Today we are joined by data scientist and astrophysicist, Dr. Kirk Borne to discuss the importance of equipping people of all backgrounds and qualifications with AI literacy and fluency. Dr. Borne has a vast background in both the astronomy and tech spaces and is currently fulfilling his passion for making complex topics accessible through DataPrime’s blue-collar AI course.
In this episode, we learn the benefits of blue-collar AI education and the role of company culture in employee empowerment. Dr. Borne shares the history of data collection and analysis in astronomy and the evolution of cookies on the internet and explains the concept of Web3 and the future of data ownership. Dr. Borne is of the opinion that AI serves to amplify and assist people in their jobs rather than replace them and in our conversation, we discover how everyone can benefit if adequately informed.
Key Points From This Episode:
Tweetables:
“[AI] amplifies and assists you in your work. It helps automate certain aspects of your work but it’s not really taking your work away. It’s just making it more efficient, or more effective.” — @KirkDBorne [0:11:18]
“There’s a difference between efficiency and effectiveness … Efficiency is the speed at which you get something done and effective means the amount that you can get done.” — @KirkDBorne [0:11:29]
“There are different ways that automation and digital transformation are changing a lot of jobs. Not just the high-end professional jobs, so to speak, but the blue-collar gentlemen.” — @KirkDBorne [0:18:06]
“What we’re trying to achieve with this blue-collar AI is for people to feel confident with it and to see where it can bring benefits to their business.” — @KirkDBorne [0:24:08]
“I have yet to see an auto-complete come over your phone and take over the world.” — @KirkDBorne [0:26:56]
Links Mentioned in Today’s Episode:
Rob Stevenson 0:04
Welcome to how AI happens, a podcast where experts explain their work at the cutting edge of artificial intelligence.
Rob Stevenson 0:12
You'll hear from AI researchers, data scientists, and machine learning engineers, as they get technical about the most exciting developments in their field and the challenges they're facing along the way.
Rob Stevenson 0:25
I'm your host, Rob Stevenson. And we're about to learn how AI happens.
Rob Stevenson 0:31
Joining me here today on how AI happens is a man who contains multitudes, I think it's fair to say he has served a variety of roles across the tech space, such as being the principal data scientist and executive advisor over at Booz Allen Hamilton. He was a professor of astrophysics and computational science over at George Mason University. Currently, he's an advisory board member over a data prime. That's only a tiny piece of the picture here. I better bring him in before I insult him any more by introducing him poorly. Dr. Kirk Bourne, welcome to the podcast. How the heck are you today?
Kirk Borne 1:04
Good. Rob, great to be here.
Rob Stevenson 1:06
When I was going through your curriculum vitae, and just kind of looking at some of the titles you've had, my favorite was VP of technology slash astronomer, which I just love that you did both. And in addition to leading a an entire department, you also spent a good amount of time looking through a telescope and doing math. I imagine. Could you tell me a little bit about just your multivariate background? Maybe we'll start with the astrophysics and astronomy because it feels like that was kind of where you spent a big chunk of your career.
