 Buenos dias. Me llamo Selena, mucho gusto. That is brought to you by 60 days of Duolingo as much as I can bring you. So I'm not prepared for this trip at all, but what I am prepared is to talk a little bit about what I believe is the foundation of advancing the data community, which is data literacy. And so throughout the talk, I will start drawing comparisons with the development of literacy, which is the ability to read and write, more particularly in the European American history of it, and pull lessons from it so that we can build better data literacy programs alongside with our data initiatives so that we get the kind of the engagement that we want. Because ultimately the foundation of open data initiatives, or if you work in an organization that tries to drive value and tries to have people to use data, is we believe that it's through democratization, collaboration efforts, and transparency that we can drive empowerment, better individual empowerment, and progress for the community that you work in. And by community, I'm going to be saying in a very generic sense, whether that you work towards delivering to the general public, or if you work in an organization, or with your team, or your co-workers. And we've actually seen that comparison before, in the sense that increased literacy was a significant contributor to the Industrial Revolution. Right before the on start of Industrial Revolution, which is the process where manufacturing has changed from handmade goods to machinery starting in the UK, you can see that there is actually a significant and steady increase in literacy of the population. And there wasn't a single moment or single pivotal moment that caused the increase of literacy or the Industrial Revolution. It was actually and taking inspiration from yesterday's keynote, it's kind of like a hive mind situation there where progress were made by individuals, contribution, and essentially standing on each other's shoulder. And what does that look like in our current data industry now? We actually see the same mirroring of that, because with the increase in computational efficiencies, it really reduced the time it takes and the cost it takes to do product to product, produce, and experiment on on different things, which started allowing us to have more and more exciting technologies. Is this the working? I think so. And so a lot of our focus and a lot of our talk internally in my organization or in a community that I'm hearing is about the exciting tech, and that's very exciting. But we I feel like we don't talk about data literacy as much. So taking Google Trends, for example, looking for the term of data literacy, we actually only start seeing a slight improvement in a past couple years. I'm good. Okay. Oh, okay. All right. It's my natural sounds too strong. So we start seeing a little bit more of it in towards the past couple years. But when we compare that with the terms of like data science and open data, can't even barely see the blue line anymore. And that was fascinating to me in a sense that in my understanding of it, the data science, the foundation of data science, the progress of it and the effectiveness of it is founded on literacy, data literacy across the whole. But yet we don't talk as much about it. And how did that hit me? Well, it hit me at work. So what's the problem of that? We talk about the tech, we talk about the methods, we talk about all those exciting things. Because as data practitioners, we are expected to bring value and transformation with data. We want to empower people, we want to show whether that's analysis or curated data sets, whatever you work in, we want to do that. And the moment I start talking about like curation, acquisition, governance, models, ethics and usability, I lose people a little bit. Or not a little bit. I lose people. I lose people. And I keep hitting this brick wall because essentially I get the sentiment of like Selena, why are you trying to make things so complicated? I just want to count things. And I believe people in this room would feel, would understand the difficulty and the nuance in that simple sentence. But with majority of the people that I work with, and most of these individuals are frontline operational individuals or administrative individuals, that's all they want to do. So my conversation went from about governance models to why we shouldn't use that spreadsheet in that person's laptop that was five years ago and in 16 versions of it now. And no, I can't explain to you what that cell means. Neither can anyone right now. And so that made me start to realize that we can really only go as far as our community's overall level of data literacy. Whatever it has is as a direct influence on our ability to impact and empower and make progress for the community that we work in. Because at best, we start creating things that no one uses. If I have a perfect library, if I have a perfect data catalog, I'm still competing with somebody's spreadsheet in somebody's laptop. And I have to figure out why. And at worst, is people start misusing what we create. People don't understand and we, and isn't actually, we're not actually doing it for the empowerment of it anymore. We're actually hindering it a little bit. So how can we improve data literacy? First of all, we can learn a little bit from how literacy itself, the ability to read and write, is developed. And it's actually a very hard concept in the history of it. It was very challenging to quantify and measure what literacy even means in the development of it. Which ironically, as I was doing my research for this, one author would say, it's an issue of counting. So counting, once again, is very hard. And literacy itself as a terminology and even as a word is actually a recent term. Does anyone here want to venture out a guess or no? What was used to gauge literacy as a proxy metrics in the UK? Signatures in marriage registries. So even in reports that were written out, it wasn't about literacy of the population. It was the amount of people that are able to sign their