AI Trust and Transparency for Business: What Leaders Need to Get Right Now

Good morning, LinkedIn friends. Value First Nation, welcome to another episode of Value First AI Daily, your collaborative intelligence report. Uh, it's Wednesday, January 28th, 2026, here with Nico and George. How we doing?

Doing good, doing good. Taking in all the trying to, trying to take in all the video over the last week from between Davos and Demis and Dario and oh my.

So how did they all start with D-sounding names, by the way? Like, what the heck is up with that? Um, keeping up with all the videos, to me, I, I hear that and it sounds like it's a full-time job, which I then worry for the mere mortal human being who has a job, a family, you know, as we do, but is trying to keep up. So hopefully they're listening to the show and we're helping maybe digest some of that stuff. Even in the green room, you guys were talking about stuff and I'm like, man, all I did was go to sleep. Like, what are they talking about? So it's— but I'm doing good. I'm doing great. Chris, how are you doing?

I'm doing good. That was like the best segue ever. Into today's commitment number 5 of the Value First AI Manifesto, because we're going to need this in order to keep up. We will design AI for transparency, not opacity. We understand that trust requires understanding. We will build systems where AI reasoning and conclusions are accessible unexplainable rather than hidden behind complexity. Um, and yeah, as you, as you mentioned, that struggle for the mere mortal human beings that, that we're trying to help. I know these three mere mortal human beings, uh, we've got mindset stuff we want to talk about, we've got news stuff that we want to talk about, And we wanna show like build stuff, like to help, like, and it is pretty much impossible to get those three things inside of 30 minutes. And so we've been switching gears here, like in January to try and say like, I think the news like will hit us kind of as needed and maybe there's a separate motion there. But if, if we can't help get your mindset right, it doesn't matter what tools are showing up at your door because there's some serious power in your browser, in HubSpot, in every which way you look right now. And if you're not looking for it, you can't use it. So luckily, like, if you're there and like it's been a little bit about like, oh, I use this thing, I wasn't sure how it was working, I wish it showed citations, and like all that stuff has gotten dramatically better because it needs to if we're gonna, you know, trust these systems. Yeah, yeah, yeah.

I mean, I think that the whole— okay, so where I want to start with this, because of the fact that it's trust, right? I mean, it's beginning 26. Um, I think the first place to start is by being, at least from a, uh, white-collar standpoint— I understand from a end user standpoint on the street, using it, hanging out in a bar, whatever. You're going to ask GPT something. It's not going to—

it may, may not.

It might hallucinate, right? I really— I think that when it comes to the white collar use of this stuff and you're using it in your org, especially if you're using Claude and you're using, you know, specified tools, basically, If you're getting hallucinations, that's amazing to me. It's utterly mind-blowing to me if you, if you get a hallucination today. Like, it— like, I, I can't— I really, I can't go beyond that statement because it's like, I, I haven't run into it in a minute right now. A lot of what I'm doing now is, is coding, so it's a little bit different because you're not necessarily asking for daily factual stuff. But the other half of what I do is research. So it's very important that all of that be as aligned and perfect and, you know, truthful and, you know, whatever. So I don't, I don't know that we— I think that we've reached the point where we at least have to start increasing the level of trust that we have in the output. And admitting that the outputs are getting better and that they are becoming more aligned to what we want and what we need. And then from there, I think there's also the other side of it where businesses have to be— you just have to be OK with it. It's getting to be a weird thing to me when I talk to people from organizations or I meet new clients in organizations And they talked to me about how, you know, oh, I don't know, because management not really too, you know, kind of iffy about the data and how data is getting, you know, handled and stuff. Again, like, to me, that says two things. One, you don't have somebody internally, and you, and you don't have like an AI anything put together. So like, that tells me that your organization is not ready to be using AI at all. The fact that there's any sort of pushback and that it's related to data privacy and security.

The—

there's a lot of reasons I bring that up, one of which, especially for, for us HubSpot users, you're doing it. If you're using any of the, any of the default things that are AI that are happening in HubSpot are using AI models.

Docker.

So like the data is going through a model, even if per se you didn't necessarily want it to or whatever, however you want to word that, right? Why is it okay? Well, for one, I have yet to see any of these guys get hacked into. Seems nearly impossible to try to get behind the wall of like an ever-thinking, ever-evolving, ever-reasoning system, right? Secondly, it's not just that, but also like who trains on what, right? So corporate versions of these platforms and even non-corporate Claude, non-corporate Anthropic doesn't have to train on your data. There's a huge difference in what I just said versus what we were saying last year. Last year we were saying needs your data to train on. This year we're saying doesn't need your data anymore. They just don't. The outputs from the models as they are now are better than the source material that they could get from people.

Hmm.