Kirk Borne 1:30
Yes, thanks. So yeah, the first half of my career, I guess, I would say was the thing that I always wanted to do since the childhood was which is being an astronomer be an astrophysicist? Well, that was always driven by the computational things I could do building models. That galaxies is what I did also, in my early days, built models of stars and other kinds of things. So it was all about modeling. But what was I modeling, I was modeling data that we obtained from telescopes, okay. So you might say was all about data and modeling data and modeling all the time. And so my first 20 years was not an any of the jobs, you just mentioned that my first 20 years was working on various NASA projects. So I worked with the Hubble space telescope for 10 years, I was the archive project, the data archive project scientists for that, again, data, data, data data. And then beyond that, I did another 10 years at the NASA Goddard Space Flight Center, where I was managing a team of people working at what's called the Space Science Data Operations Office where we basically manage datasets coming from NASA astronomy satellites. And so I tell people, I did data by day, my day job and data by night as an astronomer, but that's a bit of a misnomer, because telescopes nowadays are all remotely operated most, a lot of them are in space. I mean, I did do some looking through telescopes when I was younger, but that very little that actually happens in a professional astronomers life. But I did have an opportunity to do some of that when I was younger. So I was always interested in modeling complex systems using data, which tells us about the system that how can we build a model and what is a model is, it's a representation of the thing so that you can understand it, you can test it, for example, I can't test a galaxy, I can't go out to a galaxy and kick it and see how it responds. But I can build a model of a galaxy and crash it into another galaxy and see how it responds. And there's actually things in the universe that do that galaxies crashed into one another. So for me modeling data, computational and data science went hand in hand all those years doing astronomy. And then about 20 years ago, I started migrating more and more to other disciplines. Because I saw that the big datasets we were collecting astronomy, were nothing compared to all the big huge datasets we were seeing in business and government and healthcare, and cybersecurity and all these other domains, I mean, consumer data, etc. And so I just found it more exciting to take the skills I learned as a computational scientist and the data modeler or the data scientist and bring that into broader domain. Because what it was really satisfying within me was a deeper thing. Besides always wanting to be an astronomer, I always felt motivated to be an educator, I always wanted to train and teach people. And I feel like I can do that all the time. Now. It's it's basically training, education and mentoring, all the time for me now, in this in this role as a data scientist, so the advisory roles give me that opportunity. But also, I'm very active on social media. And I use I think I think of Twitter as my micro education platform. Okay, so I've constantly tweeting out things which I want to just sort of teach people, train people and foreign people. You'll help build literacy around data and literacy around AI as well.
Rob Stevenson 4:22
I want to ask you a question about data quickly here before we get on to some of the other stuff. But it's interesting that you say there was way more data being collected and perhaps synthesized in the the business and consumer space, then there wasn't astrophysics. That seems strange to me. Because in astronomy and astrophysics, you have, like you said, these telescopes in space, remote operated, probably constantly operating and looking out into infinity. Right? How is it possible that there's more data coming from the on Earth world than there is looking out into the endless depths of the universe?
Kirk Borne 4:59
Well, let me modify slightly in my statement, so in recent years, we're building huge telescopes on the ground, not in space. And we're doing what we call large surveys of the sky. So the datasets coming from these large telescopes on the ground, these large surveys all across the sky, the full sky surveys of the stars, the galaxies and everything there. Those are massive datasets, those are very large, hundreds of petabytes. So they are definitely larger than a lot of things. But in sort of early days, sort of individual astronomers would just collect a small data set at a telescope and analyze a particular question like you're doing an experiment in a lab, right? You don't want someone who's doing chemistry experiments in the lab, they don't have all chemistry data from the history of chemistry that they're working with. They're just working with that particular experimental data set. So that's one thing. But the other thing was when I was working at NASA on space missions, people don't realize this, but the bandwidth maybe changed and improved a lot in the last decades or so. But in those early years, the bandwidth of transmissions from space was very low, very low bandwidth. For example, when we had these fly by missions, like of Jupiter, and Saturn and Pluto, it would take months to transfer the data back to Earth, because the bandwidth from those transmission systems are very low. And so even the Hubble space telescope for a while there, people thought, well, Hubble is like producing massive quantities of data, the entire data set of Hubble, like the first 10 years of Hubble data could probably fit on my hard drive on my laptop. Because even though the data is phenomenal, high resolution, and beautiful, and just super high quality, the actual volume of it is actually relatively low. Now, again, in recent years, telescopes and communication systems and satellites, etc, um, have all improved dramatically. And so it's not strictly a true statement now. But it was true back in those days that you know that a gigabyte of data was like enormous, right. But of course, nowadays, you have a gigabyte on your phone, right? You probably have 100 gigabytes on your phone. So it's like, what we always called large back then is hardly what we call large nowadays, space is so big, yeah, you would think you could have enormous amounts of data. And that's what these large ground based telescopes are trying to do with are collecting data on like, literally, you know, hundreds of billions of objects in space. So yeah, so we are collecting massive quantities of data. And the reason ground based makes a difference, because you have all the wired internet, right? You don't have wires into satellites orbiting around Earth, right. So the transmission volume, 5g networks on Earth are very localized. So the best way you get high speed data on Earth is through wires, the internet, the wired internet, okay, so so we can do that from the ground, we can collect massive quantities, but so one of the satellites in space, and it's not just astronomy, its weather satellites, its remote sensing satellites, any kind of communication satellites, they're getting better and stronger over the years, like I said, but in the early days, it was very limited.
