everyone, welcome back to another episode of Work Tech Weekly. I'm Steve Smith, Managing Director of Growth at Rep Cap.
AI is quietly making the hardest parts of a job the most valuable ones. As it takes over the technical, skills based work, what's left over is judgment and the ability to read a room. That is turning out to be true at every level of a company, all the way up to the CEO seat.
My guest today is Dean Carter, CEO of Instill, a culture tech platform that skips the annual survey and reads what is actually happening inside meetings in real time. Dean spent 25 years as a head of HR, including a long run at Patagonia, before making the switch to CEO. He is proof of his own argument. CHROs are some of the most qualified people in the building, and almost none of them get a real shot at CEO. He thinks that math is about to change.
In this conversation, we talk about why he thinks people leaders are the next wave of CEOs, how Instill tries to measure something as fuzzy as trust without turning into a surveillance tool, and why he is convinced you cannot fake real intelligence behind an AI product no matter how good the demo looks.
If you have ever wondered what actually survives once AI eats the easy parts of a job, this conversation will give you a clearer answer.
Let's get into it.
Dean, welcome to the podcast.
Dean Carter:
Thanks, Steve. Looking forward to it.
Steve Smith:
I'm really excited about this conversation, Dean, because it's just like your track record as a people leader is pretty exemplary. I mean, you've been one of probably the most recognized people leaders in the country over the past 20 years. You've made the switch. You've moved from essentially head of HR over to CEO. Walk us through that decision. Why was now the right time to do that?
Dean Carter:
You know what? This is not a job that I sought or thought, you know, 25 years ago when I started as a head of HR, this would be the eventual path. I thought I'd just like glide into a head of HR for the rest of my life. My fourth round into it, I began to think. I got really close to CEO a couple times.
People like, oh, we're thinking considering you for this. And I was like, I don't want that job. That's for crazy people. Like anyone, like you gotta be nuts to do that job, 'cause they just saw so many CEOs who really got... really, it was a... it's a hard job, and I saw them get destroyed in the job. And so I thought, that's not what I want.
Then I began to think, as I reinvented the head of HR, I feel like I thought, well, what if I reinvent that job? I didn't have to destroy me. And I started putting it out like, I think this is something that I want to do next in terms of my next pivot. I've done this for the past 25 years. The next 20 could be something like that.
And I just kept putting it out. I'm interested in this, and all of a sudden it found me. So, this company was built by Hamet Watt, who's built several other companies, and he builds them and then he finds someone to run them. And he called me and said, hey, I built this amazing company, we capture culture.
I was like, wow, that's what I've been trying to figure that out for a really long time. And he goes, I want you to run it. And I'm really, this is a real good intersection because culture is super important to me. I'm really into data and how to drive business and connect outcomes through data.
This is the first time I have a piece of data to connect culture and behavior to business outcomes. I'm fascinated with AI, like I just, I'm not scared of it all. I recognize the terror that it could create and the havoc that, so I'm not. You know, not unrealistic about it. I think we need to put some, you know, guardrails around, you know, the broader concept of AI, but the specific, you know, use of it personally and for work.
I think it's one of the greatest opportunities ever to do really things that we've never been able to do. And I think particularly for the HR profession and for things like culture, it's been really hard to measure these things. Finance pretty, you can measure a number pretty easily. You can look up a law pretty easy.
Now HR, like the fluffy things like trust and belonging and psychological safety felt, you know, fuzzy, and maybe they're not gonna be so fuzzy anymore. That's why I'm doing this.
Steve Smith:
Well, and so I mean, what's interesting is sort of the conventional wisdom is that HR leaders don't make it to the CEO job. And you've said yourself that CHROs are probably the most qualified people in the room who never get a shot. Why do you think that it was different for you?
Dean Carter:
Well, I think the world has changed. I did have an opportunity at Patagonia when for the job I took for almost a decade. It was general counsel and CFO did report to me during that time. So I did have some experience, and my background is operations before HR, so had some experience in it, but not entirely.
