Lori Ryan - Empowering Us With the Foundational Knowledge of AI

The conversation this week is with Lori Ryan. Lori is the creator of Lorignite.com an artificial intelligence hub aimed at making AI more accessible. Additionally, with more than 20 years of experience in local TV news production, she's honed her skills in identifying and conveying what people care about in an easy to understand manner. And this skill set has allowed her to curate content on Laurie Ignite and help people navigate the rapidly evolving landscape of AI. Her extensive background in IT project management includes working with Fortune 100 companies across the globe, Implementing technology solutions and managing complex projects.

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

• Lorignite.com
• The Mel Robbins Podcast
• The 5 Second Rule by Mel Robbins
• AISALON
• Lab651
• AI Futures An Anthology by Kyle Shannon

Enjoy!

Your host,
Justin Grammens

Transcript

[00:00:00] Lori Ryan: The key thing is, I think there's still fear, right? But it's going to take your job. So how do you get folks, you know, within your organization to see that it's not intended to replace them, but help them. So essentially maybe what you were doing that was taking up 40 hours a week. Might not only take 32 hours, right?

And maybe it's taking on some of the things that you dreaded doing that to you felt like the busy work that weren't really leveraging, whether it's your creativity or strategic mind, you know, those types of things where you really add value to your organization. So it's, Figuring out how to get people on board and see how it can benefit them.

Maybe. So they're not working long hours at work, but maybe they get to go home at time and have dinner with their families.

[00:00:49] AI Voice: Welcome to the conversations on Applied AI podcast, where Justin Grammens 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.

[00:01:20] Justin Grammens: Welcome everyone to the Conversations on Applied AI podcast.

I'm your host, Justin Grammens, and today we're blessed to have Lori Ryan joining us. Lori is absolutely the perfect guest. In her bio, she says, ever since I was a little girl, I've dreamed of starting my own business. When I first experienced ChatGPT, I knew I wanted to share my journey of starting a business with ChatGPT as my sidekick.

This led me to create loriignite. com as a passion project and a side hustle, an artificial intelligence hub aimed at making AI more accessible. So I know we'll be talking a lot about Loriignite and the work that she's doing here. Additionally, with more than 20 years of experience in local TV news production, she's honed her skills in identifying and conveying what people care about in an easy to understand manner.

And this skill set has allowed her to curate content on Laurie Ignite and help people navigate the rapidly evolving landscape of AI. Her extensive background in IT project management includes working with Fortune 100 companies across the globe, Implementing technology solutions and managing complex projects.

So thank you, Lori, so much for being on the program today.

[00:02:17] Lori Ryan: It is absolutely my pleasure. Thank you so much for having me.

[00:02:20] Justin Grammens: Awesome. Well, good. So I talked a little bit about Lori Ignite and like I say, we will definitely talk about stuff that you're doing there. Cause I know you're out in the community, you're educating people on all that sort of stuff, but how did you sort of get to where you are today?

Has this kind of been like a always interested in technology, kind of got into the corporate space and worked into that, maybe give a little bit of context around how you landed where you are today.

[00:02:41] Lori Ryan: It is so random. It's crazy. So I, um, was going through some things in my last job last March. At the same time, I'm looking at my Instagram profile and bio.

I'm like, holy kittens, that has not been updated since 2011. And it said something like working mom. Meanwhile, I have. No kids in the house. And I'd heard of work about people using chat GPT. One of the main developers had used it and then a manager in our tech department's like, Hey, I'm going to try and use it for an internal communication.

So I thought, all right, I'm going to give this puppy a whirl. I'm going to see what happens if I ask it for help with my Instagram bio. And I was so blown away by the results. Now, one or two days later, I listened to the Mel Robbins podcast. Have you ever heard the Mel Robbins podcast? You know,

[00:03:28] Justin Grammens: I definitely have.

And she is phenomenal. And, uh, yeah, so inspirational, really, really big on kind of getting people to take action really and on things that they are passionate about. And, uh, I love her book, The Five Second Rule. I'm not sure if you've read or listened to that book. Um, yeah, she is phenomenal. I think there was a certain podcast episode that maybe got you fired up.

[00:03:49] Lori Ryan: There was, I don't know if it was like burned out, ready for something new, something along those effects. And I happened to listen to that a couple of days after I tried chat GPT and she was interviewing Andy Porterfield. And she said, you know, To start your own business or to figure out what you wanna do next, you just need to be 10% better than everybody else.

