Alexis
Hi, everyone. Welcome to another episode of Data Dave Dives Deeper. My name is Alexis, your host of this podcast, as well as your favorite podcast, Talk Tech with Data Dave. I am here today with my dear friend, Data Dave, as always. And we have a special guest, Ms. Karen Lopez, here to join us to talk about a lot of really fun things, of course, Karen’s story and maybe a little bit of AI. So really excited. Before we get started, I always like to remind our listeners to submit questions to Data Dave by emailing us at talktech@d3clarity.com or by reaching out to us on the D3Clarity website. If you would be interested in joining us for one of these data dives deeper episodes, you can reach out to us at that very same email. Morning, Dave. How are you today?
Data Dave
I’m very well, Alexis. How are you? I hope everything’s going well.
Alexis
I’m excited. Yes, I’m pumped. I’m pumped to talk to Karen today. We met Karen at DGIQ, and I saw her presentation title was “Data ethics and responsible AI”, and I was like, “Oh, my gosh, we should have a podcast about that.” So, Karen…
Karen Lopez
Hi, nice to meet you. I’m so happy to be here. I’m finally cooling off here a bit in Toronto, and I’m happy to talk anything. Data Dave and I chatted for quite a bit. We probably chatted way much longer than we thought because we both love talking about data.
Data Dave
Yes. So, Karen, you’re a data evangelist. What does that mean?
Karen Lopez
It means that I have my own company, so I have a wealth of titles, and I have sort of landed on that one, which is typically what I do and what I work on. And also, the title I’m kind of known by is “Data Chick” because that’s my handle on most social media accounts. And that has a long story behind it. It was a derogatory term used for women on a team during the Reagan years because now I’ve aged, dated myself because we didn’t have names. We were just one of the data chicks. Like, having a name like Karen and Lopez means if you didn’t join on the first day, you can’t grab your own name as the title that you can use in a lot of places. So, I just landed on Data Chick, and over the years, I love it because it kind of reflects a little bit of the snark in my life, a little bit of, we work on serious things sometimes, but we also need levity because it kind of works out really well.
Data Dave
No, excellent. Excellent. So, yes, I am one of the founders of D3Clarity. So, I’m lucky enough to have my own name, and then Alexis comes along and changes it anyway.
Alexis
Yeah, that’s really the story of how Dave became Data Dave. My partner in crime, our marketing person, and I were like, “Let’s start a podcast! Let’s call it Talk Tech with Data Dave.” And then I forced Dave to become Data Dave and took over his social media, where I was like, “Everyone will know that he is Data Dave. That’s my goal!”
Karen Lopez
It’s a great name, it’s a great brand. If my name started with D, I’d probably be Data Diva or whatever.
Alexis
Yes. You’re the founder of your own company, right? It’s just you and your cats, is that what you said?
Karen Lopez
Yeah, it’s me and my cat. So, I founded the company. At times I’ve had employees or contractors that work with me on large projects, but as I get closer to retirement, I like not having to manage people. But as a project manager, I love it. What I hate is the whole employee management thing. It’s just not in me. So, I hate telling people, yes, you can take this day off, or no, you can’t, and those sort of things, but I love helping teams of people reach their goals. I don’t know what PMP says about project management, but that’s how I feel about project management.
Data Dave
I agree, and I resemble all those remarks. So, we formed D3Clarity, and I managed to not actually manage any people directly. I would rather lead than manage. So, I don’t manage.
Karen Lopez
Yeah, I’m going to steal that line. I’d rather lead than manage.
Data Dave
So, that’s good. And even when I was at IBM, I tended to do exactly that. When I had a big organization at IBM, I tended to lead rather than manage and then get other people to do the management.
Alexis
We could definitely have a long conversation about leadership versus management, but I don’t want to go there. If we want to go there some other time, we definitely can. But Karen, we love to start these podcasts with your story. How did you get into data?
Karen Lopez
Okay, so I tell people all the time, because usually this gets asked on a panel or something, that I’m the oddball out because I actually started being in data as a child. So, I tell people I was born this way. I was born to be a data check. My mom has photos of me as a toddler. Like, I can’t read, but I used to love and sit with the phone book and pretend to read it. And so not only did I like data, I had this affinity to what is structured data, because anything that was in a table.
