A conversation with:
Ty Hagler

How Creativity and AI Will Save Healthcare

In the ever-evolving fields of healthcare and creativity, artificial intelligence (AI) has started to play a significant role. Utilizing AI in innovative, growth-minded discussions provides unexpected insights and expands possibilities for your team. This is precisely the topic tackled in our recent episode of the med+Design podcast, where we delved into fascinating discussions on creativity, design thinking, and the application of AI in navigating these realms.

Creativity and the MIDST Framework

Creativity is often likened to a muscle - the more we use it, the stronger it becomes. In the realm of healthcare, fostering this creativity can lead to revolutionary, patient-focused solutions. To help ignite this creative muscle, Ty has devised a framework dubbed "MIDST," which stands for Momentum, Insight, Daydream, Shift, and Try.

He created this to help people harness their innate creative energy, particularly in the healthcare space. The MIDST framework encourages individuals to generate new ideas (Momentum), see known elements in a novel way (Insight), and allow for intentional procrastination or mental rest (Daydream). Simultaneously, the approach invites individuals to consider different perspectives (Shift), and finally, it inspires them to take action and experiment (Try).

The Dance Between AI and Creativity

Integrating AI into the creative process can seem intimidating, but when leveraged effectively, AI can be a powerful tool for ideation. Tools like OpenAI's GPT-4 allow for quick answers and can navigate the notorious "blank canvas problem". It is like having a co-participant in the creative process that does not tire or run out of ideas. It serves as an idea-generating partner that can help jump-start your creative engine.

However, while boasting its advantages, we liken AI tools to 'a sociopath and a liar.' AI tools lack a moral compass and tend to simply mirror what they believe the user wants. Understanding these characteristics is crucial to make the most of the AI's capabilities without compromising the creative journey's authenticity and originality.

AI As An Aid, Not a Replacement 

It's a common fear that AI might eventually replace human roles in various fields. However, it's crucial to note that AI is a tool designed for assistance, not replacement. It is there to simplify our tasks, generate a broader range of ideas, and help us in achieving our goals more efficiently. As AI continues to evolve, it requires us to continuously explore its boundaries and potentials.

Consider AI as a new, transformational tool akin to the introduction of 3D CAD software in design. Experts were afraid that it would make several roles obsolete due to increased efficiency. However, the outcome was an expanded capacity to handle more complex tasks within the same timeframe.

As we move further into a world inextricably linked with AI, it's never been more important to understand its strengths, shortcomings, and potential pitfalls. Whether it's in healthcare, creativity, or any other field, mastering AI could be the key to unlocking more robust and sophisticated solutions.

Implementing AI in Healthcare 

A significant consideration in the healthcare field is the infusion of AI into daily operations. A potential application of AI is morphine dispensing. A more human-controlled approach brings unique risks; on the other hand, going for full automation presents different vulnerabilities. Implementing AI could enable a more reliable, safe, and trustworthy outcome considering both ends of the spectrum, providing another layer of monitoring and control.

Another potential application is AI's integration into assessing patients' medical history. AI can help clinicians save time by quickly summarizing large and complex patient data, helping to devise more patient-centric strategies and treatments.

The Bottom Line

From aiding creativity to streamlining work in healthcare, AI stands as an essential tool in our modern world. It's not a replacement for human interaction or innovation, but an enabler which, when used effectively, can broaden horizons, ignite creativity, and improve efficiencies. It's an exciting time for those willing to embrace AI and use it to forge fresh conversations in the healthcare and creativity spheres.

Episode Transcript

[00:00:00] Jared: Hello everyone. And welcome to our first episode of the med+Design podcast in 2024. It's really exciting. Huge thanks to everybody that's been on this journey with us. And, we're really so grateful to have the opportunity to talk to so many inspiring people and tell their stories as change makers in healthcare.

And today we have a little bit of a change of pace. It's just going to be me and Ty treading through it a we have a couple of different major topics that we're delving into first and foremost is the midst framework. That's actually a framework created by Ty to help folks with their creative muscle and especially in the healthcare space.

And then also we're going to be digging into design thinking and AI. And so I think first and foremost time to talk about MIDST. So midst is momentum, insight, daydream, shift, and try. And so how did you even, how did this come up for you? So you're on a walk. I'm assuming because momentum is the first one you're on a walk and you're like, I should come up with a framework for how I come up with my new ideas.

[00:00:57] Ty: Thanks Jared and a little bit of background for this. We, like most of our podcast has been to highlight the stories of our guests who come on. And so I've been building up a backlog of content because I'll give presentations and other contexts. And so this is a nice chance to just talk about some of these I think the areas that we're doing some of, what we're trying to push the boundaries of thought leadership within kind of creativity, innovation, design thinking, and so that's a lot of where my day to day work falls into.

I think ever since probably the first ideation session that I ran at Home Depot back in 2004 or so, been trying to teach people who didn't have a creative background how to be creative. And the typical context for that is a team based ideation session where you get in a room together, you get some sticky notes, and then you say, okay, go, like we're looking at problems for I don't know, design a better humidifier or, whatever the product might be, whatever the challenge might be.