And I sat there feminine and I'm like, boop, boop. Light bulbs went off. And I'm like, you know what? I'm a decent storyteller. Having that background change management, a lot of that with tech training, implementing new technology for people. And I'm like, I'm gonna dive into this, dive in. That sounds very chatty, pt, but I'm gonna get in there, get my hands dirty and learn.

As much as I can, and that's what I did. It took some online courses. What seems natural and easy for me, I guess what I'm learning is others do need some help getting started.

[00:04:38] Justin Grammens: Why do you think that is? Do you think it's a fear?

[00:04:40] Lori Ryan: Yes. And I think the best place to be right now is curious. And the phrase you can ask yourself in almost any situation now is, I wonder, I wonder what would happen or about any situation, how would it answer it?

How does it do? So it's not going in with the idea of, Oh, CHAT GPT is going to get it right. And it's going to be amazing. And it's Can it do all these great things? It's just going in there with curiosity and then being pleasantly surprised when it's like, Oh, wow, that was helpful. Or, Oh, that did help me figure out how to approach that situation.

[00:05:14] Justin Grammens: Oh yeah. I feel like, you know, you and I have maybe seen the light in some ways. I think a lot of people think of Chat GPT as just another Google, right? I think I see people using it as Google. And. It's really so much more than that. In fact, it's really not a good Google. It's, it's really not a good source for, in some ways, factual information that is 100 percent factual, right?

The whole generative AI, the reason that we're talking about this is it actually generates text. But I'm like you, I'm like, huh, okay, I need to craft an email to somebody. Give me some ideas on maybe ways that I could phrase this a certain ways. It's really more like a companion, right?

[00:05:48] Lori Ryan: Correct. I look at it as an assistant.

And I'm the kind of person when I'm in an office. I don't know why, but I'm always turning around wanting to talk to somebody, get their advice, get their opinion, hear what they're thinking. And now I have this tool that can be there with me 24 seven. Right. It doesn't take a break. Always available. If I wake up in the middle of the night and I'm like, ah, I wonder about how I should handle this situation or what should I be thinking about?

What am I missing? And I think so many people get into the frame of mind that it just works. Writes things for you and is a cheating tool and it really can help open your mind if you're asking for what am I missing or what am I not thinking of? And it can be really helpful too. I don't know if you've ever gotten an email and your first reaction is.

Uh, frustration or, you know, like, what were they thinking when I sent this? And sometimes you can put the contents of that email, assuming there's no proprietary data or personal identifying information in there, you can put in a chat GPT and you can say, Hey, here's how this is making me feel. What helped me.

Maybe figure out a better way to see this and the best way to reply to it. And not in terms of just the, how to write the text, but just open your mind to sitting in the other person's shoes and what they could be thinking about that you may not have taken into consideration.

[00:07:12] Justin Grammens: Yeah, yeah, this, this whole idea where you can say, you know, act as this type of person and then it will say, okay, I'm that type of person and you can start asking questions.

So I'm, I'm actually doing a presentation. It'll be maybe three weeks or so from now, but I'm actually going to be walking through designing basically a healthcare device, fictitious device, but this is something that would take data off of like an ECG off your heart, go through a mobile app and then put it up into the cloud.

Right. That'd be great. Something that any software engineer would kind of know how to do, but I'm basically saying, you know, act as a physician. What are the types of information that you would want to be able to see potentially, because I'm not a physician, right? And potentially a lot of these people that are building products for other people, they maybe have one or two positions maybe that are there as sort of subject matter experts, but now you can actually have more than that.

You can tell the system to basically act as the end user and then you can query it. What things, what information would you like to know about? And. Number one is gearing it for the customer. I'm also going to talk about through the entire life cycle of product development. So I am a software engineer.

What are the types of things that I need to know in order for me to do my job? Right? So as the product manager, I need to Fill out a bunch of stuff. I mean, I need to explain the system. So it just, it really helps everybody understand what the other person is doing in the entire building out of any system really, right?

So I think it's super powerful. People don't really realize you can sort of get the large language model to start to act like. Another thing, act like the end goal.

[00:08:41] Lori Ryan: And you just touched on something else that is really key. You can actually ask the tool, what do you need for me? Or what would help you, you know, along this process?

So essentially it can guide you. Into delivering, you know, the details or context or whatever it is for it to give you the best output.

