Andy Carroll - Healthcare: The Perfect Storm for Disruption by AI

The conversation this week is with Andy Carroll. Andy is not your typical chief commercial officer. He is a Six Sigma Black Belt and is just as comfortable getting into the weeds for process improvement as he is using systems thinking to develop go-to-market strategies. As a President's Club and Innovation Award winner, he leads high-performing teams to enjoy exponential revenue growth and spark innovation. He offers his expertise through consulting services to healthcare organizations and adds to those skills through graduate studies in artificial intelligence. He currently is a chapter advisor for AI 2030 and its mission, utilizing AI's transformative power for global good while mitigating its risk as well as fostering community engagement and development.

If you are interested in learning about how AI is being applied across multiple industries, be sure to join us at a future AppliedAI Monthly meetup and help support us so we can make future Emerging Technologies North non-profit events!

• Emerging Technologies North
• AppliedAI Meetup

Resources and Topics Mentioned in this Episode

• Slainte
• AI 2030
• Regression analysis
• Predictive analytics

Enjoy!

Your host,
Justin Grammens

Transcript

[00:00:00] Andy Carroll: Healthcare is really the perfect storm for disruption within AI. And you think about industries that are good candidates for AI disruption. They have to have big data, big, big, big data. And what's unique about healthcare is that data is both highly structured when it comes to electronic medical records.

And you have companies like Epic, you know, that are the big players, big structure, same fields across massive data sets and going back. A very long time, and then it also has a tremendous amount of unstructured data in terms of clinician notes, images from diagnostics and scanning and things like that, that AI is really good at looking at a very large scale to come to conclusions and make recommendations.

[00:00:45] AI Speaker: Welcome to the Conversations on Applied AI podcast where Justin Grammons and the team at Emerging Technologies North talk with experts in the fields of artificial intelligence and deep learning. In each episode, we cut through the hype and dive into how these technologies are being applied to real world problems today.

We hope that you find this episode educational and applicable to your industry and connect with us to learn more about our organization at AppliedAI. mn. Enjoy!

[00:01:16] Justin Grammens: Welcome, everyone, to the Conversations on Applied AI podcast. Today, we're talking with Andy Carroll. Andy is not your typical chief commercial officer.

He is a Six Sigma Black Belt and is just as comfortable getting into the weeds for process improvement as he is using systems thinking to develop go to market strategies. As a President's Club and Innovation Award winner, he leads high performing teams to enjoy exponential revenue growth and spark innovation.

He offers his expertise through consulting services to healthcare organizations and adding to those skills through graduate studies in artificial intelligence. He currently is a chapter advisor for AI2030 and their mission, utilizing AI's transformative power for global good while mitigating its risk.

As well as fostering community engagement and development. So thank you, Andy, so much for being on the program today.

[00:01:59] Andy Carroll: Well, thank you, Justin. That sounds so much more elegant than just a sales guy that got into artificial intelligence. So

[00:02:07] Justin Grammens: I appreciate it. Yeah, well, you're not just a sales guy. You definitely, I see you out at all the events.

You're speaking on a lot of this stuff. So you've been going deep in it for sure. So I think you're going through graduate studies at St. Mary's University, I think. Is that what it is? That's right. Yeah. So you're, you're working to get a certificate in artificial intelligence. So that's awesome. So that's kind of where you are today.

I, I always like to talk to people about maybe rewind the clock back a little bit, you know, whether it be five, 10, 15 years, however far back you want to go. But I'm sort of curious how people sort of moved into this field. Maybe what was the trajectory of your

[00:02:35] Andy Carroll: career? I often look back at that and ask myself, you know, how and why and I have a chuckle about it because it's always a happy circumstance and I saw a snippet once that said, every seven years within your career, you know, you will have some impactful or exciting change that leads you in a new direction and I'm dead on that cycle.

I came out of undergrad and backed into health care because I came out at a time when I was supposed to be with Arthur Anderson and they had the one two punch of Enron and 9 11 and I ended up at a company called Hewitt Associates that did benefits and HR consulting and outsourcing. They've since been bought by Aon.

And then seven years later, I moved to Kaiser Permanente that got me really into healthcare and empowered me to think about how healthcare should be a force for good and about the patient and how the system is broken with perverse incentives. And I had a great time there and did well. And then seven years later, I went back to school to get an MBA, where I learned that my values are a part of my everyday decision making and how I act as a leader and how I present myself.

And when there's conflict between your values and your work life, that tension needs to be resolved. And I've always incorporated my values into my work and my decision making. And lo and behold, seven years later, I was working for the non profit Medical Alley and was let go from that job and was in the market.

And here's this thing called artificial intelligence that is prevalent everywhere. It was a couple of months after CHAT GPT was released to the broad public. And I was fortunate enough to leverage the displaced worker program to go back to school. And in a day and age when Folks were taking Coursera classes online and maybe, you know, MIT or a Stanford, you know, class about an AI overview.