And it, there's definitely an art and a craft to structuring creativity for individuals. And one of the things that, as we were doing some work in building up the team context, teaching non creatives, how to do creative work, one of the things that I've noticed over the years of running these team based exercises is that you can tell a pretty profound difference between people who are trained in how to be creative, how to access that side of their personality, and those who aren't trained. We can do all kinds of facilitation for basically behavior management of how to get the, people who don't know what they're doing in a creative context to get up to speed quickly.

But there's one of the measures that I have for this is ideas generated per person in 10 minutes. I think I've just got that as a standard performance metric and you can just tell one, it biases towards younger people that people in their twenties tend to be much more agile, much more creative in how they're coming up with ideas in part, cause they don't have the scars of experience to say that's not going to work.

And so just their divergent muscle is, and it's also closer to that moment when they were a creative genius, when they were five years old, where they can access that memory of what it was like to be free of any of those kinds of restraints. So that's at least some of the impetus and the inspiration.

The other thing was, and this was talking to Dr. Jane Little so she was on the podcast last year and we had done a lot of work with her and her team to look at changing the care transition for pediatric to adult care for patients with sickle cell. And she made this observation as we were talking that the designer she works with at Trig just seems happier than the baseline, I don't know, mental wellness of people that she sees on a regular basis.

And it was just this huh is there something about that because I don't particularly feel more happy than, anybody else. I'm not sitting here dancing through the fields with kind of flowers and rainbows and whatnot. But it's but I think there's something about the creative exercises you go through for when you are going through a tough time that give you some tools and resources to better process that. All right. So that's all preamble to your core question, Jared, which is like the midst framework and where that came from. So part of it, it's trying to articulate, what is it about designers? That, and we're, we, and really people, anybody that's in a creative profession, what is it that we're doing and accessing that is leading to, one, better personal wellness being, and that also is there some consistent tools for that, that kick through?

I'll share my screen also, just if that makes sense, just to Let's check it out. Yeah. I'm going to share my screen, and for this framework of M. I. D. S. T. Which is an acronym that stands for Movement, Insight, Daydreaming, Shift, and Try. So there's five parts to this.

The first one with movement, and this comes from just an understanding in where creativity comes from, and when you're doing a creative journey, you're accessing two of your four dopamine pathways. And so one of those dopamine pathways that is where generative ideas, spontaneous ideas come from, comes from when you're moving.

So if you go out for a walk, if you go for a run, if you are in the shower, if you're brushing your teeth then you're more likely to come up with divergent new ideas because you're accessing a specific part of your psychology. The divergent dopamine pathway that has to do with movement and so regular routines for exercise and I don't really actually care what it is as long as you're just moving your body.

Will be where you spark new ideas and just have a routine for opening yourself up to new creativity. I went for a run this morning as part of my regular routine, even though it's 20 degrees outside and came back with a fresh idea for how to tackle the day, right? And so that routine is just something that want to stress as far as your own personal routines.

Alright, so the second one, insight. Really comes into like a profound understanding of a truth that's unobvious, right? So you're seeing, hearing, feeling, experiencing known elements in a new way. And one of the stories here, I'm I get teased for using too many Star Wars references.

But I'll say this, that George Lucas. when he came up with Star Wars, had a profound insight that, for his space opera, right? That there were, in his day, 1979 or so, or 78, whenever the first Star Wars came out, with the box office success of Star Wars, There were a whole lot of copycats that then emerged of, like, all these other bad knockoffs of Star Wars that emerged at the same time.

But each of them didn't understand the fundamental insight that George Lucas had, which was to pull from Joseph Campbell's book, Hero with a Thousand Faces. And understand the archetypes of different character personas that humans, as long as we've been sitting around campfires and telling stories, we have the hero who gets a call to action the hero's journey.

Right? And so there is a wise sage who comes in and interacts with the hero to give them the mission to then go off. And then there's the ruler that, of course, The heroes having to confront and so each of these 12 different archetypes show up in Joseph Campbell's literature. And actually, George Lucas had Joseph Campbell come in and consult on the movie construction because he stayed so true to that archetypal story. He had a fundamental insight to translate that. Those same themes that show up in, Homer's Odyssey, but do it in a way that is a space opera with space wizards and, lightning swords and whatnot, and tell a new story, but using old themes that everybody can relate to.

So that fundamental insight is there is one of those core things that if you can latch onto, and it takes reading, it takes experiencing in order to come up with that new insight that can then provoke and inspire a better vision for where you want to go. So that's insight.

With daydreaming, this is one of my favorite ones, which is giving yourself permission to procrastinate. That if you are procrastinating, it's actually a sign, one, that you don't have everything quite figured out yet, and that you should give your subconscious time to work on the problem, and it's also giving yourself deliberate rest, is that if you're tired, you're not going to be creative.

You've got a limited resource of the creative energy, the creative work that you can do. And so by allowing yourself that freedom and that daydreaming time that you're opening yourself up to being free to allow new ideas to come in for, allow your subconscious to work on problems in unique ways. Adam Grant, in his book Originals, talks about a study that was done where

there were two sets of students who were given a task. One was asked to immediately perform the task and the other was charged with the task, but then given permission to go play video games. And then come back to the task and the group that had been given a separate task that was fun and play and procrastination performed the task better than those who were just asked to immediately attend to the task.

So there is usefulness to procrastination. It's I want to remove that sense of guilt from it. And then also from a procrastination standpoint. Dan Sullivan strategic coach also talks about that as if you are procrastinating on a task and then it's a, it may be a signal that you should not be the one to perform the task.

