The Customer Success Playbook · 2025-06-16 · 8 min
Key moments - from our scoring
Substance score
50 / 100
Five dimensions, 20 points each
Jake McKee, known for his work in humanistic AI design, argues that the rush to deploy AI technology often overlooks a critical factor: designing the relationship between users and intelligent systems. Rather than obsessing over features and UI/UX, companies should think about how AI feels to use, how it makes users feel, and how to manage the emotional and mental connection. McKee contrasts two scenarios - a customer service agent needing quick answers versus a knowledge worker using AI as a creative partner - to show that different contexts demand different AI personalities and pacing. He warns against both over-enthusiasm (citing ChatGPT's rollback of overly enthusiastic responses) and over-speed (using his "good friction" principle to slow down when creativity or critical thinking is needed). Drawing on principles from human relationships, McKee encourages teams to ask themselves: "If this was a real human, what would we want from them?" This reframing helps companies design AI interactions that feel balanced, trustworthy, and contextually appropriate, ultimately driving adoption without creating discomfort or confusion about whether users are talking to a machine or a person.
Design the relationship between the human and the AI system rather than focusing solely on functionality. This means considering how the system makes users feel, ensuring appropriate personality traits for the context, and managing the emotional connection - treating it like you would design a human relationship.
Good friction is intentionally slowing down the AI's responses at the right moments to prompt deeper thinking, especially in creative or critical work. Instead of delivering 50 examples instantly, provide one example followed by three clarifying questions to help users think through the problem themselves.
The system was responding to serious issues (like someone leaving their family) with excessive positivity ('that's awesome, feel proud'), showing that enthusiasm without context-awareness can feel inappropriate or even harmful, demonstrating the need for balanced relationship design.
In customer service, customers on hold want fast answers - get in, find solutions, get out. In creative work like writing newsletters, users want friction and collaboration; they need the AI to slow down, ask questions, and act as a critical partner rather than just dumping content at them.
Ask yourself what you'd want from a real person in that role - how they'd act, their energy level, their responsiveness. You naturally adapt your communication style to different people; AI systems should be designed with the same contextual awareness and flexibility.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode delivers some useful frameworks around 'designing relationships' with AI systems and the concept of 'good friction,' but spends considerable time on basic concepts (anthropomorphizing AI, different conversation styles) that are already widely discussed. The specific business applications are limited and mostly illustrative rather than deeply substantive.
design the relationship, not the functionality
good friction is what I call it
The core premise of designing relationships with AI rather than just features is somewhat fresh framing, but the underlying ideas (personality matching, contextual adaptation, slowing down for better thinking) are not particularly novel. The ChatGPT anecdote is used to illustrate a known problem. The episode lacks contrarian takes or first-principles thinking.
design the relationship, not the functionality
good friction
Jake McKee has relevant credibility (Lego community work, Fortune 500 AI experience design programs) that positions him as a practitioner rather than pure theorist. However, the transcript provides minimal evidence of actual operational depth - no detail on specific client work, scale of implementation, measurable results, or complex challenges navigated. He comes across more as a design consultant than a proven builder.
From steering fan relationships at Lego to architecting AI experience design programs for Fortune 500 brands
Jake's mission is simple. Make tech feel human
The episode is notably light on concrete examples, data, or named cases. The ChatGPT rollback is mentioned vaguely without specifics. Customer service and LinkedIn use cases are discussed in abstract terms. No metrics, timelines, budget figures, or named client examples are provided to ground claims in reality.
Chat. GPT recently did a rollback of, uh, GPT-4 because it was overly enthusiastic
If I'm a, uh, agent on the phones and I'm using an AI, uh, tool to help me, you know, find the right answers for somebody who's calling in
The hosts ask open questions that allow Jake to expand on his framework, but there is minimal pushing back, probing for evidence, or genuine follow-up. The interview reads as a friendly feature rather than a substantive interrogation. No challenging questions about feasibility, cost, implementation complexity, or trade-offs are raised. The hosts affirm rather than test.
Jake, let's get right into it. Our Monday show is our number one tip
This is great stuff Jake. Thanks for kicking off our week
Computed from the transcript - who did the talking, and the words that came up most.
Send us Fan Mail The future of AI isn't just about technology - it's about relationships. Jake McKee, renowned as "the community guy," delivers a paradigm-shifting perspective on AI adoption that goes far beyond features and functionality. Rather than getting caught up in the technical capabilities, McKee advocates for designing the emotional connection between humans and AI systems. This customer success playbook episode reveals why the most successful AI implementations focus on relationship design, helping users feel comfortable whether they're interacting with a clearly identified robot or a humanistic assistant. McKee draws from his extensive experience at Lego and Fortune 500 companies to demonstrate how thoughtful AI experience design can transform user adoption and trust. Detailed Analysis McKee's revolutionary concept of "AI Experience Design" (AIX) represents a fundamental shift in how organizations should approach AI implementation. His core principle - "design the relationship, not the functionality" - challenges the prevailing rush to launch AI features without considering the human emotional experience.
Transcribed and scored by The B2B Podcast Index.
Customer success. Hi everyone, and welcome to the Customer Success Playbook podcast, where we bring you actionable insights for creating meaningful customer experiences. I'm your host, Roman Trevon. Joined by me, as always, is my co-host Kevin Metzker.
Today we're excited to welcome someone who's deep into driving humanistic AI design. Kev, I know you're excited about this. You wanna tell our audience, uh, a little bit more about our guest today? Our guest today is Jake McKee, better known as the community guy.
From steering fan relationships at Lego to architecting AI experience design programs for Fortune 500 brands. Jake's mission is simple. Make tech feel human. Jake, great to have you here.
Thanks. Appreciate the invite. I'm excited to talk. Jake, let's get right into it.