I wanted something I could put on my resume. And the reason I wanted something on my resume was that I saw that 30 percent of all job descriptions now have some aspect of AI. And we have few opportunities to pivot within our careers, especially me, you know, who's mid career or other folks who are late career.

And this is such a transformative technology. I think it's important to look at where you are and how you can incorporate these skills into your work. And position yourself to finish your career the way you want. Uh, so I'm excited to get into it. It's an eight month program. It's about a third of a master's and it's technical.

I'm learning Python right now, Justin. It's not pretty.

[00:05:05] Justin Grammens: Sure, sure. Yeah, going from an MBA to writing code. That's, that's definitely a, a different experience, I'm sure. Yeah, and I, I know what you mean about all aspects kind of have AI. I mean, people to think about today, you know, it feels very futuristic.

This whole sort of chat GPT thing that, that's there. You can actually ask a questions and stuff like that. But. I think we're going to look back in three to five years and kind of say, well, geez, that was actually pretty primitive because it's not so much even like the usefulness of it. It's how much it's going to change our jobs.

You know, how much it's actually going to change how we do our jobs. I'm already using it in a lot of different ways that I never even thought about six months ago, for sure. And no matter what you're doing, what career you're in, it's going to touch it. So it's good. You know, for people to really, of course, kind of at all levels, there's people that want to dive in deep.

Sounds like that's where you're looking to do. And then, you know, is your goal then to kind of continue to do consulting and help organizations figure out how they can use this new technology as you've been digging in?

[00:05:57] Andy Carroll: I think that is the end goal. I see a vacuum in the market coming from a commercial standpoint and working with CEOs and board members.

I can help bridge that gap between the engineering team and the boardroom to explain why this matters. Why as a board, you need to have the right approach for governance to understand how these AI tools are developed and how they fit in with the values of the organization, whether the organization is doing the right things in terms of transparency to develop trust within their employees and their customers.

So I think there's a lot of opportunity there, particularly within health care, since I've spent my entire career in health care, I'm passionate about it, and I think it's probably the ideal. Industry for disruption with AI. In the short term, I'd love to get back to work and, you know, and spend some time supporting an organization that's looking for growth, but I think the future is probably, you know, in this consulting world.

And I think there's going to be a lot of demand. Yeah,

[00:06:51] Justin Grammens: for sure. For sure. And having, you know, myself running a consulting shop, I totally understand that you basically need to walk all the walks, right? You need to be able to. have a team or yourself be able to, you know, be a thought leader, sit in the boardroom, think about that level, but then if you can get deep into it and fully understand kind of the nuts and bolts, certainly doesn't hurt.

And you've focused on healthcare, which I think is, is fabulous. Maybe you could talk to our listeners a little bit about why you think AI is so revolutionary in healthcare and like, why is healthcare different than AI and maybe some other industry? Yeah,

[00:07:23] Andy Carroll: healthcare is, is really the perfect storm for disruption within AI.

And you think about industries that are good candidates for AI disruption. They have to have big data. I mean, big, big, big data. And what's unique about healthcare is that data is both highly structured when it comes to electronic medical records. And you have companies like Epic and Cerner, you know, that are the big players.

Big structure, same fields across massive datasets and going back a very long time. And then it also has a tremendous amount of unstructured data in terms of clinician notes, images from diagnostics and scanning and things like that. AI is really good at looking at a very large scale to come to conclusions and make recommendations.

In addition, I mentioned earlier the perverse incentives. You have a broken system. It's broken from the patient standpoint where it's maddening to go through it. Especially if you've ever had to have severe care or complex care for yourself or a loved one. It's broken from the provider standpoint.

There's a tremendous shortage. We'd have to have about a half a million new providers amongst primary care, nursing, and mental health just to fill the gap today. So the math doesn't work to fill that gap, and we need to look to technology to help empower clinicians to operate at the high end of their credentials and get out of the burnout, you know, endless cycle that began way before the pandemic.

And in addition, there's a lot of money, you know, it's a four and a half trillion dollar system. Uh, it's bigger than most countries in the world. So there's a lot of opportunity. And I think healthcare is the perfect

[00:08:53] Justin Grammens: example for it. Yeah. Yeah. And it seems like Minnesota, I guess maybe that's the other thing is just the location.

It seems like this is one of our staples, I guess, one of the industries that Minnesota is sort of known for. It is,

[00:09:04] Andy Carroll: you know, Minnesota has the second highest per capita healthcare population in the country, only behind Massachusetts. So we have half a million people in this state that work within healthcare.

And as someone who didn't grow up here, that was always one of the appeals. When we moved to Minnesota and being a healthcare guys, well, what better place to be, you've got United and Mayo and Medtronic and all these wonderful, you know, healthcare companies, and then the small to midsize businesses that are very innovative and high growth and it's.

It's a wonderful place to be.