Rob Stevenson 7:41
If you'll pardon the pun, do you miss working in that space?
Kirk Borne 7:44
The space? Some days I do. I mean, certainly ever since I was a child, I wanted to be an astronomer. I mean, I got a colorful picture book of astronomy when I was nine years old. And they just fell in love with it. I mean, it was It wasn't anything technical, but it was just, I just said, Oh, I just want to know how this works. I want to understand this. Unfortunately, I had the aptitudes for math and physics and science. And we actually had a programming capability in my high school. And this is they were talking like 50 years ago, I'm dating myself now. But we actually had it link up with a computer at the local university through our high school, which was very rare in those days. And so all those things that inspire me, then I can actually fulfill in a career as an astronomer, but what I do now, if in some sense, like it taps into that other dimension that I mentioned about what really gives me passion in life, and then that's teaching and training people and seeing the look on people's face when I can teach them something that they didn't really fully understand before. And so that for me, it's almost like fulfilling the second half of what I always wanted to do.
Rob Stevenson 8:44
I see. And you've continued that you continue education, even though you left George Mason a few years back, but you have this new project that you are working on, which is sort of a more egalitarian approach to AI education. Is that fair to say?
Kirk Borne 8:56
Yes, we call it blue collar AI. So in fact, after I left George Mason, I've even had adjunct faculty positions at a couple other universities, what adjunct means is not a full time faculty and just paid to teach a class. Okay, so it's not really a salary job, you just get a stipend for teaching a class. So I kept my foot in the water, so to speak. But this blue collar AI thing is something that was generated from the data prime company, the startup where I've been working for the last couple of years. So data prime is worked with the Miami Dade College in Miami. It's probably one of the largest community colleges in America. I don't remember the number it's like 80,000 students or some massively large number of students all from the local area. I ask you a counselor a low to mid income area, and so that the community that they're serving are not necessarily you know, high income people. They're not necessarily people who are going to go to university or Harvard, those kinds of people. They're just your standard people who are maintaining the local economy and all walks of life all types of different jobs. And since data is everywhere, digital transformation is everywhere. And AI is enabling and empowering people to do business. It's better and faster, we figured, well, even these people who may not think that they want to learn the math and the science don't necessarily need to learn the math and the science, but they need to know what AI is, what automation is. And primarily the one thing that I like to tell people, it's helping people get over the fear of it. When there's so much fear of it, I mean, not just in those people, it's even in executives and businesses, they think this AI is going to take over their business, it's going to remove their decision making authority, the workers in the business are afraid is going to take their jobs away, the middle managers, our general managers have interfaith for a while now, because as soon as data scientists came along, and they discovered all these insights from data, the data scientists actually have a seat at the table with the executives now, in terms of like, what is the data informing us about our market about our business, better consumers? So the middle managers said, Where's my role here? I mean, did they, the people who report to me are now talking directly to my boss, where do I fit in? So there's like fear all over the place. And so I think I want to reduce that fear by what I teach people. And that is basically illiteracy about what it is how it amplifies and assist you in your work, it helps automate certain aspects of your work, but it's not really taking your work away, it's just making it better and more efficient, or more effective. I like to tell people, there's a difference between efficiency and effectiveness. And both of those are sped up right efficiency is the speed at which you get something done. And effective means the amount that you can get done.
Rob Stevenson 11:22
So you have educated on this topic. Now, at both ends of the academic spectrum, let's call it right, if you look at like the individuals, maybe pursuing PhDs, fantastically, highly trained, maybe have already, like worked in private sector, then come back to school. And now blue collar AI, like these are not technical people traditionally, right? What made you want to do both what made you want to move away from like the PhD type students, and work with folks who don't have as much access to tech?