I just think the moment is right. You know, things that were really focused on technical skills, the technical skills of a CFO, the technical skills of a general counsel, and even the technical skills of HR, when you're left with what's left in terms of just the human part and the EQ, the CHRO actually is left with a lot more substantial role than, to some extent, even finance and general counsel.
I mean, I can build a pretty sophisticated workforce planning and financial analysis program myself. Now I have access to pretty much any piece of legal information I need and advice for my, you know, legal agent. Now that doesn't replace, you know, those roles, but what's left after skills and tasks and technical work is, you know, AI can do.
I think this is the job at the senior level that is gonna be more important than ever, especially for the CEO too. You're, I think, the human part. And the ability to navigate that as a human-first company, I think is gonna be the companies that win.
Steve Smith:
Well, and I think you bring up a good point, because what we're seeing kind of play out in real time is it seems like. You know, you've heard all the patronizing comments over the years about, well, HR needs to earn its seat at the table. And those, you know, that kind of BS, but you know, where we're at now is like HR and heads of HR are effectively becoming the Chief AI Officer, because it's that intersection of technology and people, you know. And it seems like your background combined with the product that you're working on in the moment seems like, like, yeah, this is the right moment for you.
Dean Carter:
It is, I think it's gonna be right. I think there are a lot of really amazing heads of HR who have to lean into AI, because it is this, it's how work gets done. And you just happen to fall in the space where, like, okay, this combination of how work gets done is actually the most important concept.
Not how much money you make or, you know, are we, you know, sticking to the law. I think that we found ourselves accidentally at this spot, really prepared for it. I don't think 25 years ago you would say this, but I think today, and I think the level of sophistication, just because of COVID and a lot of things that have happened lately, had to increase dramatically.
I've had several people who are, you know, former other roles in the C-suite who've moved into the head of HR role, and they're like, this job is so much harder than I thought it was.
Steve Smith:
Yep.
Dean Carter:
Really hard. This is a very hard job. And I'm like, yeah. I'm like it's never off. I'm like, the job never offs and the CEO job is never off.
So anyhow, I think there's gonna be a natural movement towards finding more, you know, people with intense people experience moving into CEO. I think it's, I'm just probably the beginning of a wave, and I hope that I can prove this with this job. Hey, look, we can do a good.
Steve Smith:
Let's talk a little bit about what Instill actually does, 'cause I've had a chance to just kind of, you know, go through the website and look at, you know, the broad brush strokes of it. It looks like a really cool product. But you know, I mean, you've been, you probably had more tech companies pitch you than you care to remember, how you know it.
When you think about kind of the core premise behind it, that surveys are broken, I agree with that, and that real culture insight kind of lives in how teams talk to each other. Explain to me, like I'm a skeptical head of HR that's seen a hundred culture tools pitched to me and really wants to find out more about this.
Dean Carter:
And I get a lot of pitches. The good fortune of, you know, some people bringing me really amazing technology that I recognize, like Gloat years ago on a napkin. You know, like, hey, what do you think about this? This is spectacular. I think this is gonna be something that's needed for the future. And sure enough, Gloat is a significant player now.
Or Kelley Steven-Waiss, who came to me and said, I'm thinking about this company called Hitch, which is on fractional work and gig work. And she ended up being purchased by ServiceNow, by building another company. I have an eye for, you know, amazing tech. And the thing that I have been frustrated with the annual survey, but it was the best thing we could do.
I always thought someday we're gonna be able to understand what is the behaviors and the experiences in real time. We just never, and I was trying to fake it through, like, you know, pulse, everyday pulses, or fake it with, you know, some sort of sentiment analysis that you might be able to do with some things, which just never quite got there.
And what I found, and still, it was like, this is what I've been looking for forever. I wish I had it at Patagonia. When I joined the company, I said, why am I here? You don't even need me. The culture's amazing here. But what I will do is I want to dissect the DNA, or articulate, or understand the DNA.
So when this company grows and scales, we protect this beautiful culture. So I didn't build Instill, so it's not my baby, and so I can call it ugly if it was. Here's what I think is beautiful about it and unusual. And by the way, I've talked to Greg Pryor with Workday, who kind of invented employee experience, and Greg's like, I haven't seen anything like this before.