[00:09:35] Justin Grammens: You mentioned Kaiser Permanente as sort of somewhere along in your path and they're not here. I mean, I first heard of them when I lived out on the West Coast. So you must have been someplace outside of Minnesota when you worked there. Is that

[00:09:45] Andy Carroll: true or were you? That's true.

We lived in Chicago a long time. I found my way there after undergrad and lived there for 18 years. I grew up on the East Coast in Connecticut and I fell in love with Chicago because it felt like a cleaner version of New York without garbage everywhere. And it had all the culture and the good food and all the things I like.

So Kaiser Permanente has a small office in Chicago to loosely cater to the large employers based in the middle part of the country. So it was a wonderful organization, mission driven and absolutely massive and mostly based on the West coast, but it's about a hundred billion dollar nonprofit integrated system.

And back to those incentives, I think that's where it's important to understand the difference between fee for service within healthcare, where you literally get paid for every CPT code that you're able to bill, versus an integrated system that's paid, you know, either on a capitated model where you get a set fee per patient per month.

And it's up to you to take on the risk and, you know, and figure out the right treatments and care for that patient. So health partners would be a comparable model here in Minnesota. And then Mayo Clinic somewhat adjacent because they're also physician led and are mission driven and take a patient first approach.

So walk me

[00:10:53] Justin Grammens: a little bit through then how AI can actually help. I mean, it sounds like. You're trying to be more efficient, I guess, across the board, right? I mean, it's, it's an open market. So the people that are more efficient are the players that are going to ultimately, I guess, win out in the end. So if we're looking at artificial intelligence through, you know, one of these two models, this fee for service or this, or this other integrated model.

How are you seeing people use AI to sort of win in this, in this space?

[00:11:17] Andy Carroll: Well, at the basic level, and you see a lot of technologies that are in place today to do this, you're freeing up clinicians time. The average doctor spends three hours a day coding. So it's typing up physician notes, either before, during, or after a visit, typing up follow ups, following protocols.

They're spending a lot of time on administration, and that does two things. One, obviously it takes a lot of time, and it's inefficient. Two, it shifts their focus from the patient to the technology. So AI is a perfect example where it can summarize a patient visit, similar to how you might use AI to summarize a conference call.

And it can connect with the record to highlight bullets that the doc might want to be sure to cover and make recommendations for next steps. It's not replacing the care, it's just augmenting it by saying, hey, in certain cases we've seen in the past that fit this description, here were the typical protocols and treatments.

So it's like an over eager assistant that has access to a whole world of information. Similarly, in, you know, diagnostics, it can look at images and say, okay, here's cases we've seen in the past, here are issues, and here are recommendations that we see, you know, or this might be a positive case, or keep an eye out for this, and it recommends to the clinician, you know, here's what to look out for, and here's what to expect, so it's very helpful, particularly at a time When physicians are burnt out, I always book my medical appointments in the morning because I want the docs when they're fresh, you know, at the day, they've seen 20, 30 cases.

Think of your, your hardest day when you had back to back to back meetings. That's every day for a doc. That's hard. Yeah. You know,

[00:12:48] Justin Grammens: I mean, some of this stuff around transcription has been around for a long time. I guess some of the, the large language models and stuff maybe are a little bit new, but I mean, I feel like the stuff's there.

I've seen this in so many industries. Like, why hasn't this adoption really taken off? I still feel like we're still sort of. talking about how AI could be used in health care. You know, is there something unique about health care? And there probably is, but, and then, you know, maybe why haven't we been able to move this faster?

[00:13:13] Andy Carroll: Well, change is hard. And change is particularly hard when the stakes are high. And the concerns within health care are data privacy. Both clinicians and the patients and the systems. They all need to be absolutely sure that the data is safe and that it exists in an environment that is only being used for decision support, so that's hardened itself.

Physicians are notoriously slow to take on new technology. Back in my Kaiser Permanente days, I worked within the innovation lab and it was basically their proving ground for new technologies and innovations. They had a mock OR where they would bring in machines and anything that the rep was selling to say, okay, prove it works here and then we'll implement it broadly.

And at the time, this is back in about 2013 or so, they did a study and they found out that clinicians were entering their credentials into the computer 75 times a day. So they were testing out biometrics for logins. And now those are prevalent through, you know, fingerprints or key cards. And it was awful to get the clinicians to get to the point where they could switch from paper to electronic, from electronic login to biometrics.

And it's always, it's always slow. So I think it's hard to adapt to change, particularly when you're ingrained in a routine. Interestingly, even as recently as 2016, Kaiser was exploring augmented reality for IV pricks, where the AR would show the veins on the arm and make it much easier for the nurse or assistant to get a blood

[00:14:41] Justin Grammens: draw.

Interesting. That sounds really cool. I mean, so there still was a physical arm that was there. Basically, the augmented reality would sort of put the veins over it. Or would this be I can almost see this being used for practice

[00:14:50] Andy Carroll: as well. Yes, absolutely. And I think, you know, with the new Apple Vision Pro coming out and you're beginning to see those sort of capabilities with real time feedback and augmented reality, I think we're going to enter a whole new world of, you know, surgery and other applications with augmented reality.