Kirk Borne 11:51
Well, again, it goes back to that sort of deep passion I've always had, even when I was in high school, elementary school, not so much elementary High School. Even in college, I was tutoring and mentoring people who are maybe challenged a little bit, I remember where I had a short term, two year position as an instructor at University of Michigan after I got my PhD. And I was teaching introductory astronomy. And occasionally, there would be like members of the football team, who needed a little extra assistance, and passing their science course. And that's fine. And so I was occasionally drafted, so to speak up by the football department, the athletic department, I should say, to mentor some of these young men. In fact, even young women, I mean, people on the baseball team softball, team basketball, swimming, Team pitch, things like that. And so and so I found pleasure in helping people understand complicated things. I mean, it's a victory. It's like a win win. That's what I liked about it. I mean, I felt like I succeeded in achieving something. And the person who finally got it understood the concept, this complex concept, it was very pleased with themselves, okay, so there's a certain sort of mutual value that's exchanged there. When you're working with graduate students, it's somewhat different, because you're trying to push them along, to help them to get to that point where they can be an independent researcher, and come up with their own ideas and do their own research. And you work with them and mentor them and move them in that direction. But at the end of the day, real victory is their victory. I mean, I contribute, but they're the real victory is their victory when they achieve that doctorate degree. That's reassuring to me and pleasing to Me. And another way that I was able to help this person get there, but it's really their work that's being celebrated. So there's a variety of different kinds of motivations that go on when you're, when you're doing things like this.
Rob Stevenson 13:30
What are your hopes for the students in the blue collar AI course? At the end of the course? How have they changed?
Kirk Borne 13:36
Well, for one thing, it's really about becoming AI literate for most of all, not necessarily AI fluent. I mean, I think there's a sort of a difference between the way we use that terminology, and then data literacy versus data fluency. So if I can refer to data for a second, then you can sort of easily see what I mean by AI, literacy and AI fluency. So in data literacy, it starts with people recognizing data when they see it. Okay. So for example, when you are using your smartphone, almost everybody in the world uses a smartphone, I see my grandkids on all day long, they're on their iPads and their smartphones, I can't get them off. So everyone is generating data, okay? When you when you visit a site, when you click on something, when you search for something, you're generating data, on the other end, they they know what your interests are, what your intentions are, what kind of things you're doing. So they're using your data, they're creating value from data, they're making money from data, and I tell people, you are a generator and user of data when you're on that device. So why don't you get involved with this economy? Why don't you make your career I make money out of this. So AI literacy is the same sort of thing. You recognize where AI is and see what it is. And it's not a scary thing, any more than data is not a scary thing. I think that some people come to me and think data is a four letter word, and I teach them that it's not a form of a word. And so some people think of it as a scary thing, but I'm trying to teach them that it's not a scary thing. And so AI literacy, their fluency Are you going as the parallel to data literacy and fluency literacy starts with understanding what when you see it knowing what it is when you see it fluency is now you can work with it, you can actually apply it in your job, you can do something with it, you can create value from it either for yourself, your business, your boss, whatever. So AI fluency and AI literacy, it's sort of the same thing. So what So what we're looking for in this blue collar program is first to have people just aware of where AI exists. All right, I mean, so the easiest for people to grasp is actually not simple self driving car. Okay. Okay, so people can understand a self driving car. How's that car doing that? Well, it's got cameras, it's got sensors, it's got their different kinds of things that helps it understand where the road is going, what is the stop sign, say? What is the speed limit? Say? What does that school zone sign mean? I mean, slow down? How far away? Is that car in front of me? Is it moving faster than me slower than me or the same speed as me. So there's LiDAR, there's radar, there's just cameras. And so we start talking about all those data collections, and how the AI is learning how to read the sign to read the condition of the road, or the weather conditions, or read the sort of the location and speeds of the vehicles around you, etcetera, etcetera, etcetera. So it's a complex, very complex problem. But people can understand the concept of a self driving car, if I talk about a digital twin, which is a computer copy of a manufacturing plant, that gets a little hard for people to visualize what is it digital twin of a manufacturing plant? Advocate. So that's a little bit more complex. So we so we start with something where, which is also complex, but what people can understand us a car. And so so we started building this literacy around that is recognizing and understanding it, knowing what it is, knowing that it uses data, and what is it doing with it, it's finding patterns in data. If you're reading a stop sign or reading a speed limit sign you there's a pattern there, the digits on the sign, to a camera, it's just black and white pixels. It's just black and white pixels. But the AI algorithm can interpret that as this as a number. And that number refers to a speed limit, which to understand understands, it's not just a number, but what is the context and meaning of the number. It's not just a number, it's an it's a number with with some kind of meaning that's informing you about something. And that's what AI is all about is giving you insights that inform actions and decisions. So once people start grasping those, those ideas, that it's about informing actions, decisions that they can start to see, oh, this can actually help me do my job better, right? I can use this AI to help scan through my emails to find the most important emails, I can even use an optical scanner to scan through the mail that comes into my office is to identify which mail is coming from a customer which versus what, which one is probably just some marketing, email, mail or something. And so these kinds of things help people in their jobs, regardless of their jobs. I mean, so there's, there's different types of ways that automation, and digital transformation is changing a lot of jobs in the world, I mean, not not just the high end, professional job, so to speak. But the blue collar job, and by which I mean people that don't necessarily need a four year college degree or an advanced degree.