Rather, you should be looking to delegate to somebody else who has more of a sense of mission about whatever it is you're trying to accomplish with a task. That's just something as you're thinking about daydreaming and procrastination to give yourself permission to do that, because you're going to get around a seemingly intractable problem.

just by allowing for that purposeful daydreaming to occur. Okay, so that's M I D, and we've got two more, S and T. So with shift, this is where deliberately shifting your mindset to see the world through others eyes is so critical to really being able to understand the world from outside of your own experience.

And so within our courses, we teach journey mapping of going through and, let's say the journey map, you're trying to solve a problem and the problem directly impacts you, the stakeholder, because you're the subject matter expert. And so you can talk to from start to finish what it feels like from your point of view.

But then there's other people here interacting with that journey. And so then you need to shift your mindset and look at it through other people's eyes. And the case study we teach. Is the, to the experience of an annual exam for a veterinary office. And so assuming we have the vet coming in, then the vet knows their experience for an annual exam with a dog.

But then we say, okay now I want you to experience this as though you're an older dog. What is the older dog's experience of going through this? And so now you have to be like, on all fours looking up and there's weird smells in here and all of that. So you have a funny different point of view and shifting your point of view from that of what you know to somebody who you interact with daily, but trying to see the world through their eyes.

And that opens up so many opportunities for improvement and change just through better empathy for the people around you. The last one in our midst framework is to try and you've noticed I've used my hands a lot. So I do think that there is a sense, and this comes from it, like a personal wellbeing standpoint, that fulfillment comes from creating with your own hands, getting your hands dirty and so finding a practice of working with your hands, whether it's woodworking, whether it's I don't know, like legos, if it's yard work, if it's like improv acting where you're creating something new and physically interacting with the new thing you're creating, there's a positive feedback loop. And for people who thrive on a sense of accomplishment, being able to point to this IKEA piece of furniture that I created.

Yeah, I followed the instructions, but hey, look, I built it right. There is a I don't know. There's a sense of accomplishment that comes from working with your hands that I think positively impacts your mental well being. And particularly if you're feeling stressed and overwhelmed, tackling a hands on project is just can be just so meaningful.

All right. So I've been talking for a long time.

[00:13:13] Jared: just to riff off of that too, like for myself, I think it's really interesting that like when I'm lifting weights and I'm at the gym, cause that obviously counts as exercise. I am totally not creative. That's not where I'm finding my best ideas, but then I'm not like you.

I don't go on runs actually like to go on walks instead. And so on my walks, I don't know what it is, but my walks are notorious. I have a phone full of notes in my notes app of different business ideas and for, also myself as a creative person as well who just likes to work on music production for fun as well.

I have a bunch of different song ideas that I come up with and different ways on how to present these ideas to the world. And my phone is just this huge brain dump of stuff that happens while I'm on these walks. And I think for people that are listening in, I think you really have to find what that thing is that gets your mind going, because I think it's going to be different for everybody, and also going into the daydreaming side of things. I think that's another side where when you talk about procrastination in a way as well as you're going from from I insight. You get your ideas, while you're doing your work and then your insight, you have this idea that maybe you've started to maybe work through you're trying to implement it in some way.

And for me, I think about that in like the sense of creative output, like we're trying to create a song from the beginning part and you sit with it and you're sitting there for hours at a time and you're working on the same parts over and over again. And all of a sudden now your brain is fried.

You've been sitting here and listening to this thing over and over again. And you're like, you know what? I'm done. Like my brain's not working anymore. And sometimes that procrastination of taking yourself away from that thing that you've been working on, that's maybe burned you out in that moment.

You can come back to it. You have fresh eyes, fresh ears, and it just helps you. I think Ty you made a great point is it's really hard to be creative when you're tired. So if you're really tired and you've been working on something for a while that's a great time to take a break, coffee break nap break, meditation break, whatever it is that helps you.

But yeah, I think also the last piece of this as well that I just want to touch on is actually trying it, go out and do this thing, like whatever it is. You've come up with some sort of insight on whatever that may be a lot of people have this sort of initial fear of not going out and actually trying this thing.

Everyone, gets stuck in that sort of I don't want someone to judge me or think I'm, this or that, and I think you just have to have the courage to try it because thinking about all these wonderful innovations that have happened in the world. You can imagine those people had fear and they had self doubt and all that stuff as well.

And I think the try part of things is while there's, all these other great pieces to it, just go out and do it, go do the darn thing. And it's okay to fail and fall on your face because you'll learn something. So anyways, I think that was just my little piece on just the way this framework has been built out.

[00:16:08] Ty: Yeah. And Jared, I also, one of the things I so appreciate about our conversations and collaboration is that my background is as a industrial product designer. Which my education, my professional life is doing the hands on things that are physical products or digital experiences where it's a visual tactile like output, and then, which I would describe as a creative.

Profession, Jared, your background from music is I see parallels, but if we can abstract out some principles of creativity, you're accessing a completely different medium with music and the creation of music that I think you can come at it from two different points of view for the output and still arrive at something we would see.

We would say both are creative professions, would you say?