[00:15:08] Justin Grammens: For sure. Well, you know, AI is such a big term too as well. I oftentimes when I have people on the program, I'm like, what's your definition, you know, of it? Or even if somebody asks, what do you do? You say, well, I consult people on AI. Like. How do you sort of boil that down into something that is applicable to the average person that might be listening to this podcast?

[00:15:25] Andy Carroll: Yeah, it's almost overused at this point, Justin, you know, like, uh, people that have spent a lot of time in this space know, you know, AI has been around for a very long time and whether it was, you know, predictive analytics or the implementation of big data solutions, everything's called AI now, so I think everyone just has to use the term, but what has really super powered the shift from, you know, whether it's advanced multiple regression algorithms or predictive analytics to what we see today is Tremendous amounts of data and really powerful hardware.

And because of those two things, now we're able to do some really cool stuff that we were limited to be able to do before. So, at the end of the day, it's still all math. And I think it's helpful to reiterate that, you know, like whether it's a large language model or GPT. It's saying, you know, that's not truth or versus a hallucination.

It's all math and it's all probability. The models are just really, really good at it. And they just happen to produce the next word or pixel that makes sense to what you're asking it to do. But it's just really, really powerful math using big data. And once you unpack that and unwrap it from the mystique box of artificial intelligence, then people really begin to understand where it comes from.

How it can help, and the fact that it's just data and powerful hardware. It's just a matter of harnessing it the right way to accomplish your goals. Yeah,

[00:16:50] Justin Grammens: yeah, yeah. Good. It's a good way. And as you were talking about, you know, the complexities and the math and all that type of stuff, I'm curious, like, what are you coding today in Python?

You have an example project and stuff that you're, that you're working on or?

[00:17:03] Andy Carroll: No, I'm, I feel like a bit of an imposter because I've learned, you know, I'm, I'm fine on the statistics. I've learned some data engineering and I'm just kind of fumbling through Python right now. So I don't know if I'll ever get to the point where I can properly code.

I think. I can get to the point where I can look at code and understand what it's doing. And I can augment code by utilizing GPTs, copilot, you name it, to help, you know, maybe change it or do something else. Since I'm still in the program, I'm still learning and I'm trying to figure out what my project should be.

And what I'd like to do is maybe start some for my own personal use, especially now with the use of narrow GPTs. I heard a great use case, I think I'll definitely do, where you take a little narrow GPT and load up every instruction manual for every piece of tech you have in your home, whether it's your oven or your Bluetooth speaker or your router, and you can use it to say, Okay, I've got a flashing yellow light.

What does it mean and what do I need to do? Or what is the recommended maintenance schedule for my oven? And just stay ahead of it because we're, you know, we're so ingrained in using technology now, particularly with the internet of things and smart homes. It's nice to have that tech support in one little place.

And I think AI is great for just helping us gain back that extra 10 or 20 minutes a day. So I'd like to do some personal projects and maybe find some opportunities to do some broader client projects as well. Yeah, that's cool.

[00:18:28] Justin Grammens: The thing that's so nice about a lot of this stuff is there are libraries now that people have developed over the years.

And so, you know, you can stay higher level on some of these things and not actually have to get into the nuts and bolts. And obviously with things like OpenAI and their API. You know, Google, Azure, all these things, all these companies basically have ways for you to interface with them. Uh, and if you're okay, you know, with them ingesting your data, it just makes it a lot easier, right?

And so you can sort of stay at the surface level. I think, you know, where all this is going here is I'm curious to, you know, because you were talking about, well, I might not actually be an expert coder in the future. And my, my argument would be like, you don't have to be, man. You probably will never have to be.

Because what I've been seeing and what I've been experimenting around with is a lot of these, you know, code creation tools. And I'm very, very impressed with what it's able to do sort of out of the box. And so, you know, as you step back and sort of think the next 10 years, maybe for yourself, but also you just even in healthcare, are you concerned at all about the future of work, the future of maybe what doctors and nurses or people that are in the healthcare system are going to do?

what yourself as a consultant is going to do. How do you sort of see the landscape of AI sort of playing out in your, you know, lens?

[00:19:32] Andy Carroll: Well, I think that's where a lot of people begin to get nervous. And it was our our friend Matt Versace who shared, you know, he's like, AI is not going to take your job. The people learning AI are going to take your job.

And I think everyone needs to learn to coexist with the technology. And, you know, just like any technology that comes along, there's always an adoption curve and, you know, folks that are slow to adopt it may lose out in some efficiencies. I don't think it's going to be mass layoffs. And even Google recently had a few layoffs, but there are always rumblings of that and the big impact on the white collar workforce for AI.