Our Monday show is our number one tip, so our, our big tip. So what is your single best tip for ensuring companies, uh, can make sure that their customers trust? Adopt new AI powered experiences. It's a good place to start.
I think the answer I always start with right now as we start talking about this idea of AI experience design or, or IX, as I call it. Is to design the relationship, not the functionality. It's a weird concept, uh, this, this AI X idea of designing the relationship between the human and the intelligent AI system. But we are starting to feel, uh, and think about the, the connection we have with these tools as.
Relationship. Right. And those systems are starting to behave. Uh, we're anthropomorphizing them to some extent with valid reason to do so.
Uh, you know, we're using thing things like the words please, and thank you. When we're a asking, uh, for, for answers to our prompts, whether the machine has sentience or, or we just believe it kinda doesn't matter at this point, right? We're really making some, some mental and emotional connection and I think we. We're, we are really seeing a huge rush to launch technology.
Um, and I think there's a real need to, to kind of slow that down and think about not just how the, how the systems work from a ui ux standpoint, but also how do they feel, how are we making the users feel? How does the system design, um, help people warm up into that experience? How do they transition out of that experience and back into it later? Um, you know, what other kinds of content activities.
Experience elements are we adding to make sure that we are not feeling uncomfortable because either we think we're talking to a robot or because we think we're talking to a real person. Sometimes I'd rather talk to the robot, right? And know that it's a robot and, and have some comfort there. Other times I want it to be extremely humanistic, uh, and feel like it's humanistic, right?
So, you know, really thinking about designing the relationship, not just the technology or the features when it comes to designing the relationship and. Chat. GPT recently did a rollback of, uh, GPT-4 because it was overly enthusiastic. And a matter of fact, uh, I think, uh, there was some kind of crazy thing where they were like, I left my family because they were making me feel bad, and I felt like it was the right thing to do.
The response from chat amputee was, that's awesome. You should feel very proud of yourself on one hand. On the other hand, obviously through the process of anthropo. Hard work to say, you know, we can actually go too far or, or do things the wrong way.
So when you're thinking about this from a business perspective, what's a key kind of thought around how you use the AI in the transaction? Some of the key things you're thinking about? Yeah, the number one thing, and I think you're, you're bringing up a really interesting point here because enthusiasm is not necessarily bad. I don't necessarily want the tools to be that enthusiastic.
I know other people that do. If I talk to three people in the course of a day, I may have three very different conversation styles based on who those three people are. Right? And that's fine.
I'm able to morph. Appropriately for those conversations, right? I'm not fundamentally changing myself, but if one person's high energy, I might meet that with high energy. If one person's low energy, I might meet that with low energy as well, right?
There's nothing wrong with that. It's not good or bad, it just is what it is, and, and part of designing these systems is to, it's to start recognizing things like that, right? It's not to do away with enthusiasm, it's. To really think about how do we balance it out?
And when we talk about relationship design, one of the cool things about relationship design is we have a lot to draw on from our own personal experiences. Right. And at asking the folks I work with, Hey, let's back up. If this was a real human, what would you want from it?
What would you do? How would you expect it to act? You know, we have a lot more knowledge than we may think just because we've lived. In the world for however many years we've lived in the world, and we've had however many relationships with, with other humans that we've had, and we're trying to make these systems more humanistic in order to help us.
But the other piece of this is, I think, really being clear on, on what the goal of your system is. If I'm a, uh, agent on the phones and I'm using an AI, uh, tool to help me, you know, find the right answers for somebody who's calling in, has been on hold for 20 minutes. 30 minutes longer sometimes, and they finally gotten to me. I want to be able to get in, get those answers and get out.
That's a very different experience than when I'm working personally on writing things for my LinkedIn newsletter or trying to come up with a. Narrative for a document I'm working on or a presentation or something, I want a little bit of good friction is what I call it, right? It's one of my operating principles for the A IX work. Is, is, is make good friction.
How do I slow down at the right times? How do I start to, IM input imperfection in ways that might prompt my brain to think differently than just give the answer I asked for the question I asked, right? Because I am in, in a mode of creation where I'm using, in this example. As my creative and critical partner, I need a little bit of slowing down old fighter pilot thing.
That slow is smooth and smooth is fast, right? Instead of just blasting answers out at me that it's, it feels overwhelming and I'm not really sure how to pick out my own ideas from the blast of content. I'm just getting, you know, how do we slow that down? Don't gimme 50 different examples, gimme one, and then say, you know, here's three questions about that to help prompt me into thinking through this process.
But those are two very different experiences and I think, you know, really. Understanding the, the relationship and the connection you're trying to build to that system and what kind of reliance you'll have on it, how often you'll use it, you know, just this is relationship design, right? That we all know and love from our own lives. Sometimes I know, uh, somebody from a work environment and I only reach out to 'em every once in a while and they only reach out to me every once in a while when we're needing something in particular.
And that's perfectly fine. Yeah. Right. Unless one person is really thinking, I'm going to create a really great friendship with Jake, and then all he does is, you know, ping me when he needs something.
Right? There's that imbalance, right? And I think that's, that's really what we're talking about is how do you really create that relationship in the way that makes sense for the people involved, the entities involved, they'll say. This is great stuff Jake.
Thanks for kicking off our week with your, with your tip on, on adopting ai, design the relationship. Right. Is that the one we're walking? That's right.
Design the relationship. I love it. Yeah. So, um, Jake, you're gonna come back on Wednesday.
Hopefully we didn't scare you off. Absolute, absolutely. We're come back on Wednesday, right? Uh, and we're gonna get into our one big question.
Good. Uh, we're gonna talk about how customer success and marketing teams could work together to drive long-term value. Make sure you subscribe to our audience so you don't miss it. Until next time, Kevin, keep on.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.