I showed it to him. Michael Arena, former head of HR for IBM, who's one of the most spectacular technologist and social understanding people, he's written books on this, he's like, never seen anything like this. Kathi Enderes with Josh Bersin said, I have seen every single piece of tech that comes by. I have never seen this.
This is a completely new category. So now what we have is the ability, one, we always want things in the flow of work. This is in the flow of work. You, it's not, there is no app. So everything happens in email. The second thing is it's completely passive. So it would sit in this room, in this reading right now, like it's Michael and is listening, and at the end of the meeting, all the agents that come together, you know, send a note to you, Steve, and send me to me and whoever else in the room, that is my bespoke email about my behavior.
In the meeting that drove trust, psychological safety, innovation, collaboration, and it's done using Carol Dweck. It's communicated using Carol Dweck's principles of growth mindset. So, hey, these are some things you did that were really good. Here are two recommendations that you might even consider that could be even better.
It is de-identified to everyone else. So I can't see, like, Dean got, you know, Dean didn't do well on this. You can't see that from me, or neither can my manager. So it's private. It is bespoke to me, and I can see the performance at the team level, so my manager can see how the team is doing. And this number, it's actually scored, not because the score is important, but it's more of like an index, so you can see what moves up and down.
And that goes all the way up from individual to team, to function to company. And you can see this culture move in real time, in addition to culture values as well. So we plug those in. And at the bottom of it, this is the beautiful part. In addition to getting information every day, we put business outcomes at the bottom, and we can correlate business outcomes, whether it's, you know, HR people, outcomes like turnover or retention or burnout, or business outcomes like patents or revenue targets, and compare that to which one of these are actually correlate the most of that.
This is what we've been looking for
Steve Smith:
Wow.
Dean Carter:
A very long time, and it works. We work with Berkeley and Stanford to calibrate it, and there's a significant mathematical equation behind this. So we can just say AI build this. We actually built it, tested it, and built sophisticated mathematical equations around weighting and Y 62 and Y 78.
I think it is a new category, that's gonna change and challenge, you know, the old engagement survey. By the way, it hasn't changed in 25 years. Q12, still 12, still my best friend at work, and I believe those, and I did those too, but I think it's time to rethink.
Steve Smith:
Well, you know, I mean it's interesting, because it's just like, and the product sounds incredibly cool, and you know, it's got a lot of stuff built into it. Like these, the cultural vital signs thing, where it's just like you can actually quantify trust and psychological safety and inclusion, like, in real time, you know?
But also it's just like I can put on my employee hat and go, yeah, that kind of sounds scary. How do you thread that needle of trust?
Dean Carter:
It sounds really scary. What we did is we tried to design this and we work every single day to protect the privacy of the person getting the information. So this isn't something someone can do a search on and go, you know, I'm gonna use this for a performance improvement plan because, you know, the AI agent discovered these types of behaviors, the individuals, that your individual data is completely protected for you.
That's it. And so we worked really hard to create, 'cause we knew if people thought this was a spy tool, no one would use it, or they would, you know, behave in perfunctory ways. What it does is if you're uncomfortable with getting information about your behavior, you're not gonna like this. If you're curious about.
I haven't met very many people who haven't been curious, like, can you send me what it says about my behavior? I'm like, I can send you an email, 'cause I can't get your behavior, but it can send something to you. Everyone's curious, and here's what can happen. Very few people get information about their behavior in meetings, especially senior leaders, especially because it's just hard to tell your very senior boss about their behavior and what could get better the next time.
Most of them are used to not getting feedback, so they don't take it so well. Most of them, and I say feedback, or just information about their behavior, and if I get it wrong, the risk is really high. And so this is an educated, uninterested, you know, party that gives the CEO and the executive team information about their behavior, with no one experiencing a level of risk in the process.
So I get information every day about my behavior that can get better. I give you a really good example on inclusion, and I think I'm pretty good at this as a head of HR for 25 years. It said, and you did a really good job in this first meeting about connecting to everyone here, but you really just appeared to pay attention to the three loudest voices in the room.
