Media

The Road to UNBOUND - Season 2, Sixth Stop: Five Layers and the Lived Experience AI Can't Have Without You

  • Rob JonesHost
  • Chris CarolanHost
  • Casey HawkinsHost

November 11, 2026

Rob Jones drove the sixth stop of the second season with Chris Carolan and Casey Hawkins. George B. Thomas and Kyle Jepson were out. It was Veterans Day, and Rob opened with a salute to veterans: "thank you to all who serve, who served … we are two men down and we are three strong and we are ready to discuss hitting limits." Casey had hit her Claude limit the day before, "but luckily I think it was twenty or thirty minutes before it reset, so I was able to be patient." Chris had the other side of that, usage left over: "oh, man, the usage can't let it go to waste. What can I do?" He spent a weekend moving "a lot of rocks" forward and came out "feeling good, feeling the flow."

Rob would rather there be no limit: "That way I can dive deeper, whatever, sink further into the matrix." The more familiar he gets with something, the more problems he finds, like a house that was perfect the day they moved in. "I find my experience with AI is very much the same. … Now, it's very much just, this sucks. I want everything to be perfect. I'm tired of iterating." One of his unpublished narratives fits: "The CRO narrative that I kind of called the crisis of continuity, I think applies to a lot of the stuff that we're doing now, right? Every day, every week, there's a new thing, a new trend, a new best practice, a new way to use this." Couch to 5K at least tells you to just do something, but if you switch to CrossFit, then yoga, then goat pilates, then powerlifting, "Can you get the same results from constantly changing everything or is there a little bit of continuity required?"

Chris took the couch: "You don't know what the new couch is going to feel like until the couch breaks or something spills on it and gets stained and whatever." With AI, "The speed is what everybody's uncomfortable with right now, for sure. They might not articulate it that way. But I think the only thing we have control over is how we react." Usually that means using AI "to add capability you didn't have before, to add visibility you didn't have before." And he tied it to Rob's analogy: "lived experience, finding a way to do that with AI is everything, right? You can't figure it out before you start talking to Breeze or to Claude or to whatever AI assistant."

Casey has two large dogs who get on the couches. "So whenever we get a new couch, it's like this whole stressful bit for a little while until inevitably the first stain, tear, whatever happens. And then the world didn't end." It has been the same for her with AI: "I have not been using [Claude] connected to HubSpot at all until Friday. And I think if you asked me why, I probably would have said I don't know, but I think really it was I was scared that it would mess something up. And then I started and it has not built everything perfectly. But the world has kept spinning." She is on a lot of meetings and coaching calls, so focus is hard to find for a complicated project. "I was able to use Claude to do a lot of … what I would have needed to focus on. And then I could have an hour here, thirty minutes there to just QA and test and turn the wheel a little."

Chris said you have to start talking to AI: "You cannot keep up with the speed of change if you do not use or leverage the capability of the thing that's forcing so much change so quickly. … Have you ever tried to do your job lately without internet? How about without electricity?" If you want to work with other people and grow, "why wouldn't you want to use electricity and then internet and now AI, right?"

Then Chris walked through the five-layer operating system he helps people implement when he coaches. People "never had to worry about building operating systems or designing operating systems," and this one "doesn't have to be hard. It's five folders that we start with." The first layer: "we're getting hammered with context constantly." For an organization, that is memory and identity: "data identity, sources of truth, durable memory, places like HubSpot, hold a lot of that." The second: "We decide what matters," and organizationally "an explicit customer value model helps with this, like what creates value for whom and why." The third asks, "Do we have enough judgment and capability to do something with it?" That is where AI comes in, and where Chris puts people: "I have yet to see other diagrams that include humans in the intelligence layer for business diagrams right now. It's always AI focused and the OG intelligence came from humans. So let's include them. And when you do, all of a sudden you understand how important human domain expertise is." The fourth is trust: governance, safety, quality and observability, which "has existed for humans for a while," so making observability a thing for agents is "honestly not that different." The fifth is how we show up, and he has "started playing with the word communication and interface." He called it "a very simplified version of what's happening inside these layers," and said it's why he pushes people toward the conversation: "once you start having the conversation with AI, it becomes much more natural than everybody expects. … If you can enable yourself to share your lived experience and then discuss it with AI, all that stuff starts hitting." Without the foundation, people end up feeling, "I feel like I'm doing a ton more, but I'm not sure it's valuable."

Rob noticed the human in the diagram: "The intelligence and the first time you've seen a human in any graph, right? It usually defaults to AI." He wasn't sure "intelligence" was the right word and set lived experience next to it. He brought up Mike Rosenberg: "he's a writer on LinkedIn. And he is very, very incredibly anti AI and specifically for writing in the arts." By Rosenberg's own reports, Rob said, he has done nothing different for twenty years and is now paid well to write for other people. "So to have lived experience, AKA intelligence, you need continuity and you need to be able to share that lived experience through a particular lens or mechanism." AI is missing a perspective "until you give it one to operate within," and "there isn't continuity, even within threads and within chats." As Rob tells it, Rosenberg's posts say "guess what I've done? Nothing. Literally the same thing as long as I can remember. But now it's valuable." So "you have to aggregate lived experience in order to have credibility and build trust with people, so that the intelligence means something."

Chris said "what you just described is how humans have always worked, right? We draw on judgment and learn capabilities. We know when to call on someone else's expertise." Human domain expertise "is the positive way of thinking about tribal knowledge and institutional knowledge that usually people think about as a risk." It comes back to care: "Like, do you care? And when you do … all of a sudden your true lived experience can come out in spades." One disconnect being exposed now is between leadership and the front lines, where values get documented in a boardroom and "On the front lines, nobody's living those values." Everybody has lived experience in their own environment, "And if you cannot transition that to AI, it cannot possibly operate consistently or successfully in that same environment." Train an agent on the strategy documents and the front line says "this thing is garbage." Train it on real transcripts and emails, and "the lived experience comes out through the unstructured data," and leadership asks, "is that how we do business?" "We understand like ten percent of organizational context dynamics, I think."