[00:09:00] Justin Grammens: Yeah, yeah, for sure. So these are things that I think we're all just sort of learning about over the course of time and things continue to change. I mean, every, every week there's some sort of new development going on and everything.

So it's been awesome to see it sort of mature over the past 16 months. Now. Before we started recording this, you were saying you've been talking to people and you're, you're kind of, like me, surprised at the number of people that haven't, they've kind of heard of it. They're like, oh, I, I heard about Chad GPT, but I haven't really tried it, right?

What are you teaching people? What are you showing them that's kind of making their eyes light up?

[00:09:30] Lori Ryan: I think there's been, I should say, these generative AI tools have gotten a bad rap, right? It's hallucinating, it's going to end the world, there's so many dangerous things associated with it, whether it's political propaganda or deepfakes, and all of that is understandable, right?

My message is knowledge is power. And the goal is get your hands dirty, get in there, because that's where you're really going to understand how it works, what it does well, what it doesn't do well, and I try to make it fun, right? I'm like, you don't have to go in there and expect it to immediately solve all of your problems.

Examples I use, I'm going to be on a golf team. Coming up. So I'm like, wouldn't it be fun to come up with an image for our team, the party girls. And I went in and first off the image it created at five women. I'm like, wait, when's the last time you had a five summit in golf? Not to mention these five women were all white and their bodies were perfect.

I mean, there were, was not a variety of shapes and sizes. And so we kind of went through that exercise and I'm like, Now I'm asking it for four golfers and it still came out with five. Finally, we got to four, a little more racially diverse. And I'm like, you know what, maybe I should just ask it for drinks or glasses and it came out with five glasses.

So there are those types of things. And then I was with a friend who developed the American sign language program at the Minneapolis public schools. So we're like, let's see how it does with American sign language. So we said, You know, show us images of what nice to meet you looks like. And it was wrong.

I guess it started with British Sign Language and

[00:11:10] Justin Grammens: Did it even have five fingers? I mean, that's the other problem too.

[00:11:12] Lori Ryan: It's

[00:11:13] Justin Grammens: not really even good at drawing hands. I

[00:11:15] Lori Ryan: think I have pictures so bad with like many weird hands.

[00:11:20] Justin Grammens: Yes.

[00:11:20] Lori Ryan: But we couldn't get that correct either. So these are just things you notice going through it.

And my firm belief is the more people who are in there playing, the more feedback that they can give, the more things they can notice that maybe could use a little input

[00:11:37] Justin Grammens: from

[00:11:37] Lori Ryan: varying viewpoints.

[00:11:38] Justin Grammens: Yeah, yeah, yeah, yeah. It's interesting. You touched on a couple of different things there. Number one is, is obviously the diversity of the models.

It's, you know, white men or whatever. A lot of that stuff is in there. There's a lot of assumptions that are built into that. And then there's definitely some, some issues there. And then you were also touching on just sort of the. I guess the capabilities of it. So I've thrown it a chess board with a number of different chess pieces.

I'm halfway through a chess game and I say, tell me the next move to move. Right. And it's, it's wrong. It's telling me to move pieces in areas that actually don't really exist. So then I say, no, no, no, that is, that's an incorrect move. And then it says, oh, I'm sorry. Okay. No, this is the right move. Right. And you start going down these, these loops sort of back and forth where it's like, okay, at some point, it's not going to, it doesn't really understand it yet.

However, right, they are building better models to do this. And there actually is a GPT to handle chess specifically. Still doesn't do so good with images, but there's chess notation that you can say, here's the way that the board's laid out. Here's where the pieces are. Tell me the next one. And it's very, very good at that.

So it's like, you know, once it understands, I guess what, what it is, but getting it into that context isn't really so good. So. But you got to take the good with the bad. And the other thing that I think that I tell people is, you know, just imagine how far, just how far it's gone in the past 16 months. And it's not linear, you know, this is exponential, right?

So in the next 16 months, what we're going to start seeing is stuff that's just completely Even better than what we ever thought. And you see that with something like text to video, you know, if you've ever seen the stuff with Sora, right? People were saying, yeah, people were saying that that was five years out, right?

So basically last year people were saying, no way, there's no way that this is going to happen within the next five years. And now it's happened. And so it's just, yeah, the timelines are just completely shrinking. And all of this,

[00:13:15] Lori Ryan: I completely agree. And just the whole GPT store, right. And this ability to build your own GPT specific to different tasks.