own names. And so that, that itself, in itself, is actually, it helped provide a little bit of insight onto the thought of what literacy even means. And in fact, literacy from the development of it actually splits between reading and writing separately. They were taught separately. They were seen as something completely separate because it came from necessity in the sense that reading wasn't about comprehension. Reading was about spelling words out and writing was just to copy. It wasn't about writing text and creation. It was just to copy as describes as needed. And so much that the combination of reading and writing as a single thing, which feels so intuitive today, was actually considered as an innovation back then because it was actually came from the necessity of like, well, it's just easier to teach it all at once. And that is what we today call basic literacy. But as the development of literacy continues to move forward in a modern society, there's something called the functional literacy. Functional literacy means that if an individual can effectively function in their community that they are in and can continue to develop in their own and the community's development further. And that started making me thinking a little bit about, oh, we don't actually have to tackle literacy as a big, big hairy thing. We can start tackling it at a different perspective, which is how do we start determining what is the functional data literacy level required for the community that you work in, for the community that you serve? Maybe in the sense that we are trying to get everyone to be statistician or basic statistics, but maybe that's not actually the case too. Maybe it's about collecting data or even seeing things as data. So determining what that level is and start designing programs or alongside with whatever initiatives you are to gear towards that rather than trying to be like, data literacy, how do we get there? Literacy efforts in the past were also a function of market demand. What that means was that there was a need for, as I said earlier, we needed people to copy more writing, so then we had to teach people to write. During the Industrial Revolution, for example, with machineries, we needed people to learn certain types of other skills. For example, typists, you know, that's why or postmen to deliver mail. So they need to read addresses and deliver them. But with market demand, when we have up, we also have down. And what we really see is that initially in the earlier stage of the Industrial Revolution, we actually saw a stagnation in the literacy development in UK. And it wasn't until much later on that with mandatory protection from the government, with policies, that we have to send people to send children to school that we start seeing an uptake of that again. And a big problem with that was because the cost of sending children to school was high. And I don't mean the literal cost of it, but it was changing perception of workforce labor perspective. And so as we have to think about how do we have data literacy program, we have to think about how do we make it accessible and affordable in whatever definition it means. And the best way I could think about it sometimes is that what is the best method to have it as going hand in hand with the initiatives you have, so that it's not an additional thing or it's not a separate thing, but it's part of it. And with technological changes, there's always a shift in the demand of workforce, in the skill sets that's needed. And then that also has shapes and shifts what training and education we have. And a hot topic even today is the idea of de-skilling. The process of de-skilling, which is the removing of certain skilled individuals for automations and so that, and replacement of unskilled workers. And we see slight mirroring of that in the sense that with the massive need for data labelers in different industries, we're actually starting this emerging in different markets. But one of the theories in conversation about what the impact of de-skilling is that shifts in skill and demand is not, it doesn't inherently have to be a bad thing. It can be skill reducing, which is one side of the argument, but it can be skill inducing. How do we create better skills? And that's where I want to be thinking that we have to be very proactive in our literacy efforts so that we make sure that people are equipped to do more, not less. We need to induce more skills so we level the overall playing field for everyone and we're leaving nobody behind in our efforts. And my last bit here is that in the past, we know that innovation that led to increased productivity really drove great progress. Electricity, lights, paper. It's hard to not talk about chat GBT, which I thought it was very fascinating that when I sent this abstract, it wasn't even a thing. And now everyone's talking about it. That was the slide from before. Can you imagine what looks like with chat GBT? And I'm not here to talk about the technical bits of chat GBP or whatever. Like that's not, that's not my interest. My biggest focal point about this was I'm fascinated by the uptake of chat GBT, not within our community, but in the outside community with the general public. I have friends who self-proclaim, never touch data, don't use data, don't talk to me about data because I annoy them. And they all use, suddenly use chat GBP and they're talking to me about chat GBT. My friend made a rap song about her dog on chat GBT and she's like, oh, hey, look, we can do these things. And also shout out to the dude from Argentina. And so I started wrapping my head around, why is this the case? Why is everyone who says I don't use data or everyone in my life who's like, I don't use data, suddenly using something that I see as a data product, as a foundational data product. And it's because the accessibility and usability of it, it's so available, it's so accessible. And when I say usability, I meant that people wanted to use it, they can use it and it's so easy to use. It was a sentence, you know, make a rap song