A form, I used to pick up forms, you know, back before we had a lot of computers, there were a lot more forms that had to be filled out. And I would sit with my mom and ask her, “What does this mean? What is this?” And she’s a retired teacher, so she was tired at the end of the day from teaching, but she would sometimes explain to me what these data things are. And so, I went to university and I chose an information systems degree at Purdue. It was one of the early ones that had database systems. Again, we’re talking early eighties database systems in the curriculum. And I knew I was hooked once I started working with it. That was pre-relational systems and relational systems were new and definitely new at the university.
So, I’ve just never looked back. My whole career has been data information engineering, even all the way through.
Data Dave
So, I’m going to ask a quick sidebar question, because I got into databases early as well.
So, what is the difference between pre-SQL and no-SQL?
Karen Lopez
Ah, so pre-SQL is what I call… So SQL in the industry is typically used to mean relational database just because that’s the language, the query language that’s used to query them. So, I call the databases that existed before relational databases pre-SQL. So, I’m talking pre-SQL, SQL, and no-SQL.
So, no-SQL was created first by a group of people who were dissatisfied with relational databases. And so, they called it no-SQL. And even a lot of their logos and stuff had a big red circle with a line through SQL. And eventually, they learned to change it to “Not Only SQL”. But a lot of people still use it with the no word meaning no. One of the interesting stories, after all, this is a lot of no-SQL database systems, which I love. I love learning them. They’re designed usually for very specific use cases. And one of the early stories is enterprises just weren’t adopting them. And the main reason was all their developers, all their database administrators, their security people were having to learn brand new systems, brand new languages.
It wasn’t really until the no-SQL, I’ll say, vendors, but many of them came as open source projects. So, open source managers added SQL query languages on top of the non-relational underneath. And that allowed enterprises, which now we’re talking numbers of users, to get the data out of or put data in these newer databases using the languages and a lot of the tools that they were used to using. And it wasn’t until that happened, adding that relational-like language on top, did it really take off.
Data Dave
I think the reason I asked that question is because I’m always fascinated with people’s answers because I started working with databases long before there was SQL as well. So, I was writing bespoke APIs and network databases and various other things to access data in different structured forms. And then, the no SQL people suddenly came up. And then we loved SQL, right, because it was a structured standard query language that we could use against any number of databases. And while you’ve got your flavors and it’s limited.
Alexis
Is that what SQL stands for? Standard query language or structured query language?
Data Dave
Structured query language.
Alexis
Okay, I was wondering that the entire time. Sorry, Dave I cut you off.
Karen Lopez
And most of them forget the meanings of the letters. That’s right, all the time.
Data Dave
And then the no-SQL people came up and said, “Okay, no, we want you to access it using our bespoke API and we don’t want to put a query language on it. We don’t want to put a standard query language on it. We just want you to use our API.” And suddenly, I’m taken back to doing exactly the same thing as I was doing in the early eighties and saying, okay. And then they started putting a query language on it as well. I was kind of, okay, so now you. We’ve just gone completely full circle. So, we had a database that had a basic proprietary API. We added a structured query language so we could query it easily. Then we decided that wasn’t good enough, so we went back to a proprietary API and then we put a query language on top of that. Again.
Karen Lopez
Yes. The other thing that has come in, a bunch of old people talking, I mean, experienced people talking with each other about the good old days. The other thing that I think is hilarious is that, you know, when I first started taking these courses, it was all command line, like DoS prompt, like typing out letters, like messaging with your database as your program. That’s all it was. And then eventually we got applications that allowed us to interact with our tools and our databases. And the apps that end users use had a GUI interface, a graphical with buttons and drop downs and all that stuff. And it made it easier for people with no typing syntax skills to work with data. And that lasted decades. Now all the cool kids refuse to use a graphical interface or even data scientists and advanced analytics people all are typing either in a notebook, which is kind of a combination of a GUI, but still you talk to your data and structure your data by typing in words, and you have to know the correct syntax.
Now we’re in this kind of split where I get why we want coding to work with data. That’s so we can automate things. So for sure. And automation needs code. But it’s hilarious that people make fun of me that I still fire up an application to help manage a database or a set of files. You can just do that from the command line. I’m like, I’m old, I don’t want to type all this stuff. I can’t remember the exact syntax.
Data Dave
I’m the same way. And I’ll have a GUI and a command line at the same time pointing to the same database. Because sometimes one’s quicker, sometimes the other’s quicker. And I just use whatever makes sense.