And they also, you know, now have the teams, you know, the open AI has, and they continue to make updates to that. Right. So we're using that where I'm embedded at a retail store. And today I decided I'm going to make a GPT because we have to create a lot of training documentation, right. So we can take our meetings and take the transcripts from the meetings.

Pop it into ChatGPT now that's designed for our guides, whether it's the summary at the top, the purpose, the steps include the time codes, you know, from the video and it's pretty slick and just share it within the organization.

[00:13:57] Justin Grammens: So yeah, so more of a, more of an intranet back office application in that particular case?

[00:14:02] Lori Ryan: Mm hmm.

[00:14:03] Justin Grammens: Okay. And you guys are comfortable putting your information into ChatGPT? Yeah.

[00:14:08] Lori Ryan: In the Teams.

[00:14:09] Justin Grammens: Yeah.

[00:14:09] Lori Ryan: Yes.

[00:14:10] Justin Grammens: Yep. Yeah. Yeah. For sure.

[00:14:11] Lori Ryan: In the Teams account.

[00:14:13] Justin Grammens: Yeah. That's the other thing I think people are really worried about. I mean, obviously, if you just go on and you don't have an account, you're just using it just as an end user, not paying anything.

Yeah. That's one thing. But if you look at their terms of service, you know, and basically what the legalese says is they are saying they're not using your stuff to train. They're saying that they delete everything after 30 days, everything's encrypted. In some ways, you know, I see two sides of it. I see, I see the side of an organization that maybe has highly confidential proprietary stuff that they still don't trust that.

But then I'm actually working on a blog post right now where this reminds me a lot of open source software. So when I was working at Thomson Reuters, We were writing and using a lot of open source software. And then this sort of edict came down that said, thou shall not use open source software. And the, the question in me as an engineer at the time was like, well, no, but you know, well, why?

Right. And a lot of it was just concerns over what's going to happen, right? If there was a piece of software that you wrote and all of a sudden it had to be contributed back to the open source community and all that type of stuff. But what I realized the biggest defense that I had for that, I said, okay, if we're not using any open source software, well now, This thing for me to develop is going to be about 10 times long, right?

It's going to take me about six months that would normally take me about six weeks. And expensive, very expensive. And so I think that's going to happen with a lot of these organizations. They're going to say, just like moving to the cloud, right? It's like, why do you have your own data center anymore? It just makes so much more sense in most cases, not all cases, but I think in a lot of these cases.

So I think a lot of people are sort of saying, I'm not going to use any sort of SaaS based, you know, open source model or sorry, uh, you know, A large language model, because I don't want them to have my data. But at the end of the day, when you have to maintain all your own stuff in house, I mean, it's just, it's just going to kill you, I think, over an extended period of time.

[00:15:52] Lori Ryan: I do think there's some things that can be off limits. You know, and I actually just went to a presentation at a marketing company where they walked through how they implement new AI tools into their organization. And they talked about how certain things within the company are off limits for AI. But they encourage all their staff to try new tools and then figure out they had certain criteria, right?

Why do you want to use it? What do you hope to get out of it? And then they have to measure it on the other side after a trial. So I thought that was an interesting way to approach bringing generative AI tools into your organization.

[00:16:26] Justin Grammens: Yeah. As you're talking to organizations, do you see any. I don't know, that's the lack of a better term here.

Like any sort of guides or rules, it feels like it's the wild, wild West. You're like, most companies aren't spending the time kind of saying what is on limits, what's off limits people are. And so I actually think that kind of feeds this sort of fear in a little ways, because it's like, I don't know what I can do and what I can't do.

And I've been thinking more organizations should probably kind of come up with more of, it's this general term called AI governance, right? I've heard people talk about, you need to go into an organization and create AI governance. Are you, are you seeing any of that? What, like, what do you, what do you, what do you think about that?

[00:17:01] Lori Ryan: This one company set up an AI committee with representation from all departments. And they meet, I think, bi weekly. And I think, They're working on ways to bring things in and help write and put together the governance. So it's really a team effort. It's not anything that's coming right from the top.

Let's create this together based on what we're learning as we go through this process.

[00:17:25] Justin Grammens: Gotcha. Yeah, that's, that's very mature of them. Most companies aren't, aren't doing that. They're just kind of like throwing a bunch of things together, but, and then that's good. Is that an opportunity then for Lori Ignite yourself to come in and talk to organizations about that?