Rob Stevenson 17:55
So these are the people that have traditionally been kind of left out, right of the development of this tech, because this tech is being developed in such like highfalutin places, right in very forward thinking businesses with lots of money and in research facilities and academia, etc. What do you think is the risk of these individuals being left out of the development of this tech?
Kirk Borne 18:18
I think the risk is similar to the previous industrial revolutions. Let me just put it that way. Alright. So people refer to what we're going through now is the fourth industrial revolution, right? So the first was the steam power revolution, right? So people went from sort of manual labor on the farm, to having a steam engine, power, different kinds of things. And then the second industrial revolution was when the electricity was introduced into industry, and certainly changed the way a lot of things were done. And then the third of this revolution was really the computer revolution. 50 something years ago, you know, essentially, back in the days, back to back in the early days of computing, someone, a very famous guy said he couldn't see the need for more than three computers in the world. And then the sort of famous guy called Bill Gates, once even the guy who invented the basic the PC, basically said he couldn't imagine anyone needing more than 640 kilobytes. Well, of course, I got a, I got a terabyte disk drive sitting right in front of me, right? Yeah. So so the computer revolution is different. I mean, it was yeah, it was definitely Industrial Revolution, the computerizing of all kinds of business operations, etc. And so the current industrial revolution is all it's all about the big data flows, the AI, the machine, learning the interconnectivity among different things. So think of something as simple as someone operating a cash register, right. So a cash register, basically, is doing computations on things that you buy, right? It adds it up, it calculates the tax, okay? And even understanding that kind of aspect of computing as relatively simple as you might think it is. It is a valuable thing. I remember hiring a person once, who had very little capability and what I would call traditional computer technology, when I was a manager at NASA, but This guy had a really good organization skills I could tell because he was a photographer. And he and he did a lot of field work in photography. And he organizes his photograph collections in very smart ways. And so this, this guy actually has an aptitude for understanding, sorting and arranging, and collecting things, which is what data is about organizing, collecting, sorting, finding the patterns in the day. And so it turned out, he was one of my most successful hires, I basically inherited at the lowest end of the salary scale that I was allowed by law to hire someone. And they ended up pretty much near the high end by the time he moved on. And in fact, he moved on to actually have a chief technology officer job at a university. I mean, I felt like that was quite a success story. For me, a guy who basically came in just looking for any kind of job, hired him as a data clerk, the lowest ranking job we had any guy moved up in his career to where he became an officer at a university. And that was one of my favorite, my best success stories in life. And this is the kind of thing we want, we want people just to feel comfortable doing things that people are really good at sorting me, kids, kindergarteners, really good at sorting things, recognizing patterns and things, being able to sort of manipulate things. And that brings value to business, if you can organize your customers into different customer segments, marketing segments, I've heard stories at conferences, people that were just basically clerks, so to speak. And again, they are not the technology people, they weren't the database people, they weren't the programmers, they weren't the PhD scientists. They saw some pattern in the customer data and actually introduced the idea to an upper manager who had the power and executive authority to do something and they did something and actually made a lot of money for the company. And it was just started with someone just saying, Hey, I have the power to do something. And so the way that started, actually, the way that works in the real world is you get empowered from the culture of the business. So there's different ways that I've heard this one is a culture. I've heard a culture of experimentation, right test to see, does this data really telling me that this customer will buy this product if I recommend it to him? I mean, so we have recommender engines, right? So people can test different things. I mean, there's a famous Casino in Las Vegas, which expresses this another way. So this so culture, experimentation is one way of expressing it. This casino calls it another way there of culture says, Test, or be fired? Well, there's an ultimatum to you, if you work there, test or be fired. So what are they testing? They're testing? What is it that the customer wants? Okay, think of a casino. I mean, people are are in there, they're going to shows they're playing the slot machines are playing the game tables, having drinks, do all kinds of things, right? So find out what the customer likes, what's what's going to keep them there, test or be fine data