But disruption is Typically a good thing. And I'm an eternal optimist with these things too, Justin. So I think it's going to empower everyone to operate at the high end of your credentials. And it'll just take out the mundane tasks like, you know, drafting emails or note taking or scheduling or connecting with people.

Of all the AI technologies I'm most excited for, I want agents. I want to talk to my agent and say, Hey. Get with Justin, let's schedule a podcast and I don't have to worry about it and it just happens. And it's that connectivity amongst different disparate tools, like having personalized medicine combined with a shopping list and recipes for healthy foods that you can easily reference, buy at the store or have delivered to your house, and then start cooking.

And it aligns well with your treatment plan for your own chronic conditions. And then you've got your monitor, you know, to make sure that everything is going well. It's stuff like that, I think will help address labor and provider shortages. It's going to help people operate at the high end of their game.

And yes, there's going to be some disruption, but there's also tremendous opportunity. Yeah. So I don't, I don't advocate that everyone needs to stop what they're doing and do a graduate program in AI, but I think it's a great opportunity to raise your hand and say, Hey, this is interesting. I'm going to learn about it.

And. You know, maybe I can be a good resource and position myself for success in the future.

[00:21:29] Justin Grammens: Yeah, for sure. Well, yeah, you're talking to someone here who's a lifelong learner. I continue to sort of move and learn new things throughout my career. So it's really good. And you know, a lot of it is, like you said, you don't need to actually take a program.

A lot of it is just experimentation and just Googling around and just trying tools. It's amazing. How much free stuff there is out there, there's all so much freeware around, you know, generative AI in particular, that it's just, there's no reason why you shouldn't actually just experiment. And I, I actually for this podcast, well, I do the podcast, but I also have a weekly newsletter and I started experimenting around with.

Hey, what are some different logo designs that you've come up with? Right? So I used ChatGPT, you generated me a bunch of different stuff. And it was all based off of a common color set that I had, or a common theme that I had already put together. But, you know, as I thought about it, I'm like, you know, someone might say, well, look, you just took away a graphic, you know, designer's job here, you know, with regards to the stuff that came up with.

But the fact of the matter is, is actually swapping out colors. Like, well, I'm going to take a look at this logo and I'm going to try it with blue and white and try it with red and green and try it with. Like that to me isn't really what a graphic designer wants to do. I think they want to be on the front end of that.

They want to talk with somebody and conceptualize what is your logo trying to, you know, guide? What is it trying to come across at? And I still think that's a uniquely human thing. And I think that's still something that I would go and, and pay for someone to do and not funnel it all off to AI because I just, the results aren't there.

And I also think there's a lot of back and forth and a lot of human interaction as if I work with the designer to come up with a concept that I just don't think is going to be, suffice with just a human to AI interaction, or human to machine, I guess. I don't know what your thoughts are. Right.

[00:23:00] Andy Carroll: No, you're dead on.

And I think AI can encourage laziness. And when we are creating, whether it's art or client work or code, you know, we have to look at ourselves and say, okay, no one's going for perfection. You know, perfection is the enemy of productivity, you know, but when is good enough, both. And if you're lazy and you just want to use AI to create your, your logo.

You can get a very mediocre, good enough logo that will probably suit your needs. But if you want someone who will get to know you on a personal level, who understands the value of what you provide, who can incorporate your competitive advantage into a little picture, then only a designer can help do that.

So I like to think of AI as the over eager little intern. That has a wealth of information, but only if you ask it the right way. And sometimes you have to persist and keep asking and keep asking. And only once it really begins to know you, will you be able to get a good result on the first few tries.

But I think we're a long way from replicating that human creativity and human understanding we get from empathy and, you know, the emotional quotient and all of those unwritten things that humans kind of vibe with each other and understand each other. And that's really the nature of work, right?

[00:24:24] Justin Grammens: Yes, yes, it is.

I have a 12 year old. And so it's going to be interesting to see in the next, you know, once the older one graduates, even from high school, right? In the next eight years or six years or so, it's just going to be what the future of work's going to look like for them and college. And, you know, but again, I feel like there's just this whole human experience that AI can't really take away.

It can just, and hopefully in some ways, enhance. As you said, get a lot of the mundane tasks out of the way and be actually somebody that you can collaborate with rather than a threat. Right.

[00:24:53] Andy Carroll: Well, and one of the things I, I think about and I wonder about, because I have a younger child, he's turning five this month.

And you know, what does the future look like for them? And how do we balance the customization that these tools offer with us releasing our data to them? And where is the right spot? for sharing data and you see it come in swings, right? So I'm like borderline Gen X millennial where I'm one of the few that grew up without and with the internet.

And I'm always deeply skeptical of sharing data, you know, outside of what I can control. And there are younger generations that put everything out there. And now you see folks become a little more conservative and maybe they're being cautious with their data. And I know social media is the big use case here, but When it comes to sharing data, releasing your data in exchange for free tools, insights, efficiencies, all this cool stuff, you know, I mean, I don't know, what do you see and how, how are you positioning that as you're raising children?