Is that a service you offer?

[00:17:38] Lori Ryan: Absolutely. I would say the, the biggest area right now we're seeing uptick is how do I get started? You know, What are things we should be thinking about? How do we just log in? What kind of accounts should we be looking at? What are the tools? My primary focus is chat GPT, but I have to say, I've been loving the heck out of perplexity for research.

And just last week we asked like chat GPT, like, who is Lori Ryan? They gave like two totally wrong answers. Then we're like, who is Lori Ryan from Laurignite? Or who is Lori Druskin Ryan? It still couldn't get it right. But perplexity, there was a gal. Who was attending and myself, I had the pro version and she had the free version and Perplexity got it right on both versions.

The only thing that was goofy was the images that it gave, but it was really impressive, the results and everybody was blown away by how they shared where they got the information from and how easy it was to see the source.

[00:18:34] Justin Grammens: Huge. Yeah, no, I totally agree. I totally agree. No, I think perplexity is great.

I've been using it a lot more too. I can see why companies like Google are worried that again, it's a long tail. It's still, it's still quite some ways away, but you know, that they're going to lose ad revenue because people are going to start using perplexity as maybe as a starting point. Because it sort of melds the two together.

Oh, the other thing I think is cool about Perplexity is, is after I ask the question, then it actually has some really good follow ups, right? I can sit there and go down this rabbit hole of like, well, yeah, that's a good question to ask. And there's just a, you can just plus, plus, plus your way down. And that, that to me was, uh, something I'd never really thought about.

Like just suggested alternatives. And of course, Google has had those, of course, like below. Some people also search for. But theirs just seems so much more better, right?

[00:19:18] Lori Ryan: Yeah, their user experience is top notch. That whole rabbit hole thing is so fun. Cause it's like, Oh, that follow up question. Yeah, let's go there.

[00:19:27] Justin Grammens: Yeah, exactly. Yeah. Cool. I'm glad you brought up Perplexity. I actually had it here in the notes for us to, but Google now, obviously on that side, they're coming at you with Gemini, right? And so I found that to be more of my default too, as well. I hop over and sort of use Perplexity, Gemini and, and ChatGPT.

[00:19:44] Lori Ryan: So here I did a little bit with Gemini. I got a little soured, so I used to love to use these tools to compare privacy policies and they used to be really good and they would talk about how I had to compare, you know, OpenAI's privacy policy to Google's privacy and somehow it talked about Google and their search and, you know, it felt less, I don't know, wholesome to me.

I can't explain it. So I don't, I'm not using it. As much, I guess, but somebody told me the new, what is it? The 1. 5 with the million token window is off the charts and they did a quick demo of it and how well it did. With the huge token window.

[00:20:24] Justin Grammens: Yeah.

[00:20:24] Lori Ryan: So it may be worth revisiting.

[00:20:26] Justin Grammens: Yeah. For

[00:20:27] Lori Ryan: me. How about you?

[00:20:28] Justin Grammens: Like I say, I, I'm just kind of bounce around. I, and again, it probably depends on the task, you know? So, you know, Gemini and this, I was just like looking it up as we were still here. Like you can upload an image. That's all you can really upload. But like ChatGPT, you can upload a PDF, you can upload a spreadsheet.

There's Word documents. I, so I'm, I'm surprised that Gemini, they have all this multimodal functionality, but yet on their interface. Yeah. If you go to gemini. google. com, it only takes images. So again, depending on the task, right? So I've been, I've been kind of pushing the limits of chat GPT in a lot of different ways where I will actually push up a spreadsheet.

In fact, I worked with a company and they were just sort of exploring a couple of different areas of how they could improve their bidding process. Right. And back, and this is what's super interesting to me is like, you know, a year or two ago, You'd have to have somebody with a data science degree that would basically look at a spreadsheet and query a database and come up with some sort of machine learning algorithm, some sort of regression thing to figure out, okay, you know, here, here's, here's where I should start bidding out on these project, these prices.

And, uh, a guy came to me, this was like probably six months ago, and we started like playing around with it. He's like, yeah, you know, I need to figure out how to do this stuff. I'm like, just try chat GPT, right? And so he had all this data around all their prior projects that they've quoted. And then he just basically said, I'm quoting this one next.