Rob Stevenson 22:33
everywhere else. So right, you're like in this very this controlled environment, you're recording everything,
Kirk Borne 22:37
you talk about a lot of data, there's a there's a place ahead. And the third way that I that I've also heard, and I found very successful, two way to describe this concept of experimentation or test to be fired. As a CEO of an airline, which told his people, if you see something, say something, I mean, that was his that was his body. So if you see something in the day as something about the customer, something that customers like something customers are doing, just say something, whether it's your job or not, okay, no matter who you are in the business, if you see something, say something. And so he empowered people who were not the database engineer, they weren't the data scientists. They weren't the AI professional, they were just someone who understood what he meant, right? So if I see a pattern, I see a trend, I see something in the data, I should be empowered to say something. And he told stories, in a conference keynote that I heard once from this guy, and they really inspired me, I mean, the fact that this guy was able to feel comfortable as the CEO of the business to empower all of his people, no matter who they are, to speak up when they see patterns and data that can be informative for insights. And so that's what we're trying to achieve with this blue collar AI is for people to feel comfortable with it, to see where it can bring benefits to their business, even if they aren't the one to implement it as if they're doing something in their job, they can go to their boss and say, Hey, there's a beware of this new automation tool that can help me process these invoices are processed these emails or whatever I mean, they being aware of a knowledgeable of the ability of automation and AI to improve their business, even if they're not the one to implement it. That's what was the value of some of the stories that I've heard from these executives, is the people who suggested these things weren't the ones who implemented it. I mean, they didn't know how to implement it, but they were able to see something and say something, and that was and that's really made a huge difference, both in that business itself. And probably in the life of those employees. I'm sure
Rob Stevenson 24:23
this approach to education, it reminds me a little bit of Carl Sagan or Richard Fineman, these these individuals who really understood that if you couldn't understand something really simply then probably you didn't understand it well enough yourself.
Kirk Borne 24:33
Yeah, there's a famous expression. If you can't explain it to your grandmother, you really can't usually don't understand it. Or if you can't explain it to a third grader, you really don't understand it.
Rob Stevenson 24:41
You actually you worked with Richard Fineman briefly, yes?
Kirk Borne 24:45
I wouldn't say worked. I took the class from him. And I actually got talked with him separately on some occasions when I was in grad school at Caltech because he was a professor there. And it was a book that he wrote called, surely you're joking, Mr. Fineman. And I think and I think there's even a sequel Part Two, I read that book. And after I left graduate school, I knew that I knew it personally in grad school. I read this book afterwards, it was a perfect characterization of his this guy. He was just a jokester, a brilliant human being. And he knew how to play jokes on people. And then he primarily play jokes on people who were like arrogant. People who are too smart for their britches says, Hey, we use as we used to say. So my funny story with him was, I was working in the Physics Building on my PhD and I said, a physics because astronomy building was separate in any way, but the computer at supercomputer I was using was there. And one day the fire alarm goes off. And I don't know if it's just a test or a real fire. So I leave the building and I go outside the building, and I'm sitting out on the little brick bench that's in front of the building, waiting for the fire trucks to come in here. The fire trucks down the road are on their way, when you just see what happens. And while I'm sitting there, I see Richard firemen down the street running towards the building. Okay, so it's the end of the day. So he was obviously running from the faculty parking lot. He was probably leaving for the day. And he's running towards the building. And he's and he sits down next to me. I was just a young guy, I was literally just a totally young students. And here's the like world famous guy, Nobel Prize winner worked on the Manhattan Project sitting talking to me, okay. And he sits down totally out of breath. He's huffing and puffing, trying to catch his breath. And he says to me, he says, I really wanted to see the fire trucks. When I was going home, I want to see the
Rob Stevenson 26:23
12 year old boy, I can't miss the fire.
