Transcript: PODCAST INTERVIEW: Stephen Baird of TrackFly Returns

S5, Ep 140: Stephen Baird of TrackFly Returns

Stephen Baird from TrackFly joins The Articulate Fly to discuss AI and data privacy. They emphasize the human touch in podcasting and highlight the latest AI capabilities of TrackFly. Stay informed through their website, social media, and email updates.

2023, Marvin S. Cash
The Articulate Fly
http://www.thearticulatefly.com

In this episode of The Articulate Fly, I had the pleasure of welcoming back Stephen Baird from TrackFly to delve into the fascinating topics of artificial intelligence (AI) and privacy. We kick things off by going over the latest news and updates from TrackFly. As always, I take a moment to urge our listeners to share the podcast with their friends and leave a review. I express my deep appreciation to the incredible Articulate Fly community and give a special shout-out to our new series, "On the Salt with GotOne," which is sponsored by Norvise and co-hosted by the talented Captain David Blinken. Moving on to the main topic, Stephen and I begin by breaking down the different components of AI for our audience. We explore concepts such as machine learning, deep learning, and generative AI, illustrating how they are applied across various industries. Our aim is to demystify AI and present it in an accessible manner, so listeners can better understand its functionalities. Transitioning to the crucial issue of data privacy, we highlight the distinction between data processors and data controllers. It becomes clear that awareness of where personal data is being shared and how it is used is of paramount importance. I bring up the well-known saying "if you're not paying for a product, you are the product," underscoring the pervasive monetization of user data by free or low-cost applications. We also touch upon the complexities surrounding privacy policies and emphasize the necessity for users to be attentive to their own comfort levels when using certain apps. At this point, our conversation veers towards the significance of privacy by design and privacy by agreement. Stephen sheds light on the technical limitations and parameters that ensure applications operate in accordance with privacy laws. We delve into the ways in which companies capitalize on user data and explore the role of data brokers in collecting and analyzing data from multiple sources. It becomes apparent that consumers need to be well-informed about the value of their data in order to make informed decisions. Concluding our discussion, we address the limitations of AI in podcasting and stress the irreplaceable value of the human touch in the creation of podcasts. We touch upon the potential benefits and risks associated with an over-reliance on AI. Stephen takes the opportunity to share updates on TrackFly's AI capabilities and exciting upcoming features, while I bring attention to an app that aids in AI-based receipt tracking. Lastly, we express our gratitude for this insightful conversation and urge listeners to stay informed through TrackFly's website, social media channels, and email updates.

Generated Shownotes

Chapters

0:00:00 Introduction
0:03:01 Discussing Artificial Intelligence (AI) and Privacy
0:06:09 Breakdown of Artificial Intelligence Components
0:11:56 Introduction to Generative AI and Its Impact
0:14:29 Difference between Data Processors and Data Controllers
0:20:02 Setting Up for Success with Privacy by Design
0:23:36 Subtle Ways Businesses Monetize Data
0:28:37 Privacy Policies: Complex and Difficult to Interpret
0:33:44 Taking Control of Your Data: Consumer Reports' Permission Slip
0:38:19 Privacy Concerns in AI and Generative AI
0:40:48 Using Generative AI as a Tool for Creativity and Innovation
0:42:03 Frustrations with AI in Podcasting
0:46:08 The Power and Pitfalls of AI
0:47:39 Understanding the Framework of the Discussion
0:49:06 Collaboration Between Retailers and Brands on TrackFly
0:55:45 Follow TrackFly's Updates on Website and Social Media

Long Summary

In this episode of The Articulate Fly, I had the pleasure of welcoming back Stephen Baird from TrackFly to delve into the fascinating topics of artificial intelligence (AI) and privacy. We kick things off by going over the latest news and updates from TrackFly. As always, I take a moment to urge our listeners to share the podcast with their friends and leave a review. I express my deep appreciation to the incredible Articulate Fly community and give a special shout-out to our new series, "On the Salt with Gotwin," which is sponsored by Norvice and co-hosted by the talented Captain David Blinken.

Moving on to the main topic, Stephen and I begin by breaking down the different components of AI for our audience. We explore concepts such as machine learning, deep learning, and generative AI, illustrating how they are applied across various industries. Our aim is to demystify AI and present it in an accessible manner, so listeners can better understand its functionalities.

Transitioning to the crucial issue of data privacy, we highlight the distinction between data processors and data controllers. It becomes clear that awareness of where personal data is being shared and how it is used is of paramount importance. I bring up the well-known saying "if you're not paying for a product, you are the product," underscoring the pervasive monetization of user data by free or low-cost applications. We also touch upon the complexities surrounding privacy policies and emphasize the necessity for users to be attentive to their own comfort levels when using certain apps.

At this point, our conversation veers towards the significance of privacy by design and privacy by agreement. Stephen sheds light on the technical limitations and parameters that ensure applications operate in accordance with privacy laws. We delve into the ways in which companies capitalize on user data and explore the role of data brokers in collecting and analyzing data from multiple sources. It becomes apparent that consumers need to be well-informed about the value of their data in order to make informed decisions.

Concluding our discussion, we address the limitations of AI in podcasting and stress the irreplaceable value of the human touch in the creation of podcasts. We touch upon the potential benefits and risks associated with an over-reliance on AI. Stephen takes the opportunity to share updates on TrackFly's AI capabilities and exciting upcoming features, while I bring attention to an app that aids in AI-based receipt tracking. Lastly, we express our gratitude for this insightful conversation and urge listeners to stay informed through TrackFly's website, social media channels, and email updates.

Brief Summary

In this episode of The Articulate Fly, we discuss artificial intelligence (AI) and privacy with Stephen Baird from TrackFly. We break down the components of AI, highlight data privacy issues, and explore the value of the human touch in podcasting. Stephen shares updates on TrackFly's AI capabilities, and we urge listeners to stay informed through TrackFly's website, social media channels, and email updates.

Tags

Articulate Fly, episode, artificial intelligence, privacy, Stephen Baird, TrackFly, AI components, data privacy issues, human touch, podcasting

Edit Transcript Remove Highlighting Add Audio File Stop Play-on-click

Save Editor Export HTML Export PDF Export WebVTT Export... ?

