Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Spotify | Android | TuneIn | RSS
Markus Edgar Hormess offers this advice: “Never prompt alone.”
Markus was working with AI long before the current wave of excitement. He experimented with early versions of ChatGPT and quickly identified new opportunities to collaborate with both his human colleagues and his new AI coworkers.
He’s currently building a community – Teaming with AI – to study and share these new practices and to explore the future of teamwork in the age of AI.
We talked about:
- his background in strategic prototyping and how he’s applying it in his Teaming with AI initiative
- his first exploration of AI, in 1986
- one his first applications of current AI tech, a use of ChatGPT-2 to accelerate service design prototyping activities
- his work and experimentation on ways to engage AI tools as collaborators on design teams
- how to consume research on AI, but also the importance of getting out in the field since research develops more slowly than professional craft
- his insight that you should “never prompt alone” so that you and your collaborators can eliminate bias and get better answers
- some of the opportunities that AI creates for real-time research and accelerated implementation of research insights
- how important it is “to put people in the center of this”
- the benefits for design practitioners of diving in and experimenting with AI tools, always with collaborators
Markus’s bio
Markus Edgar Hormeß is a well-known consultant, practitioner and educator in the field of service design and design thinking. In his daily work, Markus helps organizations tackle complex business problems and make team cultures more agile and human-centered. The focal point of his work is strategic prototyping, where he constantly pushes the boundaries of what a dedicated team can achieve with limited resources.
Markus is a strong believer that we should break down the perceived boundaries between technology, design and business – and that cheap experiments and prototypes are efficient tools to move your company, your strategy, your team, or your project forward. Based on this mindset, he has shaped multi-year programmes to help multinationals shift towards a more hands-on, pragmatic and effective approach to customer experience and innovation.
Markus has a passion for good design, human technology, practical experiments, authentic services, and playfulness in all things. He is co-Founder of WorkPlayExperience, a service innovation consultancy which helps organizations worldwide change how their staff, partners, and customers work together – and – how they can strategically discover and create new products and services. His practice builds on his experience of service design and business consulting, and on his background in theoretical physics.
In 2010, Markus co-initiated the world’s biggest service innovation event: the award-winning Global Service Jam. This was soon followed by the Global Sustainability Jam and the Global GovJam, and Markus has been a leading figure in establishing the culture of experimentation and prototyping which Jammers worldwide call “DoingNotTalking”.
Markus co-wrote “This is Service Design Doing” and “This is Service Design Methods”, top-selling books which have become the standard reference books for many practitioners and academics. He teaches service design, innovation, and sustainability at various universities globally, and is adjunct professor for service design thinking at IE Business School in Madrid.
In 2023 he co-initiated the Teaming with AI conference and community. His growing interest centers on how AI influences our approach to teamwork and collaboration, as well as the broader impacts on innovation and the development of strategies that are resilient in the face of future challenges..
Connect with Markus online
- Teaming with AI website
Video
Here’s the video version of our conversation:
Podcast intro transcript
This is the Content and AI podcast, episode number 17. As AI tools arrive in our workplaces, we’re discovering that this isn’t just another technology adoption cycle. The generative nature of tools like ChatGPT permits rapid iteration on ideas and quicker learning about their impact. For a prototyping strategist like Markus Edgar Hormess adding these AI agents to his service-design teams has been a boon, letting him and his colleagues collaborate and experiment in ways they couldn’t have imagined just a few years ago.
Interview transcript
Larry:
Hi, everyone. Welcome to episode number 17 of the Content + AI podcast. I am really happy today to welcome to the show Markus Edgar Hormess. I first met Markus a year ago at a service design workshop in Amsterdam, and we’ve been talking ever since about getting him on the show. So it’s great to finally have you here, Markus.
Larry:
Markus, he’s one of the co-authors of the book This is Service Design Doing. He’s real active in the service design community and in that world he’s really focused on strategic approach to prototyping, which is what we first wanted to talk about. And then AI came along. So we’re on the Content + AI podcast. So anyhow, welcome, Markus. Tell the folks a little bit more about what you’re up to these days.