[00:25:54] Justin Grammens: Yeah, I mean, if you're using the product, you know, and they have your information, you are the product, right? I'm fully aware that not only myself, but my kids and all the stuff that I do online. is, you know, being monitored in some ways, right? We're all carrying around a smartphone. You got to know that these companies know a heck of a lot more about you than maybe you think that they do.

And so, from my standpoint, I've always sort of said, well, as long as it's, there's value there, as long as I'm getting a better product, a better service, a better experience, you know, I'm fine with it. If people want to have to pay for it, maybe they spend 10 bucks a month and they don't get all of those features or they get enhanced features, that's fine too.

So I think there's a comfort level there around that. I think, you know, as you were talking and I was going to bring this up earlier, you mentioned about having values being aligned, right? You said you were in a job and the values weren't really aligned. And what got me sort of thinking about this and hopefully this sort of ties into what we're talking about here is, is You know, as these AIs are being trained, what values are they being learned, right?

Are they actually, if I'm giving it good things to do and the system's going to get better based on my behavior and my path, that's actually not so bad, right? I'm not really the one that's like worried that, oh my gosh, the healthcare provider is going to find out that I did X, Y, and Z and now my rates are going to go up by 10X, right?

I'm not so concerned about that. What I'm really more concerned about is, is actually people polluting the system, right? That AI tools are going to be making decisions. Based on malicious intent, right? And so, you know, to answer your question in sort of a roundabout way, I'm okay with companies getting my data.

I'm actually okay with my companies getting my kids data, as long as, you know, whatever they market back to them is age appropriate. And once you get to a certain point that they understand the internet and that their data, anything that's out there is there forever. Those are some lessons that I think the younger generation needs to understand and learn.

But I'm not so worried about that. Again, like I say, I'm really more worried about, you know, an outside actor, somebody from a different country actually gaming the system, like our political system, right? Or even somebody else gaming, you know, a healthcare system, right? People getting excluded because of.

Certain data and certain data sets, right? That, that's what has me a little bit more concerned.

[00:27:55] Andy Carroll: Yeah, I agree. I think, you know, election year misinformation aside, that's a whole separate concern, but it's the ability of everyone to feel comfortable that their data is being used, you know, in a non malicious manner.

And there will always be bad actors, you know, there are today, I think many organizations are criminally under investing in their cyber capabilities. And what's scary about all of this innovation is the pace at which it is moving. And it might be fair to say it'll never be slower than it is today. So governments and regulatory bodies are rarely going to be fast enough to capture all the different ways that this could be used.

Maliciously. And I think it's up to the public to educate ourselves and be aware of, you know, how our data is used, how entities could use it and to just be comfortable with it, you know, I mean, going full circle in terms of my own data, I'd feel less comfortable sharing it if I didn't already know that it was out there.

Right, so if I was releasing it for the first time, I might be hesitant, but now since it's already out there and everyone's already got it, then I guess I'm a little, a little less concerned. And I think the benefits are tremendous, particularly when it comes to health care.

[00:29:09] Justin Grammens: Yes, interesting. Yeah, so I teach a class on AI, machine learning and IOT at the University of St.

Thomas, graduate class. And what I like to talk to you about is, is, you know, you might be a little bit younger than me, but we do remember the first day we put our credit card into, The website or, you know, into the internet, like Amazon or whatever, you're like, do I really trust this thing? Like, is there that, is there that little, you know, that little lock up in the corner?

Everyone said I should look for the lock, you know, make sure that it's secure. Right. And now it's just like, you can order stuff via voice, you know, on your Alexa. Right. So, and now we're going through this, I think with biometric data, right? It's the same sort of. People are very, very concerned about biometric going somewhere.

And because, frankly, the stakes are a lot higher. I tell people, I tell my class, I'm like, you know, it sucks if somebody gets into your bank account or they ignore your credit card number. Obviously, there's stuff there, but there's ways to mitigate it. There's not really a whole lot of rules and regulations on ways to mitigate biometric and companies and how they can use that.

Now, I believe in time and over time we're seeing rules and regulations and the general society relaxing a little bit on that. And so I think it's just a matter of time. It's sort of a comfort level I think people need to understand.

[00:30:11] Andy Carroll: It is. And it's, it's when you start seeing the potential of those technologies when they work with other tools, that really gets scary.

I saw an article the other day about a law enforcement agency that used, there's a startup, some AI company that will take your genetic information. and create a visual profile of what your face looks like. So they had a cold case. They took the genetic information of one of the suspects, put it through that system, generated a visual description of the person and a face, and then they took that visual description and put it through face.

Recognition software and it led to a result and then they went and arrested the guy. It wasn't the guy, but you know, it's when you take things to that nth degree and, and no one probably ever thought of doing that before and combining those two tools, but lo and behold, then we're off in this whole new direction.