Transcript

Introduction

Intro:
[0:04] Hey folks, it's Marvin Cash, the host of The Articulate Fly.
On this episode, our friend Stephen Baird of Trackfly returns.
We take a deep dive into the hot topics of artificial intelligence and privacy, and Stephen shares the latest Trackfly news.
I think you're really going to enjoy this one, but before we get to the interview, just a couple of housekeeping items.
If you like the podcast, please tell a friend, and please subscribe and leave us a rating review in the podcatcher of your choice. It really helps us out.
And as we head into the Thanksgiving weekend, we want to let you know how thankful we are for the articulate fly community.
We truly appreciate all of our listeners, supporters, and fly fishing friends far and wide.
Please have a safe and happy Thanksgiving.
And we also want to give you a heads up about a new series we've recently launched, On the Salt with Gotwin.
It's sponsored by our friends at Norvice and co-hosted by Captain David Blinken.
Give it a listen to learn how to win a Norvice tying system and a hardy reel loaded with your favorite wolf line.
Now, on to our interview.

Marvin:
[1:08] Well, Stephen, welcome back to the Articulate Fly.

Stephen:
[1:12] Hey there, Marvin. Great to be back with you. Thanks for the invite again.

Marvin:
[1:15] Yeah, absolutely. And you know, as we kind of come into the home stretch of 2023, how was your fishing season this year?

Stephen:
[1:22] You know, it was, it was great. We had, so out here in Utah, we had a banner winter last year with a ton of snow.
I'm sure you heard that in the news. Kind of delayed the runoff a little bit for us. But man, when it happened, it hit hard.
We had a nice, amazing fall with some great fishing out here.
But beyond that, I had a really fun year as I was traveling around the country, meeting with different fly shops, different brands, and really got to fish in some amazing parts of the country that I never thought I'd have the opportunity at least not in a long time.
So it was a wonderful year, I'll tell you that.

Marvin:
[1:58] Yeah, and it's been a really eventful year for TrackFly too.
I know you had a major unveiling at the most recent AFTA event.

Stephen:
[2:05] Yeah, we did. So at the most recent event, the former IFTV, the new Confluence, it was right here in our backyard in Salt Lake City where we had the opportunity to launch the first of its kind, A to the Industry Report, where we talked about true industry trends of top-selling brands, products.
We looked at what was actually happening at the ground level across supply shops across the country.
And, you know, we're just getting started. We launched a couple of really cool things, including our first-ever retailer dashboard, where retailers can start to look at these metrics and apply it to their business practices.
And we've got some really cool things coming here in the next couple of weeks as well, as a result of those discussions.

Marvin:
[2:47] Yeah, that's really neat because, you know, historically, there's been anywhere from basically no data to really really skinny data sets in the fly-fishing industry.

Stephen:
[2:57] Yeah, that's exactly right. You know, one of the things that we've been promoting

Discussing Artificial Intelligence (AI) and Privacy

[3:01] for the last year now is getting away from the survey model.
You know, that's really what was available to the flight fishing industry historically was taking time, filling out questions.
And you know, it served a good purpose, but we've advanced well beyond that technologically.
And it's been really exciting for TrackFlight to be able to work to bring these type of applications to this industry to start getting broad data sets and an accurate picture of what consumers are actually doing in this industry of ours.

Marvin:
[3:34] Yeah, it's interesting. It's always a tough nut when you do surveys, right? Because it usually only brings in people that are really, really happy or really, really unhappy.

Stephen:
[3:43] That's exactly right. That's exactly right.

Marvin:
[3:46] Yeah. And so, you know, folks, there's been, you know, Stephen and I talk more than infrequently, but not probably as frequently as we would like to about industry stuff.
And if you've listened to the podcast for any length of time, you know I'm kind of like a closet nerd and used to write computer code on a TRS-80 when I was about 10 or 11 years old and save it on audio cassettes of all things.
And so, there's been so much in the press about generative AI.
And Steve and I would keep having these conversations where we're like, gosh, no one's really kind of talking about this in a way that kind of makes it digestible for people that aren't propeller heads.
And so I wanted to bring Stephen back and say, hey, Stephen, let's take a deep dive and do something a little bit different.
And it ties into fly fishing, because I've been lucky enough, and Stephen, you know these developers too, to interview quite a few technology companies in fly fishing.
So we thought we would try to give people an overview, and deconstruct AI, talk about privacy, and help them understand what's going on in their day-to-day lives just so they can be a little bit more informed about kind of what's going on. So, what do you think about that, Stephen?

Stephen:
[4:57] Yeah, you know, it's so important, Marvin. And I know this is this is absolutely why we were excited to do this discussion together is because AI, as a buzzword is on everybody's lips right now.
You know, whether it's in fly fishing, it's really across the entire globe.
And it's one of those things that's just incredibly important to understand, not just how does it impact me personally, in my life, but especially as a business owner, you know, there are a lot of different ways to cut up this this term AI, and making sure that you are applying it appropriately, you're protecting yourself, but also you're taking advantage of it and using it as a benefit.

Marvin:
[5:35] Yeah, it's a fascinating thing. We were talking before we started recording that there's some really kind of interesting questions that we'll get to probably a little bit later about how it impacts creatives, right?
About creating content, whether it's music or podcasts or the written word, and what does it mean to have a human write something? but we'll talk about that a little bit later on.
But, you know, Stephen, before we kind of completely go down the rabbit hole and everyone's eyes roll back in their head, I thought it would be helpful to your point to, you know, there's no monolithic AI.

Breakdown of Artificial Intelligence Components

[6:09] I was wondering if you could kind of break down kind of the pieces of the puzzle that make up what people commonly refer to as artificial intelligence.

Stephen:
[6:17] Yeah, you know, it's a really important thing to understand, because there is what, you know, being published across as just this broad term of AI.
But if you actually look at it, there's some really important components of it.
And so if we just talk about, you know, the definition of artificial intelligence, you know, humans and machines, we've been interacting with each other for, you know, eons now, decades.
But what this really means is it's actually machine, It is computers, it's algorithms, it's programs that are taking on, you know, effectively cognitive functionality.
It's making decisions, it's looking at things in different ways, and it's actually producing information.
But it's coming, you know, based from a machine. And there's really three components that you need to be paying attention to. you.
You've got what's classified as machine learning, which we can talk a little bit more about what machine learning is.
Then you've got deep learning, which is in its kind of own form, a form of machine learning.
And then you've got the generative AI. And that's what you referred to earlier, Marvin.
And that's really the thing that's making the radio waves.
That's what everyone's talking about, because generative AI is truly something that we should be aware of and be cautious as we're approaching generative AI, but also very aware of what its capabilities are.