Markus:
Hey, Larry. Thank you for having me. Yeah. So you mentioned it, so I’m super interested in strategic prototyping and prototyping in all kind of aspects. And when this whole wave of AI came about, we thought, “There is no books, there is no papers that tell you how to do this, so we need to prototype our way into this new world,” and that’s why we set up an initiative, which is called Teaming with AI, where we focus on the impact AI tools have on the way we collaborate in teams. So a small group of people that have a common goal, that trust each other, hopefully, and try to make something happen in the world. Might be nonprofit, for-profit, wherever you are.
Markus:
And so we set up a couple of events, a little Unconference early last year and one in the middle of the year. Then we started writing a white paper about this. This is about to be published soon, so hopefully we get some conversation about this. But all of this is really about giving a space, a play space for people that are interested to explore what is happening there. Only a few people actually focus on that team aspect. That’s why we have a strong focus on that, because you know it, service design, what we always say, it’s what is the key skill that you have to have in service design? That’s facilitation. That’s working with a group, whether you’re part of that group or if you’re facilitating a different group. And now one part of that group is AI and how does it change things? It changes it, and it doesn’t change it in other parts, but certainly a lot of shift going about.
Larry:
Yeah. There’s two things in there that are really interesting to me. One is that we’re all still humans and we’re going to be throughout this, whatever this AI thing turns out to be, but also the fact that, I feel like, you’re living in the future a little bit, because when I met you a year ago, you were already deep into this and really exploring it. And now you’re way into this collaborative paper and you’ve given it a lot of thought and you’re going to be providing these materials that you just said didn’t exist yet. So thank you for that. But tell me a little bit about when and how did you first get interested in AI? And how does it fit in specifically with your… Because you first came to my attention or you really stood out in that workshop as the prototyping guy. And so talk a little bit more about that. Yeah.
Markus:
Yeah, sure. I gave this a bit of thought, and then I remembered something, that back in 1986, I think I was in seventh or eighth year in school, I did a big presentation about the state of AI at the time. That was during one of these first waves, big promises in AI, “We’re going to fix this by the end of the decade,” and it never happened. But that was still when there was this kind of, “Oh, we can maybe do this.” So this was time of programming languages like LISP and stuff. I think that was where I got curious. Then I forgot about it for a long time. And then just after I finished university, I started to work at the Bavarian Research Center for Knowledge-Based Systems, which basically was a spinoff of the chair of AI at the LMU University. But that was, again, during a time where we were in niche use cases. The machines weren’t fast enough to do the big stuff that we can do today. But that’s where I learned that, yeah, niche use cases can be useful and they still are to this day.
Markus:
And then fast-forward, me getting into service design and innovation. And three or four years ago, no, three years ago now, when GPT-2 came out, it was accompanied by a wave of tools that would allow you to come up with better marketing texts. And that’s where we pick them up and use them in prototyping. Because in service design, if you design a new customer experience or service, how do you make this tangible, right? And one super simple way is to create a little advertisement for a new idea that doesn’t exist yet. It’s easy to test because people know the format. So it’s a really good way to test the waters if people like that or value that way you’re trying to sell.
Markus:
And using these tools, there’s this little, “Oh, give me 10 variations of a Facebook advertisement or a Google ad.” And then the teams would just use these tools within our workshops. We get these 10 and then curate the ones where it’s, “Oh, yeah, that fits what we thought.” And they could go faster, which is, I’m not obsessed by faster. There is a caveat there, but within the design process, being able to get something faster means you can iterate more, and that means you can learn more. So you can reflect on, “Oh, what does this do? Is that what I meant? How does it test? What are people saying about this?” And then you get faster into the next iteration, which is basically the key.