[00:16:59] Jared: Yeah. Yeah. And even it also taps into a lot of the processes that go into design thinking as well as something that, you know, from working with you and learning from you about, how this all kind of happens. I thought, Oh my gosh, this is a lot of the process that people that are in the studio work on as well.

And it's just so fascinating to see the way that it's not necessarily the medium that matters. It's almost this just trusting the process of creation and I guess there's these themes that come over, over and over again, no matter what field you're in. It's interesting. We speak the same language in some degrees, it's like Spanish to Portuguese or something like that, or, it's wonderful when it comes together in certain ways. And so anyway, the MIDST framework is one of those as well. And I've ever since you've brought it up, I thought to myself, this is the way that I've been doing things forever and now there's a way to actually describe what that is that process,

[00:17:50] Ty: it's not that we're creating this out of nothing. It's more trying to describe behaviors that's common among creative people. So that because I think every human being. is a creative individual.

It's a matter of like everybody has capacity to go into the gym and move some level of weight around. And it's a question of how much training have you put into this. And then the specializations from different disciplines can lead to looking at creativity from different lens. Rather, it's a framework that's describing what's happening for creative people so that we can teach people who don't think of themselves as creative how to find better personal well being and start to practice what it is that we're intrinsically doing and trying to teach it.

One more thing I wanted to add is you were talking about managing your creative budget. I think of it as your amount of energy. It's like you have a certain number of forks that you start out each day. You think about the little forks with their tiny. And you only have so many different forks that you can spend throughout your day.

And when you're done, you just have no more forks left to give. And then you just got to put it down and move on.

[00:19:00] Jared: I'll throw on a nice little TV show or yeah, that's right.

[00:19:04] Ty: That's right. That's right.

[00:19:06] Jared: Yeah, then and maybe dig in a little bit more into sort of the MIDST framework and creativity as well, and how you see it best applied for folks that maybe are listening in and they haven't given themselves that credence to say, yeah, I am a creative person.

And they're like I said, we talked about it a bit, a fear of being judged, fear of failure. And maybe your own experience with having to overcome those things as a creative as well.

[00:19:32] Ty: Yeah, I guess that gets back to the George Land Study. Maybe we can talk about that.

Yeah. Which, you talk about fear of being judged. In the 1970s 1960s actually NASA approached a gentleman by the name of George Land to prepare a test. And so NASA was looking for extraordinarily creative scientists and engineers to hire, because that was going to help them on their mission to put a man on the moon.

And so they needed to be able to hire for these people. So George Land produced a test. And then was successfully able to identify creative geniuses in the engineering and scientific community in order to for NASA to make a more intelligent hiring decision. And as it turns out, only about 2 percent of the population.

tests as a creative genius. And George land went on and NASA, was happy because they were able to successfully recruit for that. But then George land was curious. And then he then wanted to look more broadly and say where else in the population other than scientists and engineers, might you find these creative geniuses?

And he opened it up to five year olds. And so that population tests at 98 percent as a creative genius. That's insane. Yeah. When you were a five year old, you were a creative genius. Everybody on this call, statistically speaking, was a creative genius at one point in your life. And so then what's fascinating is that by the time everybody turns 10 years old, only 30 percent of children retain their creative genius.

And then by the time you turn 18, then only 2 percent of the population is a creative genius. And you think, okay is it the school system that's doing this? And honestly no organization can be that effective. Even if you're trying to kill creativity. You could not actively squash it at that level of effectiveness to go from 98 percent to 2%.

There's something in human behavioral development that's going on there. And I liken that to say the normative pressures that happened to us in middle school of anybody who looks weird, anybody who is coming in with something that's outside of what everybody says is we're all doing this together.

There's profound pressure to fit in, to find your community, to click with people who you like, you're trying to figure out your identity. And along the way, that creative genius then gets suppressed in order to try to figure out how to conform to what your peers are expecting. And and I think, that then means that some of the people who are creative geniuses then wind up being the people who are on the margins of kind of the cool kids.

And those are the people who they act weird, they think differently, they're original in their context. And so those are the people who are able to retain that creative genius. And so I just think it's an interesting insight to the origin story for creativity. But then also when you ask somebody to then be creative who has gone through that hazing process of being, going from your, adolescent teenage years and just the, everybody has like just winces when they think about middle school.

So you just have that middle school ah, like that pops up for most people when you ask them to be creative, because you're immediately remembering the times in which you were made fun of in formative years as you were becoming an adult. And so in many ways we have to unlearn that or have the courage to stand out because honestly I think in order to do something that is novel, that's unique, that's different, it starts with commitment. That you have to make a decision you're going to do something differently. And by doing that, then you, then it's going to call upon courage because As soon as you're doing something differently, normative pressure comes in, incumbents then start pushing back against you.

And so you then also have to do new skill building in order to then, deliver on what you say, I don't think I can pull this off, but we're going to go for it. And only then, once you start seeing evidence of that, then you get to earn the right to be confident. But confidence doesn't start at the beginning.

In fact, most designers really are constantly feeling a sense of imposter syndrome because we're always putting ourselves right on the line of like where our comfort zone is. Cause that's really where that boundary is learning and we're constantly pushing that boundary forward as we're building up our practice and our craft in the creative field.

So anyway, I think that. It gets at what you were asking, Jared, right?