Marvin:
[7:41] Yeah, and so to kind of back up, my understanding is that we've had machine learning for a very long time, relatively speaking.

Stephen:
[7:48] Yeah, you're exactly right.

Marvin:
[7:50] You want to kind of elaborate a little bit on that and kind of places where people may not know that that's been living probably for the last, what, 10 to 15 years?

Stephen:
[7:59] Yeah. Well, in fact, you know, if you think about machine learning, you can actually even take it back to the earliest computer, you know, what were computers originally designed for it was to process information faster.
And so when you think about machine learning, you know, machine learning gives exactly.

[8:16] It's pretty well defined in its name. It's enabled to take algorithms, the continued algorithms that have been built for decades now with the earliest computers, but you actually provide that algorithm, what's called a trained dataset.
It looks at that dataset and learns and compounds on itself in order to process information faster.
You can give it patterns. If it identifies patterns, it can start to self-classify, information in order to process data sets at accelerated rates.
And what this really allows you to do is expand or optimize the existing computer systems that have existed, again, for going back decades, going back to the 60s and 70s, some of those very early on computers.
It's the same algorithm, but faster.
Now, what you're kind of talking about too, Marvin, of what's been evolved in the last 10, 15 years is what I would actually put more into the deep learning category.
And that's when you start to get into these neural networks where it's essentially machine learning on steroids, if you will, where these training sets are actually feeding themselves, and creating new algorithms, new networks where the machines in and of themselves, these computers, these algorithms are self-propagating.

[9:38] And again, all in the name of speed of information, all in the name of scaling out faster.
And so that's where you can talk about technology that has existed for a very long time, that's continuing to improve to ultimately process data faster.

Marvin:
[9:55] Yeah, and so I guess maybe some examples would be, for example, like UPS or FedEx using that to improve their logistics, like where the trucks go and how they load the trucks, right?

Stephen:
[10:07] It's exactly right. It takes a foundational data point or data set, and it allows you to look at information faster to make key decisions.
Another very fantastic example of this that I used to rely on heavily in my past was working with technology that did email filing.
We're all in Gmail, we're all in these different systems, and when you start to look at an email filing system, you can look at characteristics, you can look for algorithms.
You can look for words, for specific data points to identify where an email should be filed in your inbox.
That would be another really, really specific example of how machine learning can be applied in our day-to-day lives.

Marvin:
[10:50] Yeah, I suspect it also probably lives a little bit in these autocompletes that we're seeing on our iPhones and seeing in Google Docs too, right?

Stephen:
[10:57] That's exactly right. You know, that's not generative. It's not, it's not coming out of nowhere. It's being fed on prompt.
And that's really, really where that machine learning kicks in is there are data points that it's seen before that, you know, the algorithm or the machine has seen before, and it's giving you that information faster based upon your app.

Marvin:
[11:17] Yeah. And I guess the interesting thing for people, you know, algorithms are just, I guess, really, I think of it as a fancy word for a rule, right?

Stephen:
[11:25] Exactly. Yep, that's exactly right. So these machine learning sets, they just are getting fed rules after rules after rules.
But that deep learning space is where they're creating their own rule and it's a neural networks.
It's a fascinating thought. I don't know if we want to get into neural networks on this discussion, but for those data heads like you and I, Marvin, that want to go dig into that, neural networks are a fascinating piece of how machine learning has evolved in recent times.

Introduction to Generative AI and Its Impact

Marvin:
[11:56] Yeah, and we'll save that because we keep talking about maybe doing a podcast series on tech that won't be the Articulate file, it'll be something else.
But, you know, I guess the way that learning is happening is you're probably given the computer's optimization conditions, right?
And they are basically solving for rules that hit the targets.
Is that roughly kind of in a very general way how that works?

Stephen:
[12:16] That's absolutely correct.

Marvin:
[12:18] Yeah, so, you know, now that we've talked about machine learning and deep learning and people kind of have some examples, you know, of where that lives in their 2023 life, let's talk a little bit about what makes generative AI different.

Stephen:
[12:35] Yeah, and so when you get into generative AI, I mean, this is where the tools, you know, we've all probably heard the words chat GPT.
You know, we've heard about all these different tools that are coming out right now that are AI driven.
That's really where you get in the generative place, where content is actually being created, normally as a response to a prompt.
But it's actually those cognitive functions, those human cognitive functions that we would associate with the human experience coming out of these AI applications.
So it's still kind of a bit of a wild west when you think about generative AI.
There's the full scope of the impact that can be creative, the content, the text, the imagery, all of these things when you start to apply cognitive responses to machines, you know, there's still a lot of unknown about that.
There's definitely risks about it to be aware of.
There's definitely benefits that can come of it as well as it can be applied to different business practices.
But that's really the thing that I would say has crept up in the last, you know, 12 to 24 months.
You know, it's been obviously worked on for a lot longer than that, but it started spilling across businesses, operations, individual, social media, really just over that time period as it's really become available to everybody.

Marvin:
[14:04] Yeah, and so what we're going to do is we're going to kind of start to talk a little bit about privacy to work our way back to AI because, you know, the secret sauce that makes these models work is just insane volumes of data, right?

Stephen:
[14:15] Yeah.

Marvin:
[14:16] And so that to me seems like a good place to maybe kind of pull back to privacy.
And I know the last time we were together and we were talking about TrackFly and how you design the software, we talked about the difference between data processors and data controllers.

Difference between Data Processors and Data Controllers

[14:29] And I was wondering if you could give us a little bit of a refresher.

Stephen:
[14:33] Yeah, absolutely. And you hit the nail on the head there, Marvin, in that providing information is absolutely what makes all of these applications and algorithms work.
From the standpoint of, you know, we see on LinkedIn, on social media, all these AI generated profile pictures.
Well, you gotta give a reference point for that. And normally that comes with providing a picture of yourself.
And so when you start to get into data privacy, you know, there's those kind of two things you called data controllers and data processors.
When we talked about that last time, the real key thing to note is when you classify in your privacy policies as a data processor, what that means is that you are connecting data points between two entities and you are not owning that data at any point.
It allows you to enforce a lot of very personally driven or user-driven benefits, and noting that that data is always owned by the user.