Larry:
Right. I know that iteration is so important in service design work and just design work in general. And so as you’re talking about this, I’m thinking, “These are…” You talk about these new AI gadgets as new collaborators on our teams and how to work with them. And it’s kind like you get a new person in, bright kid, seems capable of some stuff, and you’re just figuring out what they’re best at. Is that sort of how you fell into that? Or was it just the opportunity that GPT-2 was one of the first generative AI things and you’re like, “Great. We can get way more messaging testing done.” Yeah. Talk a little bit about-
Markus:
Yeah. I think the big jump was when GPT-3 came out and because it was a massive step from GPT-2. GPT-2, really, the use cases were certainly more creative. So I followed some authors that were using it as a tool to get them across writing blocks and get some inspiration from. But that was writing a novel or something. It wasn’t yet in the field of fixing problems or solving problems so much apart from the text generation part. But then GPT-3 came and suddenly we realized, “Hey, there is something happening there, so this is a different beast that we have to deal with.” So suddenly we can ask it for knowledge. Well, not so much, with a big asterisk on it, but we can ask it for, “Give me some questions. Give me a sketch of what I should do,” and then work through that sketch. And with some interesting prompting techniques, you could really do stuff that was super relevant for service design, for innovation, for any kind of project that was trying to solve a business problem.
Markus:
But we also realized that when you throw that into a team that because there was no manual, how do we get to the point that the teams now run as smooth as they did before when we didn’t have these tools? Now we have these tools, there is a whole kind of change happening until we’re on the other side of being maybe good enough that we can match the quality and smoothness of our design process that we had before. And so I work with Florian Bailey, which is a CEO of a consulting company, around marketing but also strategy, and he and his team would jump in straight away. We go all-in with AI. And then we started to run these experiments together, right? So what happens in the team?
Markus:
The observations were really interesting. The teams that were using AI… We were using little jam-like structures like, “Let’s take four hours with a goal of creating a strategy or creating a campaign.” We had teams set up that would not use the AI, and we have teams set up that would use the AI and then observe what is going on. And it was amazing how fast the ones embracing the AI would go. But over time, we also saw that there is a lot of gray area of, okay, initially we worked with people that were really keen on getting into this, but there were also other people that weren’t so keen or were struggling. And then we got really interested in why are they struggling?
Markus:
And so we started to try and say, “We can’t just do this for our own. We have to set up more spaces and open this up to more people.” So we created a little Unconference where we talked about this stuff, but also had a whole day of running experiments. It was trying to use that to do a stakeholder map or trying to use that in design thinking, service design, and also run a little prompt challenge at the end, like a prompt slam. It’s kind of like a poetry slam but using AI, a little competition structure. And we learned a lot. We learned a lot because-
Larry:
Yeah. It sounds like some of that learning started with that. I don’t know anything about Florian’s agency, but that agency is big enough that you could have multiple teams to observe, which were and weren’t using it. So I love that it started with that sort of observation and then how experimentally it gets going forward. Have you codified that inquiry methodology as you’ve gone along?
Markus:
No, not yet. Not yet. So basically, and this is the problem, yes, people are doing this right now, so we’re constantly screening the academic literature, but one of the things that you have to also be aware of, the academic research process is slower right now than a lot of the tools are evolving. So we now see an influx of papers that are still based on GPT-3, and it’s because the publishing just took so long.
Markus:
So whenever you read research papers, there are really good ones out there, some really, really good ones. But always double check what is their research design? What is the tools that they’re using? And maybe it’s already obsolete in some way, or at least you have to take it with a grain of salt. So usually academically it’s sound, you just have to put it in context, I think. But that’s the way it is. Every little piece of reflection helps. So it’s all for the better. But if you really want to be at the frontier right now, I think you have to be out there in the field, not so much in your class.
Larry:
And just experimenting with it. And I love that you’re doing that. One of the things you mentioned, I love the idea of the prompt slam. That’s a great thing. And when you were talking about prompting earlier and then thinking about it in the context of collaboration, we have these new AI collaborators that we’re working with, and prompts is sort of like how you communicate with them. Is that an accurate way to think about it is that that’s your training text?
Markus:
Yeah. Well, the interesting thing is that the way they work, I don’t want to get technical because it’s one of those things, it can get very technical very quickly, and you learn very little about what you can do with it. So it’s like if you can explain to someone how a motor works, does that make you a good driver? Possibly not, right? And what I’m rather obsessed about right now is the driving part. It’s how can we get better drivers, how we can drive together? But this piece of software, it’s a different kind of software. The interface is very human-like because it’s centered around language. And language is very much connected. As Mauricio Manhaes pointed out in a conversation with me the other day, it’s connected to cognition. This is how our brain works. So that’s why this is kind of scarily similar to a conversation with other folks and also with how we solve problems. And at least that’s the idea why this actually works so well in many, many kind of tasks.