about this dog. And so when we think about literacy programs or we think about the initiatives, the data initiatives that we drive forward, we have, it's changed my mind set a little bit, is that, oh, I have to bring immediately value back to my user. I have to design things that have to bring immediate value back to the user. Because when I used to think about this before, I find that there's actually one missing component in usability in how I think about data literacy. It's because I wanted people to understand its use, I wanted people to understand its implications, I want to talk about all the all the big hairy things behind it. Nobody wanted to be in that conversation. And, but that to me was, is the foundation of data literacy, right? And so with say, as an example of chat DPD, I was suddenly able to trap a lot of my friends in conversations about this, whereas before, I would completely lose them. And so that's where I was thinking, if we want to build better literacy programs, we need to leverage technological changes that's happening to advance data literacy. We need to meet them at where people are at and start from there rather than like, okay, let's blank state, let's do better and how do we do better. And so I'll leave with the last thought of what is that chat GBT, what is that catalyst that we can, we have a choice to artificially add this in now, right? As part of the initiative. What is that catalyst in your, in your initiatives? And that's the end of my talk. Thank you. Gracias. Thank you so much. We now have about five minutes for Q&A. What questions do we have? Raise your hand and I'll get you the mic. All right. Thank you. Thank you for that. Yes, I'm loud enough. So sort of my question about literacy there is it's a topic that we hear all the time and that people think important. And I'm curious your thoughts on how, how we teach literacy might enable some people to understand things, but then are we teaching in literacy in such a way that really limits their understanding because we focus on some aspects like people are talking about, you know, if you use chat GBT to learn how to engineer prompts and you're like forcing people into a system that might end up being very narrow because they're very literate in a limited way versus being broadly literate. So do you have thoughts on, you know, very narrow literacy versus broad? That's what I'm wondering what your perspective is. My thought on that, and this is my current theory, is that I'm, I think it's a reverse of it because before with literacy we started off from the ground up. We've talked people to write and read and everything. And until that becomes a program in our education system, which I have no control over as an organization or as community, what we can do. And that's where I want to flip it up a little bit and start talking about the functional literacy part, right? It's actually true that if we could be very narrow, narrowing on, like, let's say just train prompts, you know, how can we train prompts? But the thing is that's not the only thing we can teach, right? Because after training prompts, we can start having conversation about, hey, why do you think these prompts are like, how do we shift that understanding? Versus if we start from the get go about, okay, what is data? Immediately people like, I just want to count. I just want to count. Like, I have a spreadsheet, I can count that. You know, don't talk to me, right? And so the way I'm thinking about it is kind of a reverse approach of that. Other questions? Yeah, just a quick question. What's your day job and how does this connect to it? So I work in the public sector. I work for the Canadian Provincial Government in BC. And most of my job is, it's, I mean, I started as a data scientist, I don't even really know what title is anymore, but most of my job really is trying to get enablement. People are using data, so leveraging our internal data sources. Because most, again, most of my battle, I could say I'm a data scientist, but honestly I'm trying to talk about exposed spreadsheets with people and be like, maybe we shouldn't use that, you know. So it's a complicated, it's a complicated topic for me. Any other question? For the presentation. I'd like to ask, what do you think that data visualization has a role in getting people interested enough to get in data literacy? I think data visualization or data results are very impactful initially, but the challenge I personally experience is that people only get the visualization at the very end of it, which they get what they want and they kind of just move away, right? And so that's where we come into problems about visualizations that like, you know, like chunk it at y-axis, for example, like and but then with people's get the point for it and they kind of move away from there. So I think it takes more dialogues than, but I take more dialogue than visualization itself, but it's a great primer to do that. Thank you. So I'm curious from your example of, you know, identifying what is the chat GPT of your initiatives. Do you have an example of that within your own context? Have you identified any chat GPTs of your sector yet? That's a hard one. That's why chat GPT comes so rarely. My current thought and strategy internally is identify what are big pain points that are people are using. Why was chat GPT, when I talked to my friends and family that uses it, why was it in as a big good thing? Even though they know sometimes it lies, well they say it lies, even though sometimes it's erroneous in the result, but they still use it. And so what is the biggest pain point of your organization or the community that you work in? Again, I said I work in the government, so I can't tackle the entire government. I can't tackle my entire department even, but with specific stakeholders that I work with specific pockets of them, what is their pain point and driving from, so kind of like a design centric approach to move forward that. Thank you so much Selena and everyone for joining. Let's give her another round of applause.