[15:34] Different privacy laws like CCPA, GDPR, to name just a couple, can be enforced because you can request to remove those at any point in time.
And a lot of tech companies, I think, are starting to become aware of the importance of classifying as a processor because of those privacy laws.
Now, when you classify as a data controller, it's a little bit different because that essentially means that the data in your platform you own.
And so if you start to submit your personal data to these different applications, be aware of that, that if that is going into an application where your data is being submitted to a controller.

[16:12] That application essentially has the liberty to distribute or to make use of your data the way that they seem best fit for their business.
And those are kind of the big key things to take note of there.
And there's a lot of things that come into apply when you're a consumer of data, but also when you're applying it or supplying it to these applications.
You know, we talk a lot about the cost of technology to out, you know, weight against the benefit.
And don't get me wrong, where there is absolutely immense benefits.
To providing information to these applications. It makes life a lot more convenient when you're shopping for products, when you're trying to decide which series you want to stream next on your streaming applications, when you're trying to figure out where you want to go on your next vacation.
The more information that you provide can absolutely help to dictate where you're applying your decisions.
But just be cognizant that normally the data that you're providing is going a lot further than where you think it is just for that information to make it back to you.

Marvin:
[17:18] Yeah, and I guess, you know, the easy examples for a data controller would be someone like Facebook, right, or Amazon.
But I think, you know, what a lot of people don't understand is, you know, for lack of a better word, relatively benign non-tech companies that, you give data to will go monetize that data separately from your business transaction.

Stephen:
[17:40] That's exactly right. And that's kind of to highlight that point right there.
Oftentimes, when you add things to your Facebook profile, add things to your Instagram account, we see what's on the surface and the immediate benefit.
But what happens behind the scenes on how far that data is actually traveling, that's something that's not always visible.
In fact, most of the time, it's not visible at all to the end user.

Marvin:
[18:06] Yeah. And you know, another thing we talk about is, you know, privacy by design and privacy by agreement. You want to kind of expound on those a little bit?

Stephen:
[18:15] Yeah, it's something that we've made a very strong point at Trackfly, you know, from the get-go.
When you're enforcing privacy by agreement, you're normally just withheld or beheld to the laws, right?
And you know, there's something to be said about being beheld to the laws.
People still break the law.
Unfortunately, that's just the world that we live in. And there's always, you know, abilities to enforce that. you know, there's claims that can be made, but it allows for things to happen that are outside of our purview.
And that's something that's a challenge when you go into privacy by agreement.
On the flip side of that, when you get privacy by design, and this is something that we've strived from day one at Trackfly to implement, privacy by design means that you have technical limitations.
You've put parameters in your software, in the application itself, that prevents that application from operating outside of the laws, if you will, the quote unquote laws.
And so when you've got that enforced in your agreement, that's great, you want that in place, you wanna be operating within the laws.
But when you have that privacy by design, it's that added layer that prevents.

[19:30] What you don't want to have happen from happening.
It reduces the need for submitting claims against agreements and all of those types of things.
And for most, I would strongly urge any type of application that you're reviewing, that you're evaluating, but also for any tech companies that are out there as different startups are coming in place, building applications, look at the privacy by design, look at making sure that your end users are protected.

Setting Up for Success with Privacy by Design

[20:02] It really is setting yourself up for success going forward for what we don't know what this evolving landscape of artificial intelligence of machine learning all the varying laws that are going to be changing and involving coming and coming down the road set yourself up for success by having that privacy by design.

Marvin:
[20:20] It's an interesting thing right to be really thoughtful about who the product is right and where the monetization comes from.

Stephen:
[20:28] Mark, our good friend over at Facebook said it himself, if you're not paying for the product, you are the product.
It's something to be aware of. That it's not always a bad thing to have reduced cost or no cost to use an application.
But at the end of the day, just be aware of that and know that it's likely means that you're paying for that service in a different way. You know, at Trackfly, we offer very reduced costs for our partners that come and play.
But it's also because we're very transparent to know that you're getting the value of the platform through different means. And that's normally coming back through data.

Marvin:
[21:07] Right. And so, you know, just to kind of help people understand, kind of, if we can kind of, you know, lift the curtain in the land of Oz, you know, what are common monetization strategies between free or quite honestly, you know paid apps where you know I guess consumers don't really know this because they don't realize that it would be, $50 a month that the data wasn't sold but it's $12 a month that they do sell the data but you know, what are some common ways that, platforms and businesses monetize data.

Stephen:
[21:37] I mean there's there's one that is i think stands out amongst the rest of its advertisers there are no companies across the world.

[21:46] There's some that we know there's brands that we know there's many that we don't know that make a lot of money and invest a lot of money to advertise advertise products and services to very targeted audience.

[22:00] And there's some really cool application, there's some really cool functionality.
If you're an influencer, the influencers out there, you're probably very familiar with this.
You can go into Instagram, you can go into Facebook, and you can create very targeted algorithms to make sure that your product or service is getting in front of the right and most qualified users or buyer.
And so when you think about where that data is going, I mean, it's to the bidders, to the people that will pay the money to make sure that they are getting in front of the right audience.
And so really what you're providing, you know, as a service, if we're talking social media for a really obvious and well-known example, when you're putting your data into a social media platform and you're adding things over and over, you're posting information, you're posting an image or a picture, you're feeding that information to ensure that your data is getting to the right service provider that would be most interesting to you.
And, you know, it's a two-edged sword. I mean, obviously there's negative to that data being sent out.
On the flip side of it, it can also be very convenient because it makes sure that the ads or the products or services that you are being targeted with might actually be something that would be beneficial to you.
So it definitely has a dual-sided approach or dual-sided experience.

Marvin:
[23:21] Yeah, and what are some other kind of more subtle things, like maybe things that would live in, for example, a paid app or a paid interaction about how a business might monetize the data they collect?