Kirk Borne 26:25
And that was just him. And that was just him. I mean, he just he sort of never lost that sort of childhood curiosity. I think that's the other thing I just tell people is, what are we all born with? We're all born with that curiosity. We're Natural Born scientists. I always tell people, children are natural born scientists, or unfortunately, our education system, drums that out of kids, right? They start asking a lot of questions, and it gets annoying. Why? Why? Why? It gets to be annoying. I get it. But on the other hand, we're all natural born scientists raw, naturally curious. And so people with this blue collar AI and even sort of blue collar data science, I would say is blue collar, machine learning whatever you want to call it. It's all about pattern recognition. Right? It's asking questions. Why is that? What is this? Why is it doing that? What does this mean? I mean, so we were naturally born asking questions. So when you see a pattern today a trend? Like why are people who are buying this product, also buying that product at the same time? Maybe we can create a marketing campaign around this. So now we have recommender engines everywhere, right? I mean, not just in products. I mean, even even if you if I go read a journal article recommends similar articles for me to read. So recommenders are all over the place. What does that that's an AI, right? So as soon as people realize what you naturally do, like I read this, or I see that or I buy this I can I can I say to myself, well, what else is like this? I like this. What else is like this? Well, that's what an AI is answering. It's answering that question for you.
Rob Stevenson 27:44
Yeah, and it's an important call out that whether you have the literacy or not, it's affecting you, right, it's action is being taken upon you or by the tools you use, your data is being collected and served back to you in ways the data collection piece I wanted to ask you about. Because for so long, people have been paying for Internet services with their data, basically, right? As long as you you know, click the terms of service button, I agree. And we collect this data on you. And now these things stay free, right? Do you foresee a time when people can kind of take the power back on their own data? When it's like, look, I expect to be compensated for the use of my data? Because I know that these companies are taking it selling it to advertisers or taking it and using it to design better products. Is that a realistic outcome for personal data?
Kirk Borne 28:32
Yes, it absolutely is. That's one of the definitions of what people call web three. So just as a clarification, back in the old days, we talked about Web 3.0, or internet 3.0. And that was more about the semantic web, the semantic ontology based web. And if you don't know what that is, that's fine. However, you can look it up if you want. But web three is a different concept, which is you take now take ownership of your own data. And there are companies out there which are moving in this direction. I know a company called Jenny CO. So Jenny, as in the name Jenny, Jenny CO, is doing that with healthcare data. So you basically give companies permission to use your health data, whether it's designing products or whatever, and you get some kind of small compensation for this. Okay, now ad companies have been surviving, so to speak through Cookies, cookies are going to go away that which are basically these these trackers on the internet, knowing where you are and where you're going. So that that's your data. That's, that's not being monetized by these companies. Right now, I know that you like this, I'm going to recommend products to you like this, essentially targeted marketing. So the more targeted your marketing is, the more chances are, you're going to get someone to buy the product, right? And instead of just sending the ad to just every random person on earth, and so the more you can get control and ownership of that data, then you can get some share of the data. Because even though it may not be you that buys the product, you are of a certain demographic and your educational background, your buyer persona, yes, your profile, whatever that profile might be in and I don't mean that in a negative way. Like things that we shouldn't be basing decisions on, but things like that, what's your education level kind of technical field do you work in, you know, what part of the country you live in, that makes a difference, right? I'm not gonna, I'm not going to sell snowshoes to someone in Miami. Okay? But anyway, if you're a sports enthusiast, then maybe you would want to buy sports products, okay, so these types of things, they learn from you, and then offer products to other people. And if they if they learn that from your data, if you get some small percentage of that data monetization that they're acquiring, or monetizing your data, you get small percentage of that, then that's basically web three. And so I see this coming. I mean, there's companies, like I said, just company called Gini Co. And so other companies that are working in this direction, there's this company called alchemy Exchange, which is doing this with with advertising, programmatic advertising data. And so I just I look around and see these things starting to happen. And I think there is, of course, this move now that basically cookies will be outlawed starting this summer, next summer, I mean, so companies can no longer use those tracking cookies to follow you around the internet and know who you are. And what you're doing all the time. was sort of scary that we ever allowed that in the first place. I mean, I'm originally cookies, were just basically session trackers, right. So if I'm on this web page, and I spend all this time doing something, searching through products, and then I have to go away for a while. And then I come back to the computer, I have to start from scratch all that work I did while the session cookie kept track of where I was. So I can just resume