There were two people in the room that, based on their profile, probably had something to add in this conversation, and next time you might consider calling on them or asking them, and kind of making room and space for them. So this meeting scored really low on inclusion because of that, my behavior in it.
So the next time, the very first person I called on was one of the people that it recommended that I call on. Sure enough, brilliant, brilliant. And I call on this person now first every time. Both of them, these are my first two people I call on. And they're so valuable, and the information they have, but they've different, you know, profiles.
When the person is in Ukraine, her worry is, am I gonna have electricity and get bombed? You know, so her, you know, space and, like, thinking about what's going on is real different than everyone else's. So anyhow, there's information I get every day, and if people don't want to get better, this is a terrible tool.
It's gonna give you information you may not want to hear, but 'cause it doesn't hold back, this is not a happy, smiley AI. It gives you truth. And I think we'll just have an opportunity once it's there, 'cause we don't hide it, by the way. It's not like a secret little bot that hangs out. It's in full view.
I think people get used to it, being in the room and getting information. I was with Guild for a while there as the head of HR, and we had a coach in the room for every leap meeting, a human who did what Instill does, except for a real human. And it was awkward at first, and, like, wow. And I would get Slack or text or things from her about my behavior and might consider this.
And then we would even stop the meeting sometimes, like, okay, stop. Let's talk about what just happened that, you know, how do you feel right now?
Steve Smith:
Yeah.
Dean Carter:
And I was like, whoa. And you can do that. But then Instill, you can say, stop the meeting. Like, how's it going? Feels like, do anything we need to do better? And go, oh yeah, he might consider, you know.
A little bit more on inclusion. There's a couple people who haven't spoken up, and you might wanna ask this person, and so, okay, great, we'll continue the meeting. So you can interrupt the meeting and talk to the LLM about recommendations on how it's going. Anyway, there's just, it's new, and we'll probably, as more people use it, we'll shape it. But the next feature that we're adding next week is the ability to pick a behavior. In the January 26th article of HBR, they said the best way to change culture is one behavior at a time.
So I'm like, oh, okay, let's everyone pick one behavior to change. You click on, that's the behavior I'm gonna change. We have a giant behavioral taxonomy. You pick on the behaviors, and then in 22 days I can tell you if, or within 20 days you can see your behavior is shaping or never. My one tried different intervention.
So we can track behavior change with specific behaviors tied to that. I'd go to a learning course, a training course, a coach that I. And you can determine what interventions help the most for people with a certain profile. Really.
Steve Smith:
So here's what I'm kind of curious about. We've seen this dynamic play out in real time where people seem to be inclined to share things with a chatbot that they might not share if they were talking to their manager or a coach, simply because psychologically they know, I am just talking to an algorithm. And so, in a weird way, people feel more free to be candid.
Do you find that the same kind of dynamic is going with getting feedback from your app, in terms of they're more, maybe more willing to listen to it because they know there's no judgment there? It's an algorithm rather than if they were getting it from their boss.
Dean Carter:
Yeah, it's there. There's no judgment, and there's no bias, at, like, no, I don't have recency bias. We just had, you know, an argument, or I just had anything. There's no. It is just saying what it knows from years of research about behavior. That's it. And you can go, okay, that doesn't feel right, and that's okay, and we can recalibrate it.
But it is based on, you know, data and information. Not about, it doesn't have an opinion. Does that make sense? An opinion, it reflects years of research and data, and shares that based on what it knows and what is learned.
Steve Smith:
Well, so let's shift gears a little bit and talk a little bit about kinda like the market in the moment that we're in, because it seems like culture tech is in the middle of a pretty weird moment. You know, I think we were going through a couple of years where companies have not been investing in people tech, after maybe over-investing in '20 and '21 and '22, and you have a bunch of companies out there.
You have the Lattices and Culture Amps, and 15Fives and Glints, all kind of, you know, fighting for the same ground. What's the opening for Instill right now, against those bigger, more established competitors?