Subtle Ways Businesses Monetize Data

Stephen:
[23:36] Yeah, I mean, when you start to get beyond what's on the surface, you're really looking at economics, you're looking at, you're looking at behaviors, you're trying to understand, I mean, I don't know how deep you want to go here, Marvin, but I mean, you can talk about political alignment, you can talk about, social economics of individuals. You can talk about things that really, when you start to get to the biggest players at the top of the pyramids, this is the information that drives economies.
And there's a lot of individuals out there that want to look to understand again where, and they'll spend a lot of money, they'll pay a lot of money on these application to make sure that they have a pulse of what's happening outside of their own view or vantage point.

Marvin:
[24:34] Yeah. I mean, and so an example would be like looking at correlations between people that buy certain things and then maybe that translates into certain behaviors, right?

Stephen:
[24:45] That's exactly right. That's exactly right.

Marvin:
[24:47] And so, you know, to kind of put that on steroids, if you have a business and you're collecting the data and, you know, you're the CEO and someone walks in your office and says, hey, I found a to make some new money.
We just found some Cheetos and some change in the sofa.
You know, we can sell the data to a data broker because it doesn't violate any of our customer agreements.
You want to talk to people about, you know, data brokers like Spokio and people like that.

Stephen:
[25:11] Yeah, I mean they're all over out there and this can come as simple as when your phone connects to a network and there's network providers that are able to identify movements, actions, activities, purchases, online purchases, all of those types of things.
You get to this world of data brokers that essentially sit in the middle and they're collecting and organizing it in a way so that That way, you know, the information that's relevant, on the upstream of the data relationships and of the data consumers, they're making it in a way that can be consumed at their end.
And so it gets to be this wild west, if you will.
And again, just on that exercising of caution as you get into any of these apps, to know and have that awareness is that it's not just you interacting with Instagram.
It's not just you interacting with an online store.
There is typically node after node after node of brokers of individuals of organizations and entities that are looking at different behaviors, looking at different data points in different ways.
And that is truly how data is monetized throughout a vast network of applications communicating with each other.

Marvin:
[26:31] Yeah, I think one of the amazing things I learned probably two or three weeks ago that some of the credit agencies actually have, sister companies that basically are playing in this game, even though they're not shared directly sharing the credit reporting data, they're developing products to basically generate this information.

Stephen:
[26:51] Yeah, yeah. I mean, just another really good example of it.
When you interact with your credit union, when you interact with your banking application, it's not a one-to-one relationship.
It's not the buck stops there through that process. There's information that's being shared on a very wide level.
And it's all in the name of convenience and economics and and economy, you know, growing things and making sure that consumer experiences are optimized for a fraction of them.
But there's just so many players that come in now when you interact with any type of application.

Marvin:
[27:32] Yeah, and it's an interesting thing. I mean, you know, I kind of think about this, like I don't think it's per se bad, but I think that, you know, consumers, for lack of a better word, you know, ought to be in a position to make an informed decision if they're giving away something valuable, right? And I think they own it, right? Because it's their data.
But, you know, the horrible thing is, I mean, anyone who's listened to the podcast knows that I'm a lawyer by training and occasionally I'll drop down the rat hole and read one of these privacy policies.
And, you know, one, most people don't. But then, you know, when you read the policies, they're incredibly vague, right?
Which makes me think about reading, like, life insurance policies, right?
It's like, you know, the only thing I know is that this isn't in my best interest, because I don't understand how it works.
And so, you know, if you're, you know, particularly if you're not a particularly tech-savvy consumer, you know, how do you make more informed decisions about, you know, the data you're sharing and not put yourself in a situation where you're sharing stuff you don't want to share?

Privacy Policies: Complex and Difficult to Interpret

Stephen:
[28:37] Yeah, you know, it's it's interesting. I've spent a lot of time obviously in these privacy, privacy policies with different applications over the years.
And I hate to use the word complex by design.
But unfortunately, with with legal practices with policy with with terms of service, because you have to cover such a broad, broad swath of laws and regulation.

[29:04] Unfortunately, we've bolstered an entire economy around individuals that can create language to cover.
And I say unfortunately, that's probably the wrong term to use, because it allows us to protect corporations, protect individuals through languages that, unfortunately, is not always easy to interpret.
And so as an end user to applications, it's as tough as it is to do, I mean, look at the signs around how you're using the application.

[29:35] I mean, we already talked about this, but a really easy first question is, is it free?
If it isn't free, there's normally something to know that there's something happening under the hood or behind the door, behind closed doors.
Other things to look at is, what information are you putting into it? Is it imagery?
Is it demographic information? Is it how frequently are you using the app and are you using it frequent enough that it can notate different location?
You can start to understand by your use of the application, what type of data is being tracked and shared at that point.
Really the best thing we can do, Obviously, if you can comb through privacy policy, it helps you to have an understanding.
Most of the time though, there's not a lot that an individual can do to adjust or change the privacy policy of an application.
Really, my strong recommendation for users of applications is.

[30:36] As silly as it sounds, how does it feel when you perform that function?
If you're uploading a lot of images of yourself to create a generative AI image, if it makes you feel uncomfortable, it's probably because there's something that's happening there that you should dig into or look at.
If you find yourself constantly getting your location pinged by an application, and that's something that you notice or makes you feel a little bit uneasy, pause and think about it and look at the use case there.

[31:08] And anytime those moments happen where your spidey sense is almost telling you that something's up, there's things you can do, especially on mobile devices, go and look at your location tracking.
That's one of the first places that I would always go and look if there's something that's constantly tracking your movement.
If it's not directly related to a strong value add that you're getting from the application, maybe consider reducing that or turning it off altogether.
If you're loading in data points regularly, and you're noticing that you're getting very catered experiences, or things that are happening there, review the value that you're getting from that application and does it work in putting that data into that system.
A lot of it just comes down to behavioral experiences and the personal value that you're extracting from an application, and that's really the tools that we have in front of us to make sure that we are having safe practices with each of these experiences.
You know, it's the same things kind of, you know, Marvin, to take it to another level.
It's the same things that we talk to our kids about, you know, as kids are growing up and learning and experiencing the world.

[32:14] It's kind of funny, my wife is actually an elementary school teacher, and she talks about all the time about talking to her students about situations that make you feel icky.
Pay attention to those moments that you feel icky when using these applications, don't ignore them.
And if they're in there, if you're having that experience, you can dig in a little bit further to the policy, to the functions, to the application, and ultimately step out of any situation that's asking too much.