Markus:
But also because it’s a different kind of software, people struggle with it. We have learned that software is, “Okay. I get a software, I get a manual, and this is what it can do. The developers know exactly what it can do, and I just have to ask them and then I figure it out, or I look in the manual or I ignore the manual as most people do.” But anyway, so there’s a certain learning about how user experience works and how software works. And this GPT stuff, the generative AI in general is different than that. So we have to unlearn, which is funny because the way we interact with these, with GPTs and these LLMs, is more natural than the other way. And the other way was we had to build it because computers couldn’t understand us. But now these tools can understand us, so we don’t have to go through the hoops of clicking buttons and pull down menus and whatnot, right? So-
Larry:
Yeah. I’m sorry, go ahead. Yeah.
Markus:
Yeah. But the thing is because people are not used to that and they can’t ask the developers, because developers train these systems, but they can’t tell you what it can do, so together, we still have to explore what it can do. And this is one of the challenges that we’re having. And yeah, we have to figure this out, not just for the general public, we have to figure this out for our own domain that we’re working in, in my case for service design, but then service design applied to whatever our clients do in their business.
Larry:
Yeah. And that seems really crucial because one of the things that just comes up over and over again is the importance of trust and trustworthiness and transparency and explainability in these things. But even the developers and engineers can’t explain what’s going on some of the time. So I guess just from a service design perspective, how are you helping both the teams that are working on this trust one another and trust the AI tools they’re working with but also develop trustworthy products?
Markus:
Oh, that’s a really good question. Not sure if I have a definite answer for that, but I think it starts with the people using the tools. There is a saying we came up with. It is, “Never prompt alone.” If you’re in a team and you’re working on a problem and you’re trying to solve that with AI or generative AI, then it really is about triangulating. And triangulation means that a couple of sets of eyes look at the results of the AI and judge it against their own domain knowledge. So you’re solving a problem. For example, if you run a restaurant and you’re trying to do something new and innovative, then you ask a couple of questions to the system, you get some outputs, and then maybe the team looks at that stuff and curates it and say, “Okay, no, this doesn’t work. I think that’s where it’s maybe hallucinating. But that’s interesting. That’s an interesting angle or perspective. Let’s dig deeper into that.”
Markus:
And by doing this together, what happens is that you get a shared understanding about the inputs from the AI. You get a shared understanding of the problem space, and you get a shared understanding of where the limits of the AI that you’re working with are. That’s what Ethan Mollick calls the “jagged frontier” of AI. There is tasks that you can give the AI and you get a good answer, a useful answer, and then there are things that you give to the AI that doesn’t deliver a good answer. Where this boundary is is very much different from any kind of domain that you’re in, and you have to figure this out. And I strongly believe that you only can figure this out together in a team because otherwise you bring in your own biases.
Larry:
I love that, “Never prompt alone,” advice. Because right now the concerns about hallucinations and then the introduction of things like RAG and other attempts, technical attempts, to address that, maybe it’s just a people thing. Maybe we just need to have two or three human sets of eyeballs on the problem.
Markus:
Yeah. Well, this is the thing, right? People say this software has bias. Yes, it has. Everything has a bias. We humans have a whole back load of biases that we bring to the table. But over the years in our professions, we learned how to deal with these biases. That’s why in qualitative research, you don’t do qualitative research alone, right? You have a couple of researchers that look at it from different perspectives. You train to get your own biases out of the way, by listening in a different way, right? You use different methodologies to collect your data.
Markus:
I think we have to use these principles of triangulation also when we work with AI, right? And as soon as we do that, if a person comes through the door, the new assistant, and you work with them, of course, because we’re humans, we start to figure out, “Okay, where does this person come from?” And then we roll this in in everything they say. That’s very human. Why do we expect this from the software again? Because previously old-school software was like a database with definite data. Now the software is a bit more like a human, and we have to use also different scales to do that. It’s not uncommon. We know how to deal with that. Yeah, let’s do this.