[00:24:25] Jared: Honestly, you brought up so many memories of my younger years, too. Yeah, I grew up being a piano player, and going into high school, and going into eighth grade into high school, like everyone was in the band, everyone was playing the cooler instruments.

Everyone was a guitar player, drummer, whatever. And but piano wasn't cool, like playing Beethoven and Mozart was definitely not cool, even though I was really good at it. And so in order to conform at that time, I was like, I'm going to be a bass player. No one's a bass player right now.

So I'm going to do that. And I spent my high school years. While still also being a fantastic piano player and just totally suppressing it because I just thought it was the not as cool instrument. I just lived in my high school years being this mediocre bass player and just totally being afraid of that judgment because I was so young and fearful.

And then it didn't end up being way later on that I went to music school as an adult. and ended up regaining my love for the keyboard for the piano again because music production the keyboard is such a big piece of what you're doing And yeah, I think it's this struggle of fear of failure and fear of judgment.

It's a lifelong battle folks, it's not something that just happens and you just get over through using just this framework alone. You have to have a lot of courage to try to be this creative person, give yourself that allow that creative genius to flow.

And the only tip that I will give that helped me a lot is meditation. I've noticed that when I'm in a more I guess you could say constant meditative practice, that's the time that I honestly just don't really care about what other people think. And I don't really like all of a sudden, like you actually have access to that, like inner child that you've been suppressing for so long for, all the societal reasons that, you know, or any other reasons you've discussed.

And so I would say, if you haven't tried it out yet, just just a recommendation of something to throw out the board there.

[00:26:18] Ty: Quick story there. I'm familiar with the Muse headband. Yeah. Yeah. Where it's got the EEG or EKG. I can't remember which one it is, but like it reads your brain activity.

And, in theory, if you're in a meditative state, it can tell like what the level of brain activity that you're in. And I found that, I was playing with it as like a tool at home. And I found that the mental state I found myself in when I'm running is equivalent to a meditative state.

Like I'm able to drop into that. So I don't know, it depends on I don't know, I was a cross country runner and whatnot. And so there's different ways to access that kind of, I don't know, like head clearing space.

[00:26:56] Jared: Yeah, like that alpha theta wavelength. That's really, that's interesting. And so did you, okay, so now we've talked a lot about sort of creativity framework, and I think one piece of being creative or a major piece of being creative also plays into the design thinking world. And it's a muscle that you have to have to be a great design thinker. I haven't personally met a design thinker that wasn't also pretty creative. But then at the same time, one of those things that I think works really well in this world of sort of ideation and coming up with new ideas is artificial intelligence and large language models.

And I think AI gets a bad rap these days for, all sorts of things. You just saw open AI, they got sued by the New York times for plagiarism. We'll see how that plays out. And it gets a bad rap for hallucinating and you don't want to you copy and paste it.

And I've definitely agree. I don't think you should just be straight copying and pasting any output from it directly. It's definitely a tool in your tool chest. But at the same time, it can generate tons of ideas. And so for people that are having trouble with that creative muscle and having trouble ideating, this is something that chat GPT or any of those competitors do not struggle with whatsoever, you can keep ideating, keep generating, and it can at least be maybe a kickstart to your creative muscle.

And so I think, Ty, I know you have a lot of thoughts on this as well, so I want to let you dig in.

[00:28:24] Ty: Yeah and I gave a talk on this very topic that you bring up here, Jared, which is, and this was for the Society of Physician Entrepreneurs, Dr. Arlen Meyer's organization, which was looking at design thinking and healthcare innovation.

And I'll just pop a couple of these slides up. . And it also, some of this is a reflection on, design thinking for healthcare innovation, but also where AI is going. Because how we utilize this tool can either be, it's a tool in its own right. And so the method and approach we go through about trying to make sure we have alignment with.

Human interest and what the AI can produce I've been looking at this as an empathy first approach to AI to try to, make sure that as we're using these tools, that they're helpful and beneficial because you're right, like AI does offer like generative large language models do offer some of that sense of I don't know, just like a quick answer that it can solve the blank canvas problem of staring at a like at a white sheet of paper and what do I put down first?

You can just start to think into the prompt and then get a response that you can react to and start editing. I will say that one of the ways I like to anthropomorphize or personify large language models is that it's a sociopath and a liar. Meaning that the only moral compass it has is what the programmers put into it.

Because GPT means that it's been trained on the entire corpus of the internet. So all of the text that we have ever produced as humans and put on the internet has gone in to feed the model as the data set. And And that was one of the things that OpenAI, when they first produced it, was they then had to put boundaries on what it could do.

And then say you can't go and use this for nefarious purposes and program that into the software. But if the programmers haven't anticipated it, then it's a sociopath, right? It like doesn't have a moral compass to it. It's also a liar, meaning that it tells you what it thinks you want to hear.

And when they say hallucinate, it means like it's trying to sell you something it's trying to tell you it's reflect back to you what it thinks you asked for and the truth is not, it doesn't have a grounding in truth, like other computer models you've interacted with before. It's a new thing we have to figure out how to work with.

As long as you understand it to be a sociopath and a liar, it can also be a useful idiot and that it's got a level of intelligence because it is accessing the sum total of human knowledge. In order to create a beautiful story for you. As long as that's a useful, beautiful story and you approach it skeptically, then it can be meaningful and helpful.