Marvin:
[32:39] And, you know, another thing, too, if you have apps that you don't use, you can always delete them, right?
You know, if you're not using them regularly, it's another way to not inadvertently give away data. And then, you know, we were talking before we started recording.
There was an article, I don't know, gosh, probably in the last four to six weeks in the Washington Post about kind of how to get a little bit more control of your data.
And, you know, one of the interesting things, you know, everyone's familiar with consumer reports.
And they're obviously kind of interested in this issue. And so, they've built an application that's almost like a data cleaning concierge where you basically can go into the app and give them permission to ask these companies not to sell your data.
And you get a list of companies, and they'll actually kind of run that process for you.
And I'm not vouching for it.
Your mileage may vary, but I would say in the trust world, I have a fair amount of trust that Consumer Reports is not out there doing something sketchy on the back end of the app.
So, right, right, but and then, of course, when you get all the people who refuse to delete your data, it kind of tells you how valuable it is, right?

Taking Control of Your Data: Consumer Reports' Permission Slip

Stephen:
[33:44] Yeah. How many times are you asked to not leave the application?
That's another good, another good screener. The old, the classic canceling your AT&T subscription.

Marvin:
[33:56] Yeah, and so, you know, and I think we touched on this a little bit, but to kind of bring it floral circle and kind of bring it back to TrackFly, you know, can you talk to us a little bit about how, you know, these privacy issues started to affect your kind of design considerations as you were building out your software?

Stephen:
[34:14] Yeah, absolutely. So I mean, at the end of the day, we are a data company, you know, we fit into that model of, of collecting data point, to be able to provide information that's valuable to our customers.
And, you know, you talked about a couple of different thing, we talked about, you know, the privacy by design.

[34:37] That really is in full force by the data that we collect.
So there's two different ways to collect data, especially through different integrations.
Obviously, at TrackPlan, we utilize integrations with various applications to monitor cell through data.
One of the first and foremost things, and again, this is in our privacy policies, but more importantly, it's in our technical design, which is we only collect relevant data points to the service that we provide.
That's first and foremost. We do not want to collect anything that goes beyond the service.
That's normally where you get into a realm where you have data points not visible to your customer that are being processed either through brokerage to other sides of applications.
That is not something that's aware to your customers. So that's one of the very first and foremost things that we do is we only collect the data that's relevant to our service and application.
The other thing is full transparency in data.
So anytime we work with a retailer and a retailer uploads or provides data to our system to monitor market information or sales information.

[35:52] All of that data, all of the tables, data tables for that retailer are actually completely visible.
You can go in and you can access to see what data you've contributed to the reports. Now obviously.

[36:07] That is very specific to your data that you provide.
But before all of that is then uploaded, aggregated, and shared to a collective reportable dataset.
It's processed in the strictest methods of data hashing, data cleansing to ensure strict anonymity so that no individual, individual business and or person is actually identifiable within that broader reportable dataset.
And that's, again, fully transparent in the application.
Because you're seeing that you're actually consuming the data that's being contributed.
Outside of that, the other thing that's built in by design is as a data processor, there is derivations of data that TrackFlight does own, but it's all based upon calculations and reported upon anonymous sets, so that all data that's being communicated from any retailer or brand that's on Trackfly's platform, again, has full visibility into what information is being shared, collected, or aggregated, and can make those elections based upon the value that they get out of the platform.
The final thing that I would say is we've made it easy that if a brand or retailer does not feel comfortable sharing or does not want to participate.

[37:30] Deleting, cancelling, and removing your data from the platform is incredibly easy to do.
And again, that just comes with that design. We want to be transparent.
We want to be an advocate of doing data the right way and doing that specifically so that our customers are directly being impacted and extracting the value that comes with data participation.

Marvin:
[37:56] Yeah, it's interesting. It's kind of like, you know, the business understands the ingredients they're contributing to the cake, right?
They get to eat the cake, but they don't know where any of the other ingredients came from, right?

Stephen:
[38:09] Yep, exactly.

Marvin:
[38:10] And they can kind of decide to not play at any time, right?

Stephen:
[38:15] It's exactly right.

Marvin:
[38:16] Yeah, it's interesting. And so, you know, if we kind of shift gears a little

Privacy Concerns in AI and Generative AI

[38:19] bit, and you know, earlier on in the conversation we talked about machine learning, deep learning, and generative AI, and I was kind of curious about your thoughts on kind of unique privacy concerns around, you know, AI and if there were any kind of unique things to like one flavor versus the other.

Stephen:
[38:37] Yeah, you know, and I'll tell you this first and foremost at TrackFly, we are absolutely involved with machine learning.
We're absolutely utilizing machine learning and deep learning technologies.
We're also utilizing AI technologies. You know, I think in the world that we live in, again, it would be inefficient to not look at the best class technologies that are available to utilize.
However, that being said, when you start to get into that generative AI, there's a really important distinction that comes in place, and that's utilizing that generative AI to make decisions for you or to utilize it to provide you with information that you can still use your own personal human cognitive abilities to be able to make creative, distinctive, and unique choices going forward.
That's one of the most important things that we're advocating for at TractWise, is to not let technology reduce or remove the human creative process, the ability to be unique, the ability to make decisions that are impactful uniquely for your business or for your individual self.
And so when we start to talk about AI capabilities, we absolutely, you know, prompt our AI tools to make sure that we are categorizing.

[40:05] Organizing, and constructing data that is information useful to multiple individuals looking at it.
That's a way for us to scale and to move quickly. And it's based uniquely off of our proprietary way to collect data and our proprietary way that we are communicating and utilize our AI applications.
And I would just say that that same experience goes for anybody out there utilizing those generative AI, whether it's for content, whether it's for copy, whether it's for imagery, you know, there's one thing, it's one thing to be able to feed it prompts and then just take those responses that AI is producing and just put that out there as your final product.

Using Generative AI as a Tool for Creativity and Innovation

[40:48] It's another thing to use it as a foundation, as a tool, which is what it really is, it's a tool, in order to create new ways of thinking, to be able to identify gaps, to be able to identify, you know, a thought process that maybe you would have missed if you didn't utilize it as a tool.
And then to find those gaps and fill it in with your own unique experience and your own unique take on the world that provides value for you as an individual.
That's really where it comes in very uniquely.