So there's, there's a lot of scary things that could happen, but I think, you know, having the transparency and the trust of the public to understand how their data can and will be used, that's

[00:31:17] Justin Grammens: what's important. Yeah, yeah, for sure. A hundred percent on that. Just everyone needs to be aware of how it's being used.

And then also, like you said, maybe the, just the, the possible upside of this. And after working in IoT for many, many years, it's not until you actually get the entire system together that people start actually realizing how cool the system can be. It's when you have a home security system that's also tied into your car, that's also tied into the weather and that it can actually, you know, decide that it's going to not water, you know, your lawn because the rain's already.

It's sort of just this entire ecosystem that needs to be built. But when people are looking at just one little area, they're like, well, why would I connect my refrigerator? Well, yeah, you probably wouldn't. But what if you had an elderly person who lived alone and the refrigerator hasn't been opened in three days, right?

Maybe you might want to know about that, right? So there's little unique use cases and a lot of the stuff it's there in isolation is tough to formulate it. But once you see it orchestrated together, that's really where the power with any new technology.

[00:32:10] Andy Carroll: Right. And one of the changes I would like to see, and this is a great healthcare use case, but extends well beyond it, is for everyone to have the ability to take back their data.

and to make deliberate decisions about when, why, and with whom they share it. So from a healthcare standpoint, instead of Fairview or Blue Cross or Epic having your electronic health record, you know, maybe it's on a wearable, it's on a bracelet. And every time you go to see a provider, you give them access to your data to use for your visit.

And when they recommend, you know, a solution, then you give your data to them because they're going to help you get healthier. And when it comes to personalized medicine and nutrition and drug research, you just, you do that. And then you get a personalized experience with your own health data, but it still remains yours and it should be yours.

And in the case of, like you mentioned, when you're the product, right, with Google or Facebook, you know, maybe there's a little revenue sharing there. For giving up your data, you do get a little piece of the marketing pie, you know, because you are giving yourself out as a product. Yeah. Yeah.

[00:33:16] Justin Grammens: I totally agree.

I think this whole centralized system of all the data on location definitely has flaws. to it. And this whole idea of distributed computing. We've been talking about it for quite some time, but yeah, kind of having your data on premise and opening it up as needed and then closing it back up again. I believe we'll get there.

It's not an easy problem to solve, but I think there is a path forward and, and I think at some point it's just going to have to go that way. Just for numerous reasons. The reason some of it is just going to basically be, you just don't want all the compute in the cloud all the time, you know, it's just, it just doesn't make sense.

And so there's going to be just cost savings and alternatives, people are going to want to run stuff more on prem. When it comes to getting more information and people learning stuff, like, you know, like what would you recommend? What, are there any books, any websites, any conferences? What are some things that you might recommend to people listening?

Well,

[00:34:05] Andy Carroll: what I advise people to do is to think of where they want to end up with their learning, and then they can back into the solution. So for maybe someone later in their career, if they just want to get sharp about AI and its capabilities, potential pitfalls, particularly from a governance and strategy standpoint.

Then there are a whole slew of online classes that are, you know, not very lengthy. There are some in person classes as well through your, you know, local university. I'm here, I'm sure St. Thomas does others. So that's a great place to start is just wherever it's interesting, go dig in, take a class. It's low commitment, but high impact.

And it'll get you very well armed for what's coming. You know, for someone mid career like me, I'd say this is a tremendous opportunity to add a skill set that will be very scarce. There are only so many programmers and data scientists and data engineers out there. And eventually. You know, they'll need others to understand the tools and to step in.

So, if you're in IT, I think it's a great opportunity to get into some of those cloud courses. You know, Microsoft, Google, and, uh, what's the other one of the big three? Amazon, oh yeah, that one. AWS, yeah. Yeah, we should talk about Amazon for a minute. So, they're all offering courses that can help get you a little sharper and get you certified on their system.

So, that certainly helps, you know. If your early career incorporate data into whatever you are doing, even if you're a creative and you're an artist, there's certainly an opportunity to get a little smarter with data and get a little sharper with statistics. It might seem daunting, but it will pay dividends because if you understand data, you can understand AI and there are very few.

Industrial revolutions that will cause this sort of disruption in terms of the workforce and the skillset required. So if you even have a remote interest in this, just get into data, take a course, and it'll lead you down a great path. I couldn't leave without plugging Applied AI. And the reason I love Applied AI as a conference and as a hub is that number one, it's turned into what I think is the epicenter of the AI conversation here in the Twin Cities.

And two, it pulls together such a diverse and curious group of people, whether it's students or executives or programmers or lawyers. It's just a lot of people that are curious and want to hear about what's going on. And it's such a fun room to be in because everyone just dorks out on AI. And they all come from different places.

And I think that's fun. So whether it's Applied AI or some other meetup or group that's talking about it, you can't go wrong. This is so prevalent right now, everyone is talking about AI. So if you have trouble finding a community that's talking about it, please let me know and I'll be glad to point you to one.