Dean Carter:
I'm real familiar with all those. I've used all those, so I'm like very familiar with, so I'm, I am actually the user. I didn't like, I find sometimes that founders are like, hey, I invented this cool thing, what do I do with it? And they come to me to go, you know, you're a user, but I'm actually a user of all those.
And those are all really great tools. There are tools, and I think they're really good at point solutions. If you wanna do, make sure you have a good, solid one-on-one in with your manager and you record all this stuff, then 15Five's actually really, really good at one-on-ones. Lattice has a very specific thing that it's good at.
I think about these as the interventions, good one-on-ones that impact the behavior that Instill measures. So these are all the peripheral things to helpfully get people to better behaviors and better experiences when they work together. We just have never been taking the things at 15Five and Lattice and Culture Amp in terms of pulses and surveys.
To ever know, get upstream from what that, or even downstream to some extent. Like, we did these interventions, we cross our fingers, and six months later we hope that works. Or, you know, research says that, you know what, good one-on-ones work, or good feedback works, or recognition works, or, you know, get information for how people feel on surveys.
All that was the best we can do. And that all individual point solutions, that all lead to the single thing that they wanna change behavior for better teams. The cover of HBR today was, you know, high-performing teams and how important they are, cover of HBR today. And you read the article, and I was like, ugh.
You know how you measure it? You hire someone from McKinsey to watch everyone one month, and they watch everyone again in six months. So it's $300,000 to do that.
Steve Smith:
Yeah.
Dean Carter:
That's still the answer. And I thought even HBR. I can't get outside of this. So it's a new way of validating all these tools. I don't think it replaces 'em, but I think it validates and goes upstream.
Engagement surveys, like we have a tool, but we do a little bit differently. I call it a Smart Pulse. So, with building a pulse, but it's based on the behaviors are in the room, so it'll go, and it's in the flow. So at the end of the meeting it goes, it might be a good idea to just ask everyone this question.
Ding, and then they get an email. It's at the bottom of the email, and, like, oh, I'll answer that. No separate app, no place to go to. It's specific about this particular team in this particular moment. I think there's room for Smart Pulses, and some of 'em may pick up on that. What we get Instill to do better at a Smart Pulse System. I think we may help all these apps get better and smarter at their work.
Steve Smith:
Well, I think what's interesting is that, you know, obviously we've, you know, gone through probably a 15-year period where, you know, surveys have just gone from, okay, we're gonna do a survey maybe once every six months or annually to employees, to where employees are getting pinged with multiple surveys a day.
It's just like complete, you know, survey fatigue. Then there's also this moment where employees are mistrustful, they are skeptical, they're feeling beleaguered because of AI and other things going on in the workplace. Is that a tailwind for Instill, I mean, or does that make it harder for your product to sell, because maybe leadership will be scared of what they're gonna find?
Dean Carter:
Yes, yes. And here's, I think AI had a really bad first date. You made a really bad impression on its first date. You know, you go there and everyone's like, it's here, I'm here to take your job. You know? I'm like, what? I thought you were gonna be helpful to me, and now you're taking my job.
And then also, it was more like a blind date. You're forced to go on. You're like, hey, they show up and AI's on your screen, and it's like, use it. And, matter of fact, we're gonna count how many times, you know, held its hand
Steve Smith:
Yep.
Dean Carter:
Date, and you're like, I don't even know this thing, and you're making me hold its hand and use it every day.
I just, it was a bad intro. There's so many better ways we could have done the change management and all the work into AI, then introducing, like, hey, this invasive species, we're gonna put it on your computer and see. And I have the challenge now, 'cause we are really light. Well, all we do all our AI work on the background.
We don't push Instill into your company to do AI and get like an invasive species in your company. All this is on the inside, but I get put into the AI invasive species category, so I have to go through the AI Governance Council, which is death. And so getting through the process, because we have the label AI.
It feels like we're going to destroy the company, because there's gonna have be something happen. All of ours, like, there's no more danger for us attending the meeting than me or you.
Steve Smith:
Right.