What do you think it should be? And it was really good. Right. I mean, it was actually able to sort of do what. Uh, somebody would, in my case, would charge you thousands of dollars for, right? That I would build an application to kind of do a lot of this data analytics you could actually do in chat GPT. So it's fun.

It's one of those things where I just like, well, let me just start there and see what happens and, and sort of go from there. And like, when it comes to images, right, you've got, you've got Dolly, you can do stuff, but like mid journey is also really good. Like, yeah, everyone's just trying to one up each other.

It's a, it's sort of a,

[00:22:09] Lori Ryan: but you know, what's cool though, just by having dialogues, you With people, you learn about use cases you may not have thought of, you know, last week, a woman told me how her son had this not so great flight experience. And he used chat GPT to help write a letter to the airline. And he ended up with a ton of frequent flyer miles because of that, you know, then there was a gal who talked about, she has to do business proposals, right?

It's like 20 questions. It's a lengthy process. It can take her a week. She leveraged ChatGPT for the draft piece of it, right? She still had her human piece, overlooking everything, reviewing it, tweaking it. And a colleague told her it was one of the best proposals they've ever seen.

[00:22:54] Justin Grammens: Yeah, yeah. And that's one of those things where it's like, you know, if you're not using this tool, you're either wasting a lot of time, cause you, you probably will end up getting to that best proposal, but it's going to take you many, many hours and many, many iterations.

So why not use this tool that's already there? And the other, you know, your competitors, that's the thing, as you're probably talking to businesses, you're like, look, your competitors are using this. You're going to have to use this to at least somewhat keep up with them. Right. Yeah.

[00:23:18] Lori Ryan: And I just, uh, too, the key thing is, I think there's still fear, right?

Right. Right. But it's going to take your job. So how do you get folks, you know, within your organization to see that it's not intended to replace them, but help them. So essentially maybe what you were doing that was taking up 40 hours a week might now only take 32 hours.

[00:23:40] Justin Grammens: Right. Right.

[00:23:41] Lori Ryan: And maybe it's taking on some of the things that you dreaded doing that to you felt like the busy work that weren't really leveraging, whether it's your creativity or strategic mind.

You know, those types of things where you really add value to your organization. So it's figuring out how to get people on board and see how it can benefit them. Maybe. So they're not working long hours at work, but maybe they get to go home in time and have dinner with their families.

[00:24:08] Justin Grammens: Yeah, yeah, exactly.

And each use case that I run into, I think people, you know, they, they don't want to become the horse when the automobile came out. Right. I think that's the biggest fear is people, it's basically fear of being irrelevant. And that's, that's the thing that I think is getting people is this idea that, wow, here's a technology that's going to make me irrelevant.

And the fact of the matter is, is no, this is actually technology that should actually empower you to be able to get a lot of these other things out of the way, because someone still, I still believe to this day, somebody still needs to be the master controller. As I talked about in a lot of these examples, you know, it writes code.

In fact, it writes really good code. People can poo poo it all day long, but believe me, I've been writing code for 20 years and I'm really impressed at what it does. You still need somebody with an engineering mindset to be able to pull all this stuff together the same way that, you know, as I was using open source libraries of great open source code out there, you still needed to be able to apply it in the right areas at the right time for the right solution.

So, I don't know, they just, they jump too much on it and they also, you know, think that it's going to do more than maybe it, it even possibly can today. But the fact of the matter is you still need a controller to use it.

[00:25:13] Lori Ryan: I met with a student who does coding, right? And he's like, yes, I'm using chat GPG to help me code.

But what stood out for me is that he reviews it, right? Now he's like the boss.

[00:25:25] Justin Grammens: And

[00:25:25] Lori Ryan: so when you're using these tools, essentially, you're knowing what to ask, when to ask it, right? And what and when to challenge. So in my mind, we're elevating ourselves.

[00:25:37] Justin Grammens: Yeah, yeah. What's cool is you can actually take someone else's code and say, explain to me what happens in this.

Right? So if you're picking up a project that somebody else wrote, isn't that amazing that you can basically have a co pilot? Hence why, you know, GitHub called it co pilot, but you can have a co pilot with you to help explain what's going on. Yeah. It's, it's a net huge, hugely a net positive.

[00:25:57] Lori Ryan: And, you know, the other thing we did here at the retail space was we took our onboarding guide and we.