The synthetic data is actually of the same and or higher quality consistently than the regular, regular old human data, right? Like we need less, way less. So if you're thinking about like, oh, they're taking my data, it's not yours. It's going to be like high-end mathematics, high-end physics, like really high-end, like medical and stuff like that. Like super high PhD level stuff. It's not going to be your boardroom schematics and your flowcharts and your latest findings of whatever. That kind of stuff is just, it's too easy. It's too, you know, menial. So I think there, there definitely needs to be this cohabitation, this trust that happens where you're trusting the output of the model more, and likewise, you're able to express to everyone else like, hey, we do use AI in order to do these things. We use AI in order to do this inference where it, like, where it matters, right? Yeah.

This is where, uh, we're gonna let George go, but, uh, it's coming right back around to the humans involved, as it usually does.

I, I don't know what I want to say. I mean, I do know what I want to say, but I don't know what I want to say. Um, so first of all, I think that handling Human data, properties, information is very important. I think that we should be very careful in the choices we make around that as organizations. I think security is huge. And I think us humans historically are really bad about giving a shit about some of what just came out my mouth. If you look at what's happening on the internet right now around Claude, Claude Bot, now named Molt Bot because Anthropic basically like cease and desist and like because the name is too close to them and the amount of security threats of humans who have installed something because it was the new shiny thing that YouTubers are talking about. But not paying attention to the docs to make it secure on any way that you're trying to use it. Uh-oh, whoopsie. Um, like, it's a, it's a, it's an ish show. And I'm not trying to be a Debbie Downer because I immediately was like, oh, I want to see what this is all about. Um, but I gave it a couple days to then see what the flip side of all of this, like I'm not calling it hype, by the way, because I think there's, I think there's generally very interesting things that could happen with what has been built. But I think mere mortal humans are just going after what it can do, not how I should secure it to make it so I can do the thing. And now this is just an example, right? So yes, we're supposed to have trust And I, and I like the word trust. Um, I trust my wife. Um, I trust God. Um, those two things, I might even go to the level that I say I almost many times have blind trust. Like, I know that they're not gonna make a mistake that might put me in a pickle. Or they might. God won't. My wife might not. You know, you know what I'm saying? Because we're human. The point I think I'm trying to make here is, yes, a level of trust, but a level of trust because you have guardrails in place. Yes, a level of trust because you're testing and not just doing out of like blind— this is awesome, everything. Uh, what is that Lego song? Awesome, everything is Yeah, like, it can be till it's not. Like, and I think this is why transparency is highly important. Like, how many conversations can you go back to in the AI assistant that you're using because you actually asked it to do documentation that you could go back to? Probably none of you. And I don't mean the people on the show. I mean the people watching or listening to what we're talking about. Like, where's the Kanban board of all the edits that you've actually had your assistant try to make on the website that you're building? There probably isn't one. Well, then how are you going to go back to transparency? How are you going to actually lean into trust? How are you going to be able to go back to historical actions where I feel like a lot of people, this is just the next thing, the next thing, the next thing, the next thing. Which— that's, uh, dopamine, that's endorphins, uh, less trust. I don't know if— I, I don't know. Yeah, I don't know.

No, think of the second line of that song: everything is awesome, everything is cool when you're part of a team. Yeah, right? Like, and just, uh, this is another square peg, round hole situation. Where, uh, listen to my word change, um, as I reread this, uh, we will design humans for transparency, not opacity. We understand that trust requires understanding. We will build systems where human reasoning and conclusions are accessible and explainable rather than hidden behind complexity, politics, the boardroom door, like whatever you want to say, right? Like if you can't design, like in order to transform how your organization operates, like we haven't built organizations for transparency and that makes it pretty much impossible because like Nico mentioned, we got to trust the outputs. We haven't started trusting the inputs yet. Right. Like, and that's, guess what? AI needs some trustworthy inputs to be able to give you the trustworthy outputs.

And garbage in, garbage out. Yeah.

Right. And that's why, like, we're going to keep banging on this collab. Like, we don't even collaborate effectively. Inside of organizations. Talk about silos and misalignment and all these things. So, like, this whole system is meant to be like, here's the map to transform, but it's very hard at this point to be like, you get to pick this thing, and then maybe in 3 months you can do this other thing. And, like, the reality is you have to do it. You have to do most of it together. But when you can see AI as a partner in this, you can take it on and do it much faster than anyone would expect, especially historically how this has worked. Because you can, like, when we deliver this kind of stuff, like it's all about Giving the humans the time they need to talk to each other to manage change, to create trust between themselves, to say, okay, every time we've done this project in the past, I've felt in the dark. I don't ever know what's coming, and then you show me the new thing, and I'm not ready for it. I don't know how to use it. I got a job to do, and my life sucks. So yeah, I'm not excited about this next project that you're putting in front of me. Right. We have to find a way to have those conversations because leaders rightfully so are the seats getting hot.