Rob Stevenson 31:21
right, right, keep you logged into places. If you've ever deleted your cookies accidentally, it's a painful thing to keep browsing after. Yeah,
Kirk Borne 31:27
so we use these at NASA back in the day when people were browsing our datasets. And they were looking through all kinds of data to find specific data on specific things they were interested in, in space. We wanted to keep track of that. So they wouldn't have to redo all those searches all over again, the next time they come to our site. I mean, it's just a terrible efficiency waste time waster, okay to have to do it all over again. But they they wouldn't be on session cookies to basically these tracking cookies. And they now know when you go to Amazon, or you go to eBay, or you go to msnbc.com, it's still you. Okay, so this person is looking at sports scores, they're buying sports products, they just bought some sports shoes, and you're looking at reviews of basketballs. Okay, we now know a lot about you from all these different websites you visited. So those kinds of things are getting deprecated, as it were, like you would say that is used a lot less than eventually made illegal. That was kind of tracking cookie. So there's got to be a way now that companies can continue to get value of understanding who you are. And because I don't thoroughly object to these kinds of things, I'd rather have things marketed to me that I mentioned it and not just random stuff. Right?
Rob Stevenson 32:34
Right. And that's like, No one hates an irrelevant ad basically, right? Because like, Okay, I'm looking for this anyway. But it's not always as mundane as I live in Miami, why am I getting skis? It's like, oh, I have the zip code. Why am I being denied a bank? Loan? Right. That is the that is the the challenge. That is the scary part of it. Right? So is that more why the government needs to be involved? It can't just be Oh, people's data is being used to send them better sporting goods? Like what is the government's incentive be involved here?
Kirk Borne 33:01
I think it's all of the above, it is such a huge push towards data privacy and, and going after abusers of data privacy. And I think in some sense, more and more people are realizing that your health data, your consumer data, your purchase data, these things are their private data, so to speak. So there are definitely the ethical issues of using something like yours, for example, your zip code, your your gender, or your race or something to deny you alone. I mean, those things are illegal. Obviously. It's not necessarily illegal that you're offering me skis, when I live in North Dakota, having this sort of obvious, I might need skis, or snowshoes if I live in North Dakota. But at the same time, like you say, it feels kind of creepy. Sometimes.
Rob Stevenson 33:42
It seems like at least in the United States, it kind of stems back to this like expectation of privacy, right? Almost like a constitutional expectation. And like you have your own kingdom, you have your own like home and your own space, and then like people can't trespass on it. And you know, like the police can't come in unless they have a warrant that kind of attitude. Do you think this is just the extension of that into the digital realm?
Kirk Borne 34:00
Yes. And I think actually, Europe has really taken the lead on the general data privacy regulations to GDPR. California has really taken a big stand on the data privacy. So it is a worldwide phenomenon. And certainly there's a lot of strength in the Western countries, not not so much elsewhere. I can say it that way. I mean, so I think we have a lot more concern about the ethical uses of AI and data in some countries than in other countries. So I'll just leave it at that. So other countries exploits. I mean, we know all about different kinds of hackers and different regimes and parts of the world who are perfectly fine hacking into your systems and stealing your data and holding it ransom. I mean, yeah, there's scammers and spammers everywhere, but I think there's almost like a government sponsored state sponsored behaviors of that sort. It's in some places in the world. And so we are definitely much more conscious of doing the right thing there. Of course, there will always be like I said, there's always gonna be criminals. We're gonna do the wrong thing. But I mean, in general, we have expectations that the right thing will be done.
Rob Stevenson 34:59
I'll correct I feel like we could keep going all day. There's so much more I want to speak with you about I guess I'll just have to have you back on at some point. But for now I just say thank you so much for being with me and sharing your journey experience with me. I'd love chatting with you today.
Kirk Borne 35:11
Thank you Rob. It was great.
Rob Stevenson 35:14
How AI happens is brought to you by sama. Sama provides accurate data for ambitious AI specializing in image video and sensor data and notation and validation for machine learning algorithms in industries such as transportation, retail, ecommerce, media, med tech, robotics and agriculture. More information, head to somma.com