Larry:
When you talk about triangulation too, I’m thinking of, you mentioned research methodologies and collaboration with other people, but I’m wondering if there are… Because AI is much more than these gen AI tools like ChatGPT. There’s other stuff out there. Have you looked at that? That’s almost like an architectural structural level look at it. Have you looked at that?
Markus:
Yeah, absolutely. So one of the most exciting things I feel that is coming up is that we can now embed real time research to the way our teams operate and our companies operate. So we’re just working with a company who… And I think there’s many out there, but there’s a guy called Saul Robinson, and I met in St. Louis. He’s a great guy and he’s working on systems that allow you to take the user feedback that you get on your call centers and the company chats and stuff, and actually look at them through machine learning and AI techniques. And what it does, it automatically starts to cluster the problems that come in into meaningful buckets. And he calls that, I think, character cards or something like that. I would call them like little instant personas with certain needs and certain behaviors that are shown within these chats.
Markus:
And now this I found super interesting because what happens is this is what we manually do in research, right? We research, we talk, we do interviews that take weeks, and then we take this back and we have an insight workshop, and then we derive some personas, and then we work with these personas, and we know they’re temporary and they help us along the way and build empathy and stuff. So there’s still good reasons to do them. But in this new way of working, a lot of the data constantly cannot be updated and be reflected on. And then what happens is as soon as new problems and new behaviors pop up, the system tells me, “Hey, there is a new kind of behavior that doesn’t fit into the buckets that I’ve seen before.” And that’s when the data is taken to a new team and they have to look at this.
Markus:
So the researchers now get this data that just popped up and they look at it and they bring meaning to it. “So what does that mean? What is this new kind of behavior? Do we have to deal with it in a different way?” And then they can take action on this. That’s what I love about this stuff is that we used to wait for projects to be kicked off and get a budget, and then we’d do the interviews, and this is often long after the behavior changed. And here now we get enabled to actually get this in closer to real time, and we can still make a decision, do we have to deal with this right now or not? But at least we can again speed up our iteration, speed up the learning loops and say, “Okay, let’s look at this,” just screen the new behaviors, and if we feel it’s just a weak signal then we put it in a certain bucket and we deal with it later, but if it gets really important, we just jump right in. But now we’re way more in the driving seat and way less reactive.
Larry:
No. As you’re describing this, I’m just picturing that flywheel spinning faster and faster, that interplay between qualitative and quantitative research and the insights and then further discovery. Yeah. I think that’s kind of how we all feel about AI right now, this is this crazily spinning thing, but what you just said is a way to capture that insane pace and use it to your benefit.
Markus:
Yeah, absolutely. So this is actually if you implement that in your business, but there is the other side. There’s a human side of this. And the human side is that if you look at how experienced designers and innovators work and actually all kind of business people that solve complex problems, what you see in their cognitive patterns is that they are very able to go between different abstraction levels. They go from big picture to detail and back, and they do this very, very quickly. They go from, “Oh, this is a problem. What could be a solution? What are the new questions?” So they go back to or move forward into problem space again with new questions. And they can do this back and forth.
Markus:
And the studies tell us, this is, I think, Nigel Cross doing some interview studies, protocol studies on how junior designers and senior designers work, junior people stick with a certain position, so they stick with problem space, they stick with solution space, and they have a hard time going back and forth as fluidly as an experienced person. And that comes with time. But now, interestingly enough, we have these LLMs and we can use them as our own training tools to help us move between these different perspectives much faster. So in a way, these tools become our own kind of tutor and coach for these different cognitive… To train and strengthen these cognitive pathways that we need to solve the problems that are ahead of us. And they’re usually quite complex.
Larry:
Yeah, I love that you brought in that this always comes back to the human stuff and augmenting our human stuff. I’m thinking of Doug Engelbart and people like that, and finally, 50 years later, we’re actually seeing some of the stuff those kinds of folks envisioned. Hey, Markus, I can’t believe it, we’re coming up close to time already. These things always go way too fast, and I have 20 other things I’d like to talk about, but I need to wrap it up. But before we do, is there anything last, anything that you want to revisit from the conversation or just share with the folks before we go?