Right.

And whether it's from the board or just constant news cycle about AI, AI, AI, they've got to find a way to start leveraging this. And, you know, that's why we're here to try and help figure that out. But it's not just plugging in the next tool.

Yeah. See, here's, here's the thing that worries me and about today's conversation is that you, you say they keep hearing, they keep hearing, they keep hearing the hot, you know, the seat is hot, it's getting hot in here. And here's the thing. When the seat gets hot, you start making decisions out of anxiety. And when you start making decisions out of anxiety, you try to move fast. And when you move fast and make decisions and you don't think about the humans, the security, the data, all the things, you end up making dumb decisions. And so like where I'm struggling with this is it's almost like this pattern that's being created on the internet to that, like, if you don't do this yesterday, you're stupid. By the way, some of our shows may have actually even kind of felt or said or been that way. And I'm just at this point, and it's maybe today, it's maybe for a week, I don't know what it's for, but I'm at this point right now where it's like, hold up for a minute. Like, what, what narrative is driving and who is driving and why is it being driven? And like, all of these things that we think are like frolicking in the— like, I'll stop there. I'll stop there. Like, I don't think we're in a flower field. I'm not saying AI is not important, and I'm not saying don't use AI. I'm just saying if you're the person watching or listening to this and you feel like you're in the hot seat and you're confused and it's creating anxiety, then you need to slow down, find a coach, read a book, do some research, like something.

Talk to AI?

So that you— well, yeah. Talk to your pastor, talk to your doctor, talk to your wife or husband, your partner, talk to somebody. Like, the point is that I'm trying to make here is like, stop and think and make sure that you're making this decision and many other decisions in your life from a rooted place, not a routed place, because life will route you. That's a different podcast.

I'll just— Yeah, no, I think like the challenge and the whole goal of this is like you do need to feel urgency, right? Like that's— it's— I think it's healthy for leaders right now to feel urgency, like especially in a world where, you know, The, the old ways have been broken for a minute, especially related to marketing. Everything's going down. You're wondering what to do next. You've been throwing money at it. It hasn't been working, right? So the urgency is real, and— but don't make any rash decisions. Like, this is a moment where, uh, there are a number of correct paths to take. But if you don't take a minute, and again, this is always the challenge of SaaS in my opinion, like if you just, the right pitch like hits you at the right moment and then you get the demo and you're like, all right, let's go, this is gonna change, this is gonna do the change for us, right? It's not, it will not, no matter the tool, like it's not going to because those humans in your organization need you to lead them to change. And, and the software, like, most companies are not equipped to come in. Like, this is where— this is why the, you know, the big consulting firms of the world, like, make the big bucks, because it's like, all right, here's the team of people basically coming in and, like, guiding you step by step through change management throughout your entire org, right? And they get like millions of dollars to do that, right? Uh, it's that level of change is required. But now, like, it— you don't— AI enables the scale in a way where you don't need like the 12-month project, the whole team, the millions of dollars spent. You just need to know how to bring what's available to you together. And we're working hard to figure that out where we're getting a lot of insights every day as we learn on our own learning curves, as we work with each other inside of this, you know, the content and what we're building in our various websites and apps and all these things. We're trying to figure it out for this purpose, right? Like, it's painful to watch. Like, look at all these beautiful humans that are trying to do good things. And, you know, whether it's signing up for the next tool or, you know, sharing the story about, yeah, I went over to Gemini and then it was just hallucinated, like, right away. And, you know, so maybe in a, in next year, we'll try it out.

Right.

That's so painful to hear. And what we'll tell you is when we can sit you down and like spend a half an hour with that guy, that flips like immediately.

Well, but I, but I think too, I want— yes, it flips. But I think that, Chris, it might be helpful for people to understand that, like, because we come on these shows and we talk about these things that they should do and, you know, we're trying to help people out, but they don't see the behind the scenes. Like, I don't know about you guys, but I've had some painful moments with AI where I'm like banging my head against the wall of like, why do you not get what 50 effing pixels is? I said 50 pixels down, not 500. What the frick? And why is this yellow bar that is the same way on every other hero section only 10 pixels tall when it's like 175? Like, I told you to look at this and you said you looked at this and this looks like a turd on the page and you're just going back and forth. Like, but, but the thing is, like, it's painful us to go through the learning curve so that we can bring the things that were talking about. Like, what I don't want people to think is like, oh, these three guys get on there and they talk about this cloud stuff. And sure, we might talk about some cloud stuff, but it's because we're in the dirt every day banging our head against this stuff trying to figure out how to make it work right and being able to share like these pain points. And so please, by all that's holy, do not look at this show and go, oh, they're the one-shot wonders. No, we're the guys who just stayed persistent. We're the guys who say, well, I'll try Gemini next year. We're the guys who say, well, I'll try Gemini in an hour, or I'll try Gemini in a day. Like, I—