Uploaded it to see how it might help our stylists or how it did with answers. And our area director went in and was asking questions. He's like, Oh my gosh, I can't believe how great these answers are. And then I not only asked questions based on what was in our fine tube guide, but I thought I wanted to check and see.

How it would answer questions that weren't in the guide. And he's like, wow, it was so great. It gave me ideas for what we should include, you know, as a follow up guide, you know, or additional training. So I thought that was exciting, right? And having a stylist, you know, could have this train and all of a sudden we can scale the training.

If everybody could have this in their pockets.

[00:26:40] Justin Grammens: Yeah, for sure. For sure. Well, if I was just getting into this, you, you touched on it a little bit, I guess, exploration and try it out. But I mean, maybe if I was even very early in my career, for example, maybe I'm just graduating from, from college or what, what have you, where, where do you suggest people go?

Maybe who should they connect with? Uh, what should they do?

[00:26:58] Lori Ryan: Well, they should definitely become part of the Applied AI community. It's unbelievable, right? The meetups that you offer, the conferences that you offer, the people coming together who are so supportive on a variety of topics, anyone in Minnesota in this area should definitely become involved in that.

Nationally, there's something called the AI Salon, which has all these different guilds and areas of expertise. There's like education and business and. They have what has been all free gatherings. And then there's a Tik TOK live at most nights at 9 PM with a guy named Kyle Shannon, and there's another woman there and Murphy, who's all about empowered fundraising because there are a lot of opportunities, right?

For nonprofits to benefit. They don't have a lot of money, right? Nonprofits typically. And so here are these tools, whether it's for grant help with grant writing or help with. How to approach getting fundraising or all of these different tasks. That's another huge opportunity area with these generative AI tools.

And the other thing is that there are a lot of online courses, whether it's. LinkedIn, I think Google's offering some free courses and generative AI Microsoft. They've done some cool demonstration videos. And then if you want more than just the generative AI tools, you know, you can take classes like I did, uh, you know, Product management for machine learning, where you can learn about all the things that can go wrong with AI projects and all the things you need to be thinking about when you approach these projects and making sure you're aligned with companies like Lab 651, who will talk you through all of those important things.

So, yeah.

[00:28:41] Justin Grammens: That's awesome. That's great. And we will be sure to put all these links in the liner notes along with the link off to your, to your website, right? You want to talk how people can maybe reach out to you?

[00:28:50] Lori Ryan: Absolutely. The website is laurignite. com. So it's like my name, Laurie, L O R I. And then that I serves as the end of Laurie.

And the beginning of Ignite, because the goal was to inspire and ignite folks to maybe explore and have fun with these tools that have become available since November of 2022. They're all in their infancy and it'll be fun to see how they grow.

[00:29:15] Justin Grammens: Yeah. Yeah. And so, like you said, it's your focus is on generative AI, right?

Yes. Yes. How do you explain generative AI to people versus regular AI? Do some people get a little confused?

[00:29:26] Lori Ryan: You know, like you say, AI is so broad. It's so many different things. So I just say, My focus is the piece that talks about generating something new. Hence the word generative, right? It's going to be new text.

It's going to be a new image. It's now audio or video, and it's pre trained, right? And all this information that it's collected over all of this period of time from the internet, everything it's read and done, and then the transformer piece, that's where people like you come in handy, Justin, to explain it to me.

It's like all the plumbing that's able to take information. from here, figure out the number vector stuff, what it aligns with back in that corner, pulling out the key information from that corner of the room and bringing it back to the context that was asked and spitting it out the other side. Right.

That's where I need your help.

[00:30:15] Justin Grammens: Yeah. Yeah. Well, a lot of us are just riding on top of, you know, the work that a lot of data scientists have done and basically deep learning and machine learning really is sort of, is what became sort of the breakthrough, right? Right. And that. The transformer architecture made it be allow us to do the stuff even faster, right?

I mean, this, all this stuff could have been done years and years ago before the transformer architecture. The problem was it's just super slow, right? So that architecture has been able to speed it up. But you know, are you reading any, any books on this subject? Is there any, any ones interesting that you, that you like to take a look at?

[00:30:46] Lori Ryan: You know, the folks from the AI Salon wrote the AI Futures book. And what I like to think is it's a time capsule. Because whatever was written in 2020, early 2024 or late 2023, I can't wait to see how much it changes in the next year. So, I don't know. I just think now is a good time to save. You know, images that were generated or text that was generated because we're going to look back and say, remember when, I don't know, kind of crazy.