Casey asked, "what do you think the first step towards giving a context is?" Chris: "Talking to it." He told the group that at the Admin HUG the day before, "HubSpot's rolling out a pretty massive change to the base layer of HubSpot, which is CRM records and the way that they look." The admins are "not wrong in this reaction." Asked how likely they were to use Breeze to help, on a scale where "One was definitely using Breeze, five was not," "there was nothing below three." But if you haven't tried it lately, "you can't possibly know that Breeze has turned into coach strategy mode pretty hardcore." He asked the product team if that was intentional: "And they're like, nope, it's very intentional." As for which AI, when and how, "None of it really matters, though," without this: "The conversation has to happen."

Casey said "sometimes Chris is good at thinking kind of above where I am. … I often just feel like I'm in the dark, just trying to make my way. And it feels like Chris has a much clearer map and path." So today: "Like, okay, let's make it happen. We care a lot. Let's execute."

Rob tried to disagree anyway: "it's not even a disagreement. I agree that you should try it. I agree that you should give context." The day before, he had asked a recurring task "did you take any of the feedback from yesterday and apply it? And it was one of those, oh, I didn't, I'm sorry. Here's where I went wrong." That is "consistently my experience with LLM is like, I have to point out errors." With a colleague, "I wouldn't value that as much as somebody saying, oh, let me fix this before it gets worse." He has had those conversations "with both Claude and people and the response has been much better on the people side."

Casey had heard the same from two family members, casual AI users, whose AI "got something very obvious wrong in a way that it affected their trust of AI overall." Chris took it to the relationship: "So if I work with Rob and all he can do to help me be successful is tell me over and over, hoping that I just get it, I'm not going to enjoy working with Rob." When it should work and doesn't, "We take a step back. What do I need to do to make you successful? … that is exactly what the five layers are for." "This is all my lived experience for the last year, working with AI every day, seeing what works, what doesn't." You need to trust "that the correction you just made is unlikely to happen again, and you know exactly why because it documented it in a certain way in this space and you can go double check it, right? You don't have to rely on this hidden memory."

Rob held his ground: "The point is some of this doesn't work. Everybody's aware of it. And I need you to have the wherewithal and initiative to go and figure out what does." Chris asked whether he judges the outcome of those conversations, voicing Rob's side back to him: "Or you're just like, okay, that sounds right. But I can't see any of the memories that you're supposedly saving. … So that system sounds like it's not working, bro. Maybe we can build something else." Rob: "It's not the five layers. I haven't applied that. It feels very whack a mole to me." Then a story. His team needed a wow factor and someone to do content for a conference, and "nobody knew." A team member messaged an NBA team's content lead on Instagram and got a meeting. "Neither one of us would ever have been able to put coaching in place or to put that guidance in there." That's the initiative he wants: "I want it to be in the AI to proactively find things. I have that in people currently more than I do AI."

Chris closed on the goal: "to put all humans in a position to use their expertise constantly, and that's it, and not have to learn systems and things that are not valuable." "Humans have never been more accessible on social," he said, even as people aren't coming to the websites. "Humans wanna help other humans a lot of the time. So when you can reach out in the way that cares, in the way that says, there's value to be had here, I don't know what it is, but hey, I like what you're up to, right?" And as AI content goes through the roof, "the value of being recognized as a human being doing anything online is going very far right now." Next week Chris wants "to talk about this gap with Kyle and George."

Moments from this episode

Key takeaways

You can't figure it out before you start. Chris Carolan: "lived experience, finding a way to do that with AI is everything, right? You can't figure it out before you start talking to Breeze or to Claude or to whatever AI assistant."

The fear is bigger than the stain. Casey Hawkins: "I think really it was I was scared that it would mess something up. And then I started and it has not built everything perfectly. But the world has kept spinning."

Let AI hold the focus work. Casey Hawkins: "I could have an hour here, thirty minutes there to just QA and test and turn the wheel a little."

Start with five folders. Chris Carolan: "doesn't have to be hard. It's five folders that we start with."

Put humans back in the diagram. Chris Carolan: "the OG intelligence came from humans. So let's include them. And when you do, all of a sudden you understand how important human domain expertise is."

Lived experience needs continuity. Rob Jones: "So to have lived experience, AKA intelligence, you need continuity and you need to be able to share that lived experience through a particular lens or mechanism."

AI needs a perspective you give it. Rob Jones: "there is some perspective that is lacking from AI until you give it one to operate within."

If AI can't get the lived experience, it can't do the work. Chris Carolan: "if you cannot transition that to AI, it cannot possibly operate consistently or successfully in that same environment."

The conversation comes first. Chris Carolan: "once you start having the conversation with AI, it becomes much more natural than everybody expects."

Ask what it needs to succeed. Chris Carolan: "We take a step back. What do I need to do to make you successful?"

Initiative is the gap. Rob Jones: "I want it to be in the AI to proactively find things. I have that in people currently more than I do AI."

Being human online is worth more. Chris Carolan: "the value of being recognized as a human being doing anything online is going very far right now."

Transcript

This transcript was generated by AI and may contain errors. It is provided to help you skim and search the conversation.

Keep learning

You just watched the work. Now build it.

The next step is doing it yourself — live, with people on the same path.

Get the next conversation in your inbox

One signal-dense note when a new episode lands. No noise.

Join the list