yeah, I mean, I'll, I'll agree with that. Persistence is definitely, definitely key. Um, you know, I, I— it, it's kind of, it's funny, a lot of people have different takes on it. Spoke with somebody yesterday who is pretty, they're wading in, they're getting pretty good, but they're pretty much GPT-bound, but they use Manus. So there's this want to sort of explore. And there's something to be said about you know, what, what tool you're using in general, and like, do you need to actually go that deep into stuff like Claude Code or whatever, you know, stuff that we do. And, you know, showed them NotebookLM and it was just boom, right? And it's like, yeah, because this tool combines all of these things into this one place in a very cohesive way so that you can very easily understand how to work with it. Um, even though you can do this with other stuff and you can make it customized and you can put your own models behind it and all that, it's really nice to have that in one spot, in one package where you can access it, right? That, that whole Microsoft Office aspect of it, right? Um, so, but, but what they were saying that, that I thought was interesting was they felt that it takes a certain mind they felt that there was like a separation that happens that, uh, and I, I, I see that and I understand it, right? Because again, we have white-collar jobs, we have blue-collar jobs, you know, even when it comes to the different collars, there's different segments in there and sections in there. Um, so I do see that separation of like, you know, AI— I don't want to say, uh, like AI interested versus AI disinterested is kind of the way I see it. Because it's not something that's like— it's not difficult. It's more that like you have to, you have to be a little curious, I think. Um, otherwise, otherwise you're applying effort, right? Like it's something that if you were curious and you start playing with, you just— you're— everything just starts unfolding and unlocking, and you just keep going, and it gets really cool. I think otherwise you're having to apply a level of effort, but I think there's a point at which that kind of— the amount of effort that you apply starts to not matter because the output that you're getting is so grand, like grandiose, and has so much value to it by comparison to the input. So it starts to break down that barrier of this is too hard to even bother to get involved in because of XYZ. Um, so, you know, I— yeah, adoption, you know, let's change management, right? Like Yeah, how do we, how do we break down not just the, the barrier, but how do we, how do we identify which type of barrier it is?

Yeah, I'm gonna say something that I don't know if I ever thought I would say in my life. And I'm curious, and Chris, I know you want to talk about this, and we've quickly running or ran out of time, but I've never asked myself, when is enough interest enough? Like, you use the word interest and there's like a little bit of interest, a medium amount of interest, and then there's like the Taylor Swift fanatics interest, right? And so like this, and I'm just using that as an example, like Swifties are Swifties and they're always going to be Swifties and we're talking about AI, but like Do you have to go all the way to the right of the bar of interest, or for who you are and what you do, is just a small slider or a medium slider enough? I don't know. Again, I never really thought that I would hear myself ask, when is enough interest enough? But that's where my brain is on today's show.

Yeah. Now, I think I I think it's about bringing it to the interest, wherever, wherever that interest, either level or like topic-wise, the thing that you're interested in. Right. Um, and I brought this up because it's definitely like mental model of some sort, right. That has to shift here. Like, this is a paradigm shift. And I brought this up. I'm good. I'm not going to read from the screen. I'm going to read from. This feedback from, uh, like, speaking of transparency, we do Value First office hours, uh, every Tuesday, Wednesday, Thursday, um, and it's free. And if you sign up, even if you don't show up, you get access to these recordings of what we talk about, right? So like, this is what I'm talking about with transparency and just having conversations, not targeted out in the open and watch what happens. Somebody who I didn't even know was signed up, right, uh, says, listening to the, the VF office hours from this morning, uh, regarding Carolyn's mention of an AI-native resourcing model where SMEs are equipped with tools like Cloud Code, GitHub, and Vercel with output stored in company-owned repositories. I really wish we had that at blank. An on-demand GPT-like searchable library of all communication and knowledge and ongoing capture of all daily knowledge for blank that lived in a repository any of us could call upon. Am I crazy to think that that is even possible with today's technology? First of all, folks, like, I could not have, you know, like, manufactured a better product market fit response.

Yeah.

And that's just because we're talking about it, like, out in the open and making it available. And if we don't find a way to do that, right, people like they gotta find their own way to ask the questions in their own way. And then like matching up this, this, you know, problem versus solution is because it's such a paradigm shift, right? Like, and that was just so powerful to me to say like, yes, office, like this is what I mean when having, you gotta find a way to have these conversations out in the open. So that we can all learn together. But the short answer is, hell yes, it is possible with today's technology. And that's why we're here, because it's not about the technology. It's about the humans and your mindsets. We're over time. Thank you both so much. We will see everyone tomorrow. Thanks, guys.

Peace out.