That's

[00:36:51] Justin Grammens: great. Thank you, Andy. I appreciate the plug for Plaid AI. And it's good to hear that. Yeah, you have the same sort of mindset that I do too. It's like, let's get a bunch of smart people together that are curious. And no matter your background, if you're a geek programmer, or you're somebody at the CEO level, or no matter what industry.

This is going to have an impact on you. So at least, you know, sit up and take attention and learn and then, and then share. Right. That's the other thing that I think that I'm really, really happy about is, you know, most people that come in and present, you're going to be presenting in a couple of months that presented our conference that are on this podcast are people that have just picked it up and learned it.

Right. I mean, it's just like, yeah, I'll get the occasional PhD here, this, that, or the other stuff. But I mean, most of the other people have just, they've dove in deep and they've learned stuff and then they've come back up and they said, I want to share this with the world. Right. And I want to speak. For 45 minutes at a meetup or a conference or want to be on the podcast, Justin, and do it.

So it's just that full circle, right? And this is the thing that's so exciting is, you know, we're just in the first inning of this, of this baseball. You know, this is going to last for many, many, many years. And that's what I'm so excited about is I don't see any technology out there today that just has the biggest potential.

to really have such an impact on, frankly, the human race than how artificial intelligence does. So

[00:38:01] Andy Carroll: I totally agree. And I think what's interesting about AI compared to other technologies, when you look at like the advent of the internet and the creation of the web browser, you know, with AI, you have the ability to dig into a category or area where you are passionate about it.

And, you know, even if you picked it up today, you could spend your next two months. On your free time doing nothing but playing with AI image generation. And after doing that for a couple of months, you could be an expert. And you could be the person that could speak knowledgably about prompt engineering and tools and pitfalls and governance and all these amazing things.

You know, we didn't get that with the web browser. You couldn't go browse the web for two months and say, I am an absolute expert at Firefox. And, you know, and that people would want to hear you talk about it. So there's so many areas that you could dig in and become a functional expert, you know, both on your personal end and especially at work.

That's why I say it's a great time to raise your hand and, and say, hey, I'm going to take on this AI thing and I'm going to get sharp about it and it's going to open up doors. Yeah, for sure.

[00:39:05] Justin Grammens: For sure. Well, I love your passion for it. Other people that are listening to the program, I'm sure will as well.

Andy, how do people get

[00:39:11] Andy Carroll: ahold of you? Uh, I think LinkedIn is the greatest way to do it. Just Andy Carroll, you'll find me. It's a big smiling bearded face on LinkedIn. And my consulting firm is Slancha. That's a Gaelic for health. It's S L A I N T E and it's slanchaai. com. And you can email me at Andy at slanchaai.

com. It's awesome.

[00:39:34] Justin Grammens: Yeah, we'll, we'll have liner notes and transcript from all of this stuff, which I'm working on a GPT right now that will take the transcripts from all of, you know, the 85 episodes or so that we've done and put them into a knowledge base that we'll query and ask that stuff. But yeah, people will be able to see all of this information in our notes as well.

I always ask people, yeah, is there something else you wanted to talk about? Is there some sort of key point or thing that maybe we didn't, we didn't discuss at all that you'd like to share?

[00:40:00] Andy Carroll: I think it's such a crazy time and it's fun. I'm, I'm a little twisted because I savor disruption and this is such a perfect storm for disruption.

And folks might be tempted to claim the sky is falling and that robots are going to take over and AI is going to take my job. And, you know, at the end of the day, whether it's a new technology, a new process, you know, new industry. It's always just made us better, and I, I might be preaching from a little AI optimist soapbox here, but the reason everyone is so passionate about this is because we see the potential, and for every one person that might be acting maliciously, there are hundreds who are creating tools for good, that are making healthcare better, that are making the customer experience better, that are disrupting industries in a good way, and I'm just excited to see where it all goes.

Thanks. That's

[00:40:53] Justin Grammens: beautiful. That's great. That's great, Andy. Yep. Amen to that and your preacher soapbox for that. I think I 100 percent agree with that. And I think there's probably a lot of people that do as well. So I think if we keep that attitude and that mindset as we continue to use this stuff, it will only, uh, you know, sort of self perpetuate to do good with this stuff.

So, well, cool, man. I appreciate the time today. And I know you're a busy person. We'll be keeping in touch here in the future days and months ahead. I know you're speaking at some other events, so I will direct people off to your LinkedIn page and follow you. Hopefully catch up and meet you in person at some point.

So thanks, Andy. I appreciate you being on the program today.

[00:41:27] AI Speaker: You've listened to another episode of the Conversations on Applied AI podcast. We hope you are eager to learn more about applying artificial intelligence and deep learning within your organization. You can visit us at AppliedAI. mn to keep up to date on our events and connect with our amazing community.

Please don't hesitate to reach out to Justin at AppliedAI. mn if you are interested in participating. for listening.