Marvin:
[41:20] Yeah, it's a gosh. It's a phenomenally. I mean gosh.
This is the understatement of the millennia. It's an insanely complicated issue to kind of work through, um, you know in the sense of um, you know One like, you know as you're talking about like not creating stuff or not having stuff created for you and sharing it as your own um, but there's a really kind of a difficult thing in the sense of, You know some things are not don't necessarily have to be personal right to be effective So, for example, if you leave your office every day at six and go home, you don't really need a personal touch to send you a reminder to send your wife an email that says, hey, I'm going to be late, right?

Frustrations with AI in Podcasting

Stephen:
[42:03] Although my wife might appreciate that, Marvin.

Marvin:
[42:05] Yeah. But you know, it doesn't – you don't have to like look at that and craft it for yourself.
But I do think there's some really, you know, interesting things like one of my frustrations, you know, when we were talking about this before we started recording as a podcaster and kind of creating in that space is you kind of, you know, the big push right now is for the most part is around show notes, time stance, and social media creation.
And, you know, you kind of look in some ways, you kind of are getting into this trap where you're using a machine, you know, an AI AI bot to basically create the post on social media to get the social media post bot to then share it to the world to get a listener.
And that seems slightly silly to me.
And of all the things in the – if any developers are out there wondering what Marvin Cash wants from AI in the world of podcasting, the thing that to me that would be phenomenally helpful is there's a lot of stuff that you have to do to put out a podcast, right?
Whether loading stuff in content management systems, cropping images, all that sort of stuff, that if someone could start to streamline that so you could kind of, for lack of a better word, teach the algorithm how you like to post content, that would be huge.

Stephen:
[43:22] I love it and you know it's it's funny. I've got a really funny story for you actually.
So at the beginning we talked about the after Confluence event that we had you know a great launch there, great meetings with people.
There is a sticker that we created that is circulating right now and if you were at Confluence and you took TrackFly stickers, you may have this one.
We actually created a AI generated fly and it was just to kind of put this on and it's a little Easter egg. We didn't promote it. We didn't talk about it.
But if you've got a track fly sticker from confluence, look at that fly.
And tell me if you think you'd catch ever ever catch a fish with that.
And if the answer is yes, you're not looking at it close enough because there are fine details that you need to pay attention to.

[44:12] That if you just use what is produced from AI, you miss out on all of the under all of the behind the scenes work, all of the little nuances that need to take place for it to be an accurate experience.
And it's not to say AI will get better, it will experience that, it's going to continue to get smarter, it's going to continue to be able to identify these things.
But to your point Marvin, you cannot just take the responses from AI because you will be missing things, you will be missing all of the details, all of the work that goes around creating a podcast, identifying, you know, the technologies to make it what it is today and to reach an audience that you have set up with the articulate fly.
There's a lot that goes behind the scene that AI simply cannot and likely will never be able to do.

Marvin:
[45:05] Yeah, it's an interesting thing, I mean, because I think at some level, particularly when you start looking on the creative side, and it doesn't really, I mean, there's creative that's, you know, in kind of today, day-to-day business, You know, it fundamentally begs the question, you know, what does it mean to be human?
Yeah, right and it's an it's a fascinating thing because you know The other thing and flip side of it is and you know, we've seen this before You know if you get an extra two hours a day, but you just spend it on Instagram It probably doesn't really matter whether you have generative AI in your life or not, right?
You know, you could go to the gym. You could read a book That a person wrote, So it's an interesting thing, right?

Stephen:
[45:44] Absolutely absolutely it's the world is going to continue to evolve and this is why this this topic is so important and you know it's again I know I said this a couple times now but it really is so important the world's going to continue to evolve.
I am here it's not going anywhere learn how to utilize it learn how to.

The Power and Pitfalls of AI

[46:08] To avoid the pitfalls though that can also come by relying on it too heavily.
And it's kind of funny, there's a lot of comparisons between when the internet was launched and with AI right now. And I think it's an accurate comparison.
We can't imagine a world right now where we didn't have the connectivity that comes with the internet.
In five, 10 years from now, we're probably going to say very similar things about AI.
But just like the internet is a powerful tool, it also has some damage that it's produced across the years as well.
You can have positive and negative to everything in life.
And if you can utilize AI to make good decisions, to promote your own creativity, to grow yourself personally, and you can supplement it with going to the gym, reading a book, doing things that are involved in the human experience, we're going to see that it can become very powerful.
That being said, it will also have the side as well that if we stop doing those human things and we start to lean on it too heavily, it will have some large and wide-span, negative impact, both societally and economically.

Marvin:
[47:25] Yeah, it's the it's the seduction of convenience and I know you've heard me say this before. I always say, you know, if I have a shovel, I can dig a hole or I can hit you on the head and it's really kind of up to me which one I want to do, right?

Understanding the Framework of the Discussion

Stephen:
[47:39] Well, I don't know which one you'd do to me, but I'll wait to find out. I'll hold up on that one.

Marvin:
[47:45] Yeah, there you go. And so, you know, hopefully, folks, this has been kind of helpful to kind of understand, at least to kind of give you a little bit better framework for what's going on.
And you know, before I let you go tonight, Steven, you know, I know you're probably frantically putting together new features and product updates for either late this year or early next year. or do you have anything you wanna share with our listeners?

Stephen:
[48:08] Yeah, we've actually got a couple of really exciting things that are coming out here just in the next coming weeks and months.
First and foremost, so in the topic of AI and machine learning, we are refining that aspect of TrackFlight every single day to get more dialed in or accurate product categorization and trend, which ultimately leads to really accurate trend, consumer trend information that's happening for specialty retailers.
Again, all of this is visible and given directly back to the customers utilizing TrackBlight.
That's one thing that's just continually ongoing that I would be remissed if I didn't shout out to.
Additionally, two really cool features that are coming out here in the next couple of weeks, is the ability for our retail partners to connect with their brand partners.
This is something that is really, You know, it's existed for a long period of time and very costly.

Collaboration Between Retailers and Brands on TrackFly

[49:06] Expenditures for very large brands to do with very large retailers.
And we couldn't be more excited to bring this capability to the specialty retailers that are utilizing TrackFly and the brands that they partner with.
And really what you're going to start to see is retailers inviting brands to collaborate with them on TrackFly and vice versa.
Brands starting to invite their retailer partners to really start to collaborate on inventory.
And if there's one thing that we know is shaking out right now, and will continue to shake out over the next 12 months, is cleaning up inventory issues.
And we believe that this type of connection capability is going to really support the means of flatlining that and solving a lot of those inventory issues over the coming months. So keep an eye out for that.
We'll do a market launch on it when it's available, obviously.
But then also pay attention when your retailers and brands are inviting you to connect. That'll be really exciting.
And then the last thing that I'll promote here is starting early next year will be the first version of the brand dashboard.