[00:31:17] Justin Grammens: I was thinking about a couple of different things regarding a time capsule. I mean, first, you know, the video games now I've just been, are just insane in my mind. And I talked to my, I have 10 year old and a 12 year old and, you know, I, Tell them, and we even take a look at old Atari games, right? Just like basically space invaders and stuff.

And I'm like, can you imagine from space invaders to, you know, the vision pro now, right? People are wearing these things and it's just so realistic. Can you imagine where we're going to be in another 30 years from now? But. The time capsule thing is interesting, you know, as well. So I'm sure there's a lot of startups that are doing this.

I haven't really dug into it, but I did kind of have this epiphany around training a GPT. Like imagine if I could talk to my great, great, great, great grandfather, right? Like basically, you know, before I pass away, for example, like, could I just load all this information that I have, all the diary journals entries that I've, because I journal a lot, like, could I put all of that stuff into a thing?

That, you know, 300 years from now or whatever, you know, some future generation of mine could actually ask me questions about me. Right. So that's fascinating, isn't it?

[00:32:19] Lori Ryan: And I thought about the same thing. We're taking a family trip this summer with my parents and my kids and my son's like, I want to shoot a documentary.

Well, he's there with her and his cousins and I'm thinking, wow, with this video, with this data, with this transcripts, imagine the kids and the kids being able to talk with my parents someday, right? And have, I don't know if it's audio conversations or just texts. I don't know. It just, the whole thing got me really excited.

[00:32:44] Justin Grammens: Yeah. Yeah.

[00:32:45] Lori Ryan: Or for some people that might be creepy. I don't know.

[00:32:48] Justin Grammens: Well, yeah, the video part might be a little interesting, who knows? But even just, even just typing, you know, I just think it would be, If somebody was, and I'm sure people are, if they're, they're very, very good about, you know, just writing journal entries and putting their thoughts down.

You can do it with the tools today, right? I mean, there's nothing to say you couldn't create a GPT today that you could just load all this information into. So the tools are there. The tools are literally there to do it at a simple level, as a space invaders level, I guess, is what I would say. And then 30 years from now, like I say, it's just going to be amazing.

Completely crazy what we'll do.

[00:33:21] Lori Ryan: And yeah, most GPTs are a way to like almost build a proof of concept. Without a lot of cost, it can at least get you, Oh, okay, this is kind of how it would turn out. Here's where it does well.

[00:33:31] Justin Grammens: Yeah. Yeah. So Emily McCarthy here from Applied AI, she attended one of your sessions that you did a couple of weeks ago.

She said she had a blast. She said it was great. Everyone was learning from each other. You know, everyone was sort of like cross collaborating on stuff. I mean, that, that's one of the things that, that you guys did, right? Was kind of create your own GPT on whatever you wanted.

[00:33:49] Lori Ryan: We sure did. That was exactly it.

And everybody had different use cases. And I think Emily's interest is the humanities and, you know, figuring out how all of that ties into what's happening today. We had another woman who was like, Hey, wouldn't it be great to have a tool that would help write my link or my Airbnb reviews. So she was able to do something that would save her time and make it easier to leave Airbnb reviews.

So the use cases were across the limit. Or, wide. They were wide.

[00:34:18] Justin Grammens: Good. Yeah. No, that's what's so fun about this technology is this, it can be used literally anywhere and that's what I'm so excited about. I can't think of a use case where it's not going to have an impact on us in the future. So, a lot of fun.

Like you said, we're in the first early stages. Well, Lori, I appreciate the time. Was there anything else you wanted to? Make sure you, that we touched on or do we pretty much hit, hit it all today?

[00:34:40] Lori Ryan: I just hope that people try things out, try logging in. And then if they have questions, reach out because I just get so much joy seeing people's faces when they've logged in and asked a question or seen things in action for the first time.

Cause I do, I think it helps take down that fear and answer too. When they see there is good that can come out of it. It's not all evil.

[00:35:03] Justin Grammens: Yeah, very good. Very good. Well said. Well, great. Great. I appreciate the time today, Lori, and we'll definitely keep in touch and you'll, you'll be speaking at the Applied AI Conference coming up here in May.

So we look forward to having you share more of your knowledge with our community. Thank you.

[00:35:17] AI Voice: Very much looking forward to it. Thank you so much. 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 in a future episode. Thank you for listening.