But I don't think we've been with the technology for long enough in order to have fully anticipated everything that could go wrong with it. And it's also moving very fast where the capabilities that it has now compared to a year ago when we first had Dr. Harvey Castro on to start thinking about this and looking at it, it's just night and day different in terms of its capabilities and has continued to grow at that level.

So I think then it's stepping back to some just design principles for AI, which is where I really like the the concept of traffic lights versus roundabouts meaning, if you think about from a traffic design standpoint if you are setting up AI to be where it's going to be traffic lights, where you have much more centralized control and gates on it, then you're removing agency from the local individual and rather trying to set up centralized controls that have more and more centralization.

Versus traffic circles have less accident points than a stoplight and allows for more local agency for the individual driver to look for oncoming traffic and make a local decision on what's the best use of this tool, how to proceed with this intersection. So I personally prefer more decentralized design methodologies.

If we can, when we're approaching a problem, try to implement traffic circles rather than try to be like central control impulse to try to like, tightly control everything and put traffic lights in place. Just as a design principle, that's one thing I like to put in place. and Dr.

Ben Schneiderman wrote a book called human centered AI and that came out in 2022. And so he 1, this is 1 of my favorite frameworks out of the book, but he talked about this 1 dimensional thinking that has been, it's also this false dichotomy that we talk about traffic lights and roundabouts.

Is that if you reduce human control, then you're increasing computer automation versus or you can move back along this continuum that as we bring AI into place, we're taking more and more humans out of the loop. And so we have to actually break out of that one dimensional thinking when we're approaching AI, because really, it's a two dimensional concept where you've got high and low degrees of human control.

We've got high and low degrees of computer automation. And so somewhere in a use case where you've got low human control and low human computer automation, that's where you've got basically like ditch digging low tech, highly labor dependent like activities versus high level of human control versus low computer automation.

That level of human mastery, like playing a piano, being a brilliant chess player, like those are things that individual humans build up a unique skill and expression that are high in human control, and you're not relying on a computer to do those tasks versus high levels of computer control. This gets down to a pacemaker.

Where you don't want a human interacting with controlling the pace of the heart and how the pacemaker operates. In fact, any time a human interacts with it, you're really trying to eliminate any kind of human control over that because that's an autonomic system. And you shouldn't have human control interacting with this.

And so then we're trying to maximize reliable, safe, and trustworthiness where you've got humans in the loop. And computer automation to enable a more, a better outcome that's more efficient. That's ultimately the goal, is not giving up human control, but rather making it so we're making better decisions as we go about the traffic circle with our, AI helpers, right?

So that's a, framework for that. And actually one of my examples that Dr. Schneiderman gives in the book. Is he looks at morphine dispensing and the lowest tech, most labor dependent thing is like a morphine drip bag where it's up to the patient to just, or you can just have a raw feed of morphine that goes into the patient.

There's no control over it, except for what the nurse comes and clamps off and that's pretty risky. You know where you can overdose on morphine pretty quickly versus high degree of human mastery where it's up to the patient to guide how much is coming in there. And that also carries particular risks.

That's the local judgment of the patient. Versus we then say, you know what? We're just going to take humans out entirely. We're going to have an automatic dispenser, which is where the FDA gets really nervous about that because any failure then falls back to the system design and we've removed expert control from that.

Rather, I think the design ideal from how to dispense morphine for a patient who's suffering and doing pain management is making sure the patient has input to it, but it's clinician monitored to make sure that no adverse events happen. And so this is a good way of thinking about how can AI fall into just pain control from a design standpoint.

So there's a couple other ones.

[00:36:32] Jared: That was interesting. Wait a second. So you're proposing that the patient should be able to administer morphine based on their pain levels.

[00:36:41] Ty: They should have a say in how much morphine that's coming in, like doc, I'm in pain, help me with this, but having controls in place as well, so that.

It could be that there is a, like the patient doesn't need more because it would have an adverse effect, but rather there could be a placebo effect that comes in to the patient will feel better if a placebo is administered.

[00:37:05] Jared: I like that. I like that because I was thinking, just the hospital, how are they going to trust patients and they're going to be.

ODing themselves over and over.

[00:37:14] Ty: This is a tricky topic to try to get through, right? And particularly if you're getting into how does AI play into this. The answer is not to just give up human control.

[00:37:25] Jared: It would, yeah, and it would be interesting because then, if it, if you are giving them a placebo, how do you keep that big secret from the general public?

In theory, these devices are going to be publicly available information for the most part. And so then this really placebo effect only works on people that are completely oblivious to medical technology, which is quite, quite a few people. And then I guess for the people that know, they just know, and the secret's out and maybe it won't work on them.

But no, I saw that. I think that's really interesting. I was trying to imagine what that would look like,

[00:37:57] Ty: yeah. And I think if the patient isn't aware, I can't tell the difference between placebo and non placebo, then I think the psychological effect would be the same.

Honestly. Yeah. Even if they're aware that a placebo might be used and honestly, if that's, I think somebody who's in pain probably isn't going to be reading the literature on that either. No, but that's just an example of like how this two part framework comes into play and there's other examples too, that he talks about.

Which as we've been playing with large language model designs, that's what I think is entertaining, right? So Jared, you've got your you come back from say your walk and you come back with lots of ideas. You've just got a Evernote document going with all of your ideas, right? You come up with.