Dean Carter:
It's there, it gets the information, it works it, and puts on an email, no different than email you and I would send to anyone else, and it goes to URL, just like you would go to that, like a Google search.
So that's the, but it, because it's AI, is a challenge. But here's what I wish. The world, it started with Instill, and this is why Instill is there, to help your behaviors get better every day and save your job. This is a way of seeing how you behave with other people, that you can help them be better and you get better every day.
If we get introduced AI as this tool, this is designed to be in the room, to whisper safely how you can get better every day, versus tell me your skills and tasks so I can augment you. Like, I don't know about you, but I don't, I'm not, whatever that word augment means, I don't think I want it done to me.
Like, no, I don't wanna be augmented, whatever augment, I offer augmentation. So I just want you to help me be better at my job. And I think that we need more tools that are focused, like Instill, and they're gonna emerge, that are really helping people save their jobs and be amazing at the work, and so amazing that you won't replace them completely by AI.
And I think you can't AI trust, you can fake it, but you can't AI trust because it's so, so dynamic and changing every moment. Or psychological safety or innovation or awe. There's a really great podcast about awe, the power of awe. Like, wow. Like, wow, that's cool. And so I'm just, I've been studying a lot about the things that are uniquely human.
And how do you embed those into technology? Like the power of creativity and awe, like, ah, or the power of wisdom, like AI doesn't know all the experiences in my life. Doesn't know, there's no way it can have the judgment that I have, because of all the things happened in high school and college, and all the times that nothing is recorded about my life. It cannot know that ever, even though it knows a lot about me, but not that.
And I think the power of, like, wisdom and earned experience, the power of awe, the power of play, you know, being really playful at work, I think we can bring a lot of those back. We can be a lot more playful. We're gonna have more awe and creativity. We can honor experience a lot more. Yeah, you can AI a technical skill now, but you can't AI my life.
Steve Smith:
Well, and it seems like that's where the opportunity is. It's just like, if, you know, can AI, I mean, if you, we wanna look on the bright side of things, AI can make you more human, and it's, seems like the kind of feedback it's getting from Instill, helping a person understand their behavior, and with non-judgmental feedback, but feedback that it seems like it has the opportunity to, at an individual employee level, increase the humanity in those work experiences.
Dean Carter:
You have 500 people at a company, and 300 people in that company every single day, for some reason, use Instill. And 300 people a week later have behaviors that are trackable and improved better than the week before. And the next week they're a little bit better, and the next week they're a little bit better.
Can you imagine, like, a workplace where everyone is getting better at their behaviors every single day, like 60% of their own choice, not because they're, you know, had a performance improvement plan
Steve Smith:
Right.
Dean Carter:
Because they have, you know, helpful information about their behaviors, and they can make choices around what they want to do.
Steve Smith:
Well, and it's interesting, 'cause it's just like, there's also the scale issue, because it's just like, I mean, and you've seen this, it's just like we've dumped so much line managers, and it's just like, so great, you need to give direct feedback to each one of the 12 people on your team.
And then how do you measure if there's any progress against the feedback? It's just, it is at a human level. It's not possible, but what you're talking about makes it, not only the feedback, but then the measurement of progress, highly possible.
Dean Carter:
It does, it enables, then you can have conversations like this. Over the past month, what have you learned about your behaviors? Like, what are you working on? How can I be helpful to work on the things that you've been working on? Because I'm not in the room with you every day, but you are getting information about your behavior every day.
What have you learned versus me doing a 360 feedback, calling all your peers? All your peers said, you know, you're this, and they're like, no, they didn't. Who are you talking about? Becky's the one who said that, right? I'm like, no. Was Becky, but I'm not supposed to tell you, it's Becky. So that's all that 360 feedback and all those things.
This is really, and everyone, and Marcus Buckingham, will say 360 feedback is the worst feedback ever. It's the only worst feedback you get is self-reporting. So I'm really biased about my performance, and I'm even more biased about everyone else's performance. So having an interested party that isn't you, and that isn't, you know, a human being, that actually gives you unbiased information every day. Now that doesn't mean and have information go, hey, this is how you made me feel, and have conversations about that. This doesn't replace the human reaction, it actually enhances it. And maybe you get the opportunity to change your behavior before someone comes to you and says, you know, you'd never call me in meetings.