[50:09] We launched at Confluence the retailer dashboard. The brand dashboard is now coming up here in the next couple of months.
And that's going to be really a way for you as a brand to monitor the market trends, the value of your specialty retail channel, and ensure that you're making all the right decisions to grow and see those those trading partners succeed.
So those are just a couple of things obviously there's a long tail of development projects that are in the work right now but definitely keep an eye out for those when they're gonna be launched during the coming weeks and months.

Marvin:
[50:37] Yeah and kind of one of the interesting things too I mean because you and I have talked about it is there's gonna be at some point kind of in the evolution the ability to bring all of this technology not to replace manufacturers reps but to actually help them be better business partners with their brands and their retailers.

Stephen:
[50:55] Absolutely, and I'm glad you called that out. We are actually meeting with many of these manufactured reps right now, developing a very impactful tool for them.
So if you are exactly that, if you're a manufacturer or an independent rep, please reach out to us at TrackFlyInfo or TrackFly.com.
We would love to get your input to make sure that this is being designed in a way that that supports your success.

Marvin:
[51:19] Yeah, absolutely, because I mean it sounds to Like, you know, what you're really doing is you're using AI at TrackFly kind of as a wrapper around your older versions of machine and deep learning to actually help you kind of clean up the data faster, right?
Because, I mean, we've talked about it's kind of like someone giving you a million shoe boxes full of pieces of paper and saying, please file my tax return, right?

Stephen:
[51:40] That's exactly right. Yeah, that's exactly right.
It sounds like you and I have probably done that before as well.

Marvin:
[51:51] No, I've had a scan-snap for a very, very, very long time.

Stephen:
[51:55] There you go. You're exactly right, Marvin.
We apply machine learning and AI so that any retailer can continue to use your existing system, continue to track your receipts in the shoebox, in a point of sale system or however you're doing it, use the processes that you're currently doing the systems that are in place and we map it and we can map it quickly because of these technologies and it gets you intelligence that otherwise would have taken a lot of time, a lot of resources, and and which is why it's normally been reserved for those companies that that truly just have the most resources because it historically has taken that.

Marvin:
[52:35] Yeah and so I'll just for for just individual small business owners or any other people.
I have an app that I love called Receipts that you can scan or put you know PDFs in of receipts and you do it one time and you basically tag the data with the manufacturer and it basically will read the amount and going forward you just put it in there and you know at the end of the year you have all of your receipts and it totals it up and it's all categorized. It's pretty killer.

Stephen:
[53:00] Amazing.

Marvin:
[53:00] Yep and then I get to spend more time on Instagram.

Stephen:
[53:04] And give it data, right?

Marvin:
[53:06] Exactly. Just feed the beast. And so I know you are probably trying to get to the million mile man on United.
Where can folks find you and your team on the 2024 show circuit?

Stephen:
[53:21] Absolutely. We're looking forward to this circuit. We're looking forward to talking to retailers and brands as we make aware across the country.
Specifically, you'll find TrackFly at the Denver Edison Atlanta and Bellevue Fly Fishing Show this year.
We'll be there in partnership with our good friends over at AFTA.
So please come find us. If you want to sit down and talk about the data work that TrackFly is doing or get involved and participate, we'll be there to answer any questions. and to do a couple of fun presentations as well.
So you won't want to miss that.

Marvin:
[53:55] Well, very good. And is there anything that I've left out this evening, Steven, you want to share with our folks?

Stephen:
[54:01] You know, just always want to be appreciative to you, Marvin.
I've always enjoyed the conversations that we've had.
We dig into some of the hard topics, which is really enjoyable.
The only other thing that I would say is anyone that wants to continue to dig into this, you know, we've got a great team of data engineers, of data scientists, And obviously, I just love nerding out of this at any opportunity I can.
So please reach out, you can send an email to info at trackfly.com or come to our website www.trackfly.com.
And we would love to talk to you about how our data practices are set up, how you can participate and how you can start to extract really important business information out of TrackFly to fuel profitable business decisions.

Marvin:
[54:46] Got it. And then, you know, for folks that may not be, you know, a retailer or brand or a rep, you know, any resources you can put, you want to point folks to, like, you know, Electronic Frontier Foundation or anything like that to kind of get more information and kind of stay up to date on kind of the state of play and privacy and AI issues.

Stephen:
[55:05] Yeah, you know, there's there's a couple of really good, good platforms that I go to look at.
And, you know, one of which that that I get notifications on pretty frequently is just from the McKinsey and company.
They're, they're doing a lot of research in there on generative AI really frequently.
I don't, I don't have probably anything else to say. I don't know if you want to include that Marvin or not, but No, I probably don't have anything to add there.

Marvin:
[55:33] Yeah, if I have any great brainstorm, I'll drop it in the show notes, folks.
And you know, Stephen, if folks want to kind of keep up with all the cool stuff you're doing at TrackFly and follow the company, where should they go?

Follow TrackFly's Updates on Website and Social Media

Stephen:
[55:45] Yeah, go ahead and head to our website, www.trackfly.com.
We're posting updates there frequently. Pay attention to our Instagram channels at, excuse me, at trackflyfish and or trackflyinc. You can also find us on Facebook, LinkedIn, all the standard channels there.
But then most importantly, if you want to get in touch with us, shoot us an email at infoattractfly.com.
And we are eager to talk to anybody that wants to explore this fun data world we live in.

Marvin:
[56:13] Well, there you go. Well, Steven, I appreciate you carving out a little bit of time for me this evening.

Stephen:
[56:19] Likewise, Marvin, as always, appreciate it, and we'll look forward to the next one.

Marvin:
[56:22] You betcha, take care.

Intro:
[56:24] Well, folks, I hope you enjoyed that as much as we enjoyed bringing it to you.
Again, if you like the podcast, please tell a friend and please subscribe and leave us a rating and review in the podcatcher of your choice.
Happy Thanksgiving and tight lines, everybody.