Yep. Great. And so one level of just there is some automation in place where like. There's some spell check in there. Like we have some augmentation already that comes in with how we're digitally like writing things down. And then, I don't know, did you ever play with God mode?

[00:38:59] Jared: I couldn't figure it out.

The API wasn't working for me. So

[00:39:03] Ty: it took me it was a solid weekend of one of those side projects where I was finally able to get the API to work where so this is a tool that allows you to connect your chat GPT account to this program that will like automatically run prompts through chat GPT.

And you can say, Hey, I want you to build a website for me. And assuming, God mode could actually function, then it would then go through, I think somebody said it to I don't know, like plot to take over the world. Oh my gosh. And just okay, we're going to put this in God mode and see what happens.

And I was playing, I didn't give it that parameter, but I gave it like a, Hey, design a new website for me as a parameter. And what was interesting was that just the AI by itself with a very basic parameter like that, it spun its wheels. And just didn't get anywhere. And it was the conversation that then said actually, that's not what I'm talking about at all.

What I'd like for you to do is this, and it was that back and forth between human and AI that started to make progress. Eventually, it fizzled out and didn't get me anything useful but I gave up on it. But it was interesting to see how little the AI can do without human help.

Yeah, versus, I think there's a interesting thing we could do with say like a natural language survey for patient history. And so then that's where you're then going through and you throw in like a large body of text. It'd need to be HIPAA compliant, but throw a large body of text for, say, the doctor's use, to then say, please interrogate this information and return back information about this patient history so I can have a quick summary and save me the time of reading back through all of this stuff.

And so that could be one way to utilize large language models, but train it on real data. So it's got a ground truth that you're feeding it, but then allowing for a more natural conversation with it so that you can more quickly summarize information. So those are just like different applications and different ways that, the design could come into play.

[00:41:11] Jared: And I see a huge need for this as well. Like I'm on a Part of my TikTok algorithm is healthcare and healthcare related issues. And so a big thing there right now is people talking about they're basically mad about their doctors or their physicians, their specialists not being able to diagnose them they're having certain things in their blood work go on and that don't really make sense. And the doctor just goes I don't know. And I'm like, yeah I can see why people would be so mad, why you would see that as unacceptable, but if there was a tool for the physician to use that could help them out to make sense of some of this data, because at the same time, yeah, they're, of course, physicians gone through so much school, but they still are not an encyclopedia of every medical outcome that could possibly happen.

So I think that using this, what we've talked about before, we can be friends, the expert paired with the AI, I think that's going to be really where we start to see a lot of really beautiful things happening. And this, I hope this use case actually plays out.

[00:42:09] Ty: Yeah. And I think these are just examples of, you can switch your topic out from other like big intractable problems and play with this two by two matrix and figure out so if we go all the way to human mastery on one end.

Then you arrive at certain types of solutions, or maybe that's like our status quo versus we then have the opposite end of the spectrum where we have complete computer control over the situation. And so then you have centralized errors that happen at scale versus trying to then have the most safe and trustworthy, where you continue to retain the expert.

And you then benefit from the access to a large data set in real time. I think that's where you start to get meaningful, interesting new outcomes. Okay. So one of the things that we did in this exercise, and I don't think we have time to get into it, but was to what I found surprising is that there's a lot of people talking about AI, but not many people are following the midst framework.

Okay. And actually trying it. And so the exercise that I gave during the talk, and I encourage you to check it out, was to have all the participants go through and pick their favorite large language model. And then to go through and then put this prompt in, which is, list solutions to, and then in quotation mark, put the problem statement in.

And the problem statement we had, the job of keeping a record of the patient's medical history often conflicts with delivering care to the patient. So that's a problem statement that I think anybody in clinical care, there's an overwhelming amount of documentation that happens as a result of electronic health records that is less efficient than paper records that we move from, right?

So there's a whole thing that electronic health care management has led to more work, not time savings. Okay. So then the prompt that if you have your chat GPT window open is to put this exact prompt in list solutions to this problem statement, where a combination of human mastery and computer automation lead to reliable, safe and trustworthy outcome.

So for anybody listening at home or actually we'll put this in the show notes and you want to try this. Following our framework, right? Then I would say, drop this into your chat GPT or whatever the Microsoft thing is, I think Microsoft's going to own open AI anyways, but like whatever your large language model and just ask it to generate, say 10 solutions to this statement and then see what comes up, by the way, what comes up will be different for each time you do this because it's generating a fresh response each time, which is really fun to say, okay, generate 10 of these for me. Okay, cool. Generate 10 more and it just like generate 10 more and keep doing it. As a wild exercise, try to get it to go 300 times for you and just see what comes up with. And it's in the wealth of ideas that come back.

You might have something that's novel that the large language model might pull from a completely different industry. In order to try to get that next idea, but it might, generate something for you that in your job as an innovator shifts from generating these ideas, the large language model can generate lots of ideas because it's trying to make you happy.

And it's a useful idiot, right? And so your job shifts from generation to curation, and now you're applying judgment, and then sifting through all these different ideas to try to come up with what's meaningful, what's valuable, and attribute meaning and story to an idea where this could actually be pretty good, and could start to address this problem.