Karen Lopez
Yeah.
Alexis
Does GUI stand for something or is it just a word that means a graphical way to look at something?
Karen Lopez
Graphical user interface. And I’ve also dated myself because the new term to use is UX. That stands for “User Experience” or UI for “User Interface”.
Data Dave
Okay, but the old term is GUI, spelled G-U-I. Yeah. And most of us like using “Graphical User Interface”.
Karen Lopez
And it’s fun to say. That’s why I still say it.
Data Dave
That’s right, exactly. It’s fun to say. It just brings this connotation to the mind.
Alexis
Oh yeah, it sounds a lot more fun than “User Experience” or “UX”. I’m like, “What?!”
Data Dave
Let’s just talk about data for a minute. Right.
So, what’s your perspective on data? What is data, Karen? What do you consider data?
Karen Lopez
I have a super wide definition, I’m not going to give a dictionary definition… I feel like I’m on an interview now… Data is anything that we collect and use to help make decisions. And I don’t care if it’s video or sound or emojis or for the experienced people, what were those things before emojis? Wingdings or something like that. That was a font…What’s that?
Data Dave
Emoticons?
Karen Lopez
Yeah, yeah. So, any of those like could be data, like even texture, how something feels. Data sonification, which is something I got to work on on a NASA project, is where you turn data into sound to see if you can hear differences in data that you might not be able to see in the raw data or see as a visualization. And that’s used a lot as well.
Data Dave
I’ve got a very simple definition that I use. Data is simply evidence of history. It’s the evidence of event. So it can be stored anyway. So, you can have any kind of evidence, but data is simply the evidence of an event. And it’s the fact that you have evidence of history that makes up data. You can, you can store it as sound, you can store it as pixels, you can store it as words. You can store it as anything. The data is simply the fact that you have evidence of history.
Karen Lopez
Interesting way of looking at it. I hadn’t thought about events, but, yeah, I guess my name exists because my parents gave it to me, and that’s the event.
Data Dave
So, it becomes that evidence of history. That is that stream of events that led to something. The reason I use it that way is because it falls into the scientific method, because then you start saying that the scientific method is a whole bunch of observation of history and observation of events, and then compiling those events into a prediction. Mathematics is now the language of prediction.
Karen Lopez
I have a friend who believes all predictions are bogus. So I don’t agree with him, but I agree with what you’re saying.
Data Dave
Yeah, so I agree with him, too, a little bit. Because mathematics is built on an imprecise model, right? Because all it is is a model, a mathematical model, that takes the history and allows me to predict another event. So, if I drop something on the floor, I can predict exactly when it’s going to hit the floor. Right. That is using a newtonian model of mathematics and a newtonian model of physics to predict an approximation, because the model isn’t precise. An approximation that is close enough for 99% of what we do as to when that object is going to hit the floor.
Alexis
9.8 meters/second squared.
Data Dave
Yeah, right.
Alexis
The one thing I remember from physics. Yeah.
Data Dave
So, that’s kind of the model that I use. And then that leads us into sort of AI, because if I collect all this data, myriads of data, which is all history, and then what is AI really trying to do? Is it’s trying to use that history to predict an outcome or to say something is good or something is bad.
Karen Lopez
So, I don’t think all data is for predictions. Like, some of our teammates refer to my obsession with data, but some data is really boring. My chart of accounts, with all my accounts that. I mean, as a business person, I do use it predictively, analytically. My bookkeepers and accountants, they’re 100% focused on filing my taxes. Like, they’re using that, the historical data, in order to pay my taxes. But they…. This is the first thing I learned about most bookkeepers and accountants I’ve worked with, is they don’t care that I want to use the data to help manage my business or anything.
Data Dave
No, I agree with that. You don’t always use data for prediction. I’m not saying all data is always used for prediction. Got it. Saying that data is collected and it can be used for prediction in some instances when you want to scientific type data. Because I use accounting type data, right. Because we all do forecasts and we all do various things. And part of the forecast is look at the history to decide whether I can use my history to forecast. I did so many hours of work last month. Therefore, I’m probably going to do that or slightly less this month because I worked too hard last month. No, I’m just kidding.
Karen Lopez
Yeah. There’s a bit of human factors that people forget in predictions.