Steve Smith:
Right. Well, you know what's interesting is there's a lot of AI washing in tech right now, and, you know, what does, you know, AI actually look like in a culture product versus a traditional tool with AI slapped on the label, in your opinion?
Dean Carter:
We are AI native, so this is everything we do. Everything we do. So we are. Our entire system is AI, and everyone in the company is like AI native. We all are in it, every day. We are experimenting with open claw. We are, we're really trying to stay at the edge of what is possible and what is terrifying, every day, and really trying to find ways of, you know, using it to help people do their work better.
So that's our focus. We're all thinking about this every day. I don't know, there is a lot of AI washing, I think, if you're designing something to replace humans, I think that's gonna scare people.
Steve Smith:
Yeah.
Dean Carter:
And I could, like, you see the post on the phone, like, hey, we just made an HR business partner, you don't need an HR business partner again. Like a million dislikes, like.
Steve Smith:
Yeah.
Dean Carter:
No one's like crazy when, like, we made an AI business partner, because it's a ridiculous concept, that is, you know, that there are things about that HR business partner job that are so human that you need, now there's 60% of it maybe that isn't, and you can free them up to do a hundred percent of the time the work they've always wanted to do.
But that, I think there's a lot, you can't just tag AI on something and have it do a little bit of intelligent called AI program. We are from day one, everything we do, as a matter of fact. It's a little challenging today, 'cause Claude is down.
Steve Smith:
Yeah, there's that.
Dean Carter:
So we know when Claude is down, we are down. We just wrote an SLA around, you know, 99.9% uptime, 'cause we are never down as long as, and the, with the statement after, as long as Claude's up, we're up. Well, we did.
Steve Smith:
That's a pretty key asterisk there.
Dean Carter:
Today. So we are working on that, but what I think about AI, AI isn't AI. Everyone thinks it's the same, and it is not. I joke about it, and I think this concept of, like, AI readiness, and you need to go to AI school and learn how to use AI. It's the stupidest thing ever.
How many people went to phone school?
Steve Smith:
Yeah. Right.
Dean Carter:
Went to phone school? Like, we all, everyone knows how to use this thing, and it's pretty darn complex. Like, AI is like phone school, just start using it. And if I could, like, raise a magic wand with every person who wants people to use AI, give 'em Perplexity, Claude, and ChatGPT for free to use at home.
Pay for it. Use it at home all day by using it and playing with it at home. They will use it at work beautifully and know what it does when they know that Claude is wildly different than ChatGPT. ChatGPT is wildly different, Perplexity, and there was a fourth one, it will be wildly different. You don't get the same answer from Claude, ChatGPT, and Perplexity.
That's important to know. AI isn't AI. And you have to know what's behind the curtain of AI, because you can build a lot of junk using AI to build junky things. But if you don't put, what I think we did was like two years of intelligence behind it that has math and tested, and not just let AI build a bunch of slop. We actually helped it think better by giving it, this needs to be, it's like a color palette. You add a little more blue and a little more yellow. AI wouldn't have done that.
Steve Smith:
AI can spin up something that looks like insight in an afternoon. Whether it is actually insight is a different question, and it is the one Dean spent this whole conversation answering.
Once the technical skills get automated away, what's left is judgment and the ability to read a room. That is why a career HR leader ended up running an AI company instead of an HR department, and it is why Instill exists at all.
Dean's team calibrated Instill's scoring system with researchers at Berkeley and Stanford, building real, tested weighting behind the numbers instead of letting an AI model guess at what mattered. That is the difference between a real product and AI slop wearing a product's clothes.
The tools are everywhere right now. The judgment about how to use them well is still entirely human.
If you enjoyed this episode, make sure to subscribe to Work Tech Weekly on Apple Podcasts, Spotify, or YouTube. And I'll see you next time.