Anyway, as an injunction to everybody who's listening to the podcast, I want to encourage you to play with that. And yeah, and I can drop in that into the chat and the show notes as well.

[00:46:09] Jared: Another thing that I think is interesting about just the integration of AI into sort of the workforce, I guess you could say it's, there's this group of people that I'm sure you're very well aware of that are like there's certain people that think, Oh, it's going to replace us.

And that's one way of looking at it. Then there's the other group of people that don't even want to integrate it and they think Oh it's not ready yet. And then there's the people like yourself that are like. Ooh, there's this new tool. Let me find out and see what it does. And let's break things along the way and figure out, how best to use it.

And so I guess, for yourself, like what's the advice you have to people that are not using this technology yet, they're afraid of it. They have maybe an adverse view of it. And how can you shepherd more people along just this sort of new tool, like I said, it's not the end all be all, I don't think. Of course it's not the magic genie. That's going to do everything for you, but in the day being a useful tool.

[00:47:06] Ty: Yeah, it gets back to what we were talking about earlier about it requires commitment to exploring the boundaries of what's possible. It requires some courage. In some of this you're pushing the boundaries of what the technology can do.

And those people who are on that leading edge of this are the ones who uncover a new value that nobody else has found yet. There's a moment of delight. And if a positive feedback loop for Oh, I found something that's this and playing with these tools that don't really work yet and everything, but you can start to see that spark.

And then also you can follow it as it comes to fruition. I think you can also then start to, in a safe control environment, start to see where it doesn't work. And by playing with it, by experimenting with it, then you have a better concept of it. And actually, I think you gain more confidence and this thing isn't going to do what I originally feared it was going to do, but this one might be like, it's creepy.

And so just you have deeper understanding of it by playing with it. And so I think it's a transformational new tool That I think we've seen change some of our workflows because it does make certain tasks easier And then it's annoyingly not helpful in other contexts and you just have to work with it in order to figure out okay So I have a fancy new like circular saw, and it has some really cool laser guided tech to it.

And like the laser kind of works and it doesn't. And you have to sit there and play with it, like any tool. So as designers, we use 3d CAD pretty extensively. Just as an example of like technology that when 3D CAD first came out, there was this fear that there are gonna be a lot less designers, a lot less engineers needed because 3D CAD software is going to make the work so much more efficient.

As it turns out so there was a certain amount of time that was available when you were doing paper drafting of a new product. And then that same amount of time has still been in play. It's just the level of complexity that same team can handle has increased tenfold, but the time it takes hasn't changed.

And then you'd probably have more people that are more specialized to use these 3D CAD tools as a result of that. So I think AI is a tool that I think is going to be a. It's a really critical part of being participating professionally in the future, and the more that you understand its strengths and shortcomings, the opportunities and risks, just the better prepared you'll be as this tool continues to change on an almost monthly basis.

[00:49:47] Jared: Yeah, and I think just last little piece on this before we head out is. Okay. So maybe the financial opportunity that people have as well as a subject matter expert to also train your own GPTs. Have you seen that, Ty, that you can do that? I haven't. Please tell me more. Yeah. So basically if you're a subject matter expert in any niche whatsoever, and you have proprietary information that you can train the chat GPT model on, you essentially share it with them and they brand it as your GPT.

You name it as whatever it is. And then off of your expertise that you've shared with it, it now creates a sub language model for you. And then based on the amount of usage that your GPT gets is they pay you out from that usage. So it's an interesting model. I think it's almost the model that they're going to be going to, because of this lawsuit from the New York times as well, which is having to pay the people that have actually trained the model and a people for the output that it's getting.

But yeah as people out there that are subject matter experts, I think it's an interesting way to Also train a model based on your expertise and then mess with it after and see like what output comes out of there. And if you're happy with it and almost like testing is that the information that I'm putting out there even sufficient enough?

Because this whole language model is just learning from you only. And so if the output, it's not giving you. Sufficient. Maybe you need to relook at some of the materials you've got going on and then materials that you're sharing with folks. Maybe as a litmus test in a way, so

[00:51:25] Ty: it reminds me of one of the early things I was like trying to figure out what this is and it might be fun is someday let's say Gary V who has so much content out there.

Could just use just train a model to be the Gary V AI and just, you could have a private Gary V come talk to you and here's the AI model that you're then interacting with that point of view and I don't know, it's some form of like subscribe to access your favorite persona personality.

Maybe it's a subject matter expert, but it's a if you've read what was it? Re shift. What's his name? But he talks about basically human consciousness being uploaded into the virtual reality and yeah, exactly. So it's like a, Version of training the model into the specific mannerisms and whatnot of individual humans.

And so anyway, it starts to approach that except you'd have to spend a lot more time with it, I think, to get it there and you need to get a lot of training data, but it was fascinating that we're starting to go that direction.

[00:52:23] Jared: Yeah. Yeah. So I think we've left everybody with a lot of food for thought today.

Yeah. And you just want to thank everybody that was listening in today. Thank you for joining us for this fun, creative journey that me and Ty were on today. And definitely go try out where we will leave those the prompt in the notes for your chat GPT exploration and in whether it's on YouTube or whether it's on LinkedIn, please share with us your results.

Cause we're really interested to find out. So thank you so much. And we'll get, see you next time.

[00:52:50] Ty: All right. Thanks everybody.