Markus:
That’s a good question. Let me think. But there’s one thing. I made a couple of asterisks along the way, and one of them I think is that we have to put people in the center of this. And when you’re trying to use AI in your own team, then maybe you are excited about this and you see the potential and you want to do things with this, but other people just see danger to their jobs. They see a threat to the, “Where does this leave me as a designer? Or where does this leave me as this expert on X? Now that the AI gives actually scarily good answers or good enough answers for a lot of people, do people need me anymore?” And you have to be aware of this. So I think that’s the first thing. Be aware of that.
Markus:
And then like a service designer, talk to people and understand where everybody is at. Where does that fear come from? And have they tried AI tools or have they just heard about AI tools? There is a lot of stereotypes out there that come from various places. Because it’s a software that we don’t know how it works and we haven’t seen before. It’s a different kind. “It’s software, Jim, but not as we know it.” So we have to give people that foundation, and that only comes by doing, right? And then they see, “Okay. Oh, it isn’t as powerful as I thought. Or it’s more powerful than I thought in this little bit, but I contribute this.” So it’s by using it, you get rid of all the preconceptions that might even come from Hollywood and movies, and this is all what’s shaping this picture in the press. We’re right in the middle of the hype cycle. So let’s get out of this hype cycle and just put a foundation of real experience underneath.
Markus:
And that’s where I would say, “Okay, try these tools.” Don’t just try them by typing some stuff up in the free versions, unfortunately. So try to get your hands on the actual best model that you can get for the use case and shell out these $20 or ask a friend, and then really play around with different prompting techniques. Not everybody needs to be a straight answer sometimes, just get some answers. Sorry, ask for questions. Ask for, “How shall I do this?” So there is now more and more material out there that you can use. And join some community.
Markus:
Don’t stick to problems like writing a bedtime story for your kids. That’s nice, but how meaningful is that for your job? Actually try to solve business problems. And be aware that you need a safe space to do that. Safe space meaning don’t ignore safety concerns. Don’t ignore privacy concerns. Talk to someone who has a good feel for that. Where can you do this? What is a problem where it’s safe to do that? And then try to solve that problem using AI. And then only then you can judge whether the output is useful for you. Are you this side of the jagged frontier? So is it a useful thing? Or is it not yet a useful thing?
Markus:
The problem is if it doesn’t work on the first trial, it doesn’t tell you a lot. We might have to figure out together how to make this work. So get into a community of people that are from your background. If you’re a content designer, find the other content designers that work with these things. If you’re a service designer, same thing. If you’re in a certain industry, find the industry guys that are doing stuff with this and see where they succeed and where they fail. And that’s where you learn. And that’s also where maybe some of the fear then goes away and it’s replaced by some trust in some parts and a very healthy distrust for the parts that don’t work yet. But now it’s rooted in reality and not just hearsay and stereotypes and hype.
Larry:
I love that advice, and I love that it’s always focused on people and relationships and empathy for the folks who are maybe less excited about it than you are, but also curiosity for the folks who are, and just the idea of seeking out community. That sounds like a perfect approach. Hey, one very last thing, Markus, if folks want to connect online or stay in touch, what’s the best way to reach you?
Markus:
Well, yeah, there’s a couple of ways, but the easiest way is to find me on LinkedIn and just ping me, and I’m happy to have a conversation on Teaming with AI or prototyping or service design, any of these topics. The other thing that you can look up is the Teaming with AI website where we will open up the community very soon and where you also will find the white paper as soon as it’s ready to go. This is not written by me. It’s written by a whole community. So, Florian and I kind of orchestrated that, but all this stuff we put on Google Docs and ask people to give feedback because we don’t believe there will be one white paper. We believe that it will be an ever-evolving sequence of white papers or other kind of tools that we have to put out. But this is about building that community and doing this together and having some fun, not just by talking about it but by doing this stuff and collecting our experiences and growing the experience together.
Larry:
Walking the talk as you always do. So, thanks. Well, thank you so much, Markus. I really enjoyed the conversation.
Markus:
Thank you, Larry. It’s really great to be here.