Data Dave
Right. That’s the thinking there. And so, we get into, and you bring up a really interesting point, which is you said that your accountants and people collect data for your accountants so you can pay your taxes and they’re not doing any prediction. They are simply using it, a reflection of history to make sure something happens in and of itself, which is pay the right tax amount. What is interesting in that is then we get people who say, this data was collected in order to pay my taxes, but now I want to use it to predict what I’m going to do. And then they say, now the data is at fault because it’s dirty or not correct or not good enough. So, what’s your thoughts on that? Because you’ve changed the purpose of the data. You collected it for one purpose for which it was great.
Karen Lopez
Yep.
Data Dave
But now you’ve changed its purpose to something else for which it isn’t so great. And you’re blaming the data.
Karen Lopez
Absolutely. We see this all the time. And this kind of ties in both data literacy and data ethics. So, I’ve had many a discussion on project you’re trying to use data to. So, I’ll go back to, I like to talk about one of my clients a lot because not only are they gone, the whole industry is gone. The whole industry only lasted about twelve years too. So I’ll let people think for a minute.
But anyway, this industry had customers that they had to collect much more personally identifiable information, or PII, because it was a rental business. So rental businesses have to know much more about you, your credit history, a lot more contact information than most retailers need to know. But a lot of people,,, I know this is shocking… might lie to a retailer when they collect that information, like “What’s your zip code?”
And retailers are collecting zip codes because they’re trying to establish where their customers are coming from. But they don’t realize that there are people out there who’s like, no, I prefer not to give this, but their cashiers, their people at checkout are motivated or punished by how many zip codes they collect in the US. And therefore, the cashiers won’t push back. They’ll enter the store zip code or their own zip code or the White House zip code or 90210. Here in Canada, we have postal codes. So, a lot of people give “H-0-H-0-H-0”, which is actually the Santa Claus postal code because the North Pole is in Canada. What happens is that, yes, it was collected originally to just try to get a feel for how many customers walk to the store. Because especially postal codes are almost building level compared to the US, where zip codes are much broader. Whole city sometimes, or whether they came from afar. It really kills them in the US. When I get asked this, I’m like, “I don’t have a zip code.” And I get this big eye roll, and I’m like, I have a zip code. I have a postal code. Well, let’s try it. “Can you put letters in there?” Well, let’s see. No. “Okay, I need your zip code.”
“Sorry, don’t have one.” One cashier asked me, “Are you in prison?” And I’m like, “No, prisons have zip codes. And I wouldn’t be out shopping, so.”
But it was collected for that. But then someone sees that we have a zip code or a postal code tied to every transaction and they want to use this. What everyone in the business knows is a very low-quality piece of data because we incentivize the people collecting it to always collect. There’s no verification, and people lie. And so then they want to use it for something that’s much more business critical, like, go figure out the demographics of their customers based on zip codes, which you can do if you have any quality to that, if you got.
Data Dave
Quality in your data. My famous story on that is actually “Mickey Mouse”. When you’re asked an email address, everybody gives Mickey Mouse. Oh, yeah, right.
Karen Lopez
Or drop downs for a state a lot of people pick. I don’t know, is Alabama the first one? Because it’s the first drop thing and a drop-down is that people don’t realize that drop-downs alter your data, whether you intended it to or nothing.
Data Dave
My other favorite one, having worked in healthcare data and healthcare, is that the most common malady when people go to the hospital or the doctor is “broken bone”.
Karen Lopez
So, they’ll be seen first?
Data Dave
No, it’s just because it’s the top one on the list. So, when the registrar is entering, it’s just. It’s just that. So, you’ve got a list of a thousand, and the first categorized in the first HIPAA categorized malady is “broken bone”.
Karen Lopez
There you go. And it could also be skewed. Like, I was raised without healthcare. Kids in the family weren’t covered by our parents’ insurance because they were government workers. And so we really only went to the doctor if it was something serious. If you broke your toe, they didn’t take it to the doctor because one, they don’t do anything for a broken toes. And two, if all it resulted in was a bill and a recommendation to take aspirin at the time, then there was no point in going to doctor.
So, I always wondered if visits to doctors were influenced much more based on the socioeconomical, you know, basis of their local population.
Alexis
Dave, Karen, this has been fantastic. I love everything you guys are saying. I’m super happy that we got to talk about this, and maybe we can keep having this conversation in the future. But thank you again for joining us today. Karen, it has been an awesome time.
Data Dave
Yes, thank you.
Karen Lopez
Thank you.
