WIA 2021: Keynote Noelle Silver

This is so cool. Okay. Awesome. I just flew in yesterday. And first and foremost, I don't know how many of you in here, it's like your first time like out in the real world, meaning at a conference. Right? For yesterday, I may have gone through a couple of like, maybe I should just tell him I can only do it remotely. Mostly that was due to the fact that my 6-year-old started bawling in a way that I've never heard him bawl, because the last time I left, I was 4, and it was cool that mommy was going away. So what a difference a couple of years have made. And so I'll tell you a little bit about my journey so far. I guess I will start this clicker thing. Do I, can you see it? Oh, I can't see it, oh boy. So that is me. I have done, I've gone on a bit of a journey in my career. I brought my phone because I always like to take a selfie. We'll see if I get to it. But one of the things that, I do this for, and I always like to share with you the reason that I'm on this stage and how I got here. It's kind of humbling. This is my gaggle of people. You could see the 6-year-old I just left bawling. We are not actually in Africa there. We are at the zoo on a green screen, but it's cool, a cool picture. But those are my— that's my gaggle of humans that I have the honor of raising And I got involved in artificial intelligence specifically because 16 years ago my first son was born. His name is Max. He's there holding the heart. And he was born with Down syndrome. And I don't know if you know anyone with Down syndrome, but 15, 16 years ago, the geneticist sat me down and said, you probably want to think about adoption or— and I was a first-time mom— or an institute. You know, there's places places for people like this. Now, little did she know I was me. I was like, psh, I got this. But then I had someone come to me and tell me in that moment of craziness where you feel like everything was going well and then something crazy happens, and that happens to all of us in different ways. But this woman came to me and said, you get to choose. You get to choose how he experiences this world. And as a technologist, that meant something different probably than it meant to other women she told that to. And so I started literally on a rampage of who can I work for that can help me make his life better. And luckily that was relatively early in my career. So I started working, I went to IBM, I went to Red Hat, I'm back at Red Hat now, which is kinda nice. But I went to Red Hat, I went to VMware, I went to, Amazon, was a certified, a principal architect on AWS, and then Alexa was born. And here's the key to this is that I had to be looking my entire career for me to be ready to say yes when Alexa showed up. Because if you don't remember those days, 2013, Alexa came on the heels of a horrible product failure. Called the Fire Phone. Awesome product, by the way. I still own mine. It's fantastic. But horrible product failure. So no one wanted to be on the team. No one wanted to be part of something that actually didn't work. Does anyone remember Alexa in 2014? It didn't really work that well. It was cool, and that was good. But most— it wasn't even really data science at that time. It was a lot of regex and decision trees. And I helped write a lot of that. But what I found was, is that when I was on those teams and I started building, I realized this is gonna be huge for people like my son. In 2014, another very interesting thing happened that shaped my AI career, which is my dad. He's sitting right there in that green shirt, and he was hit by a car crossing a street in Seattle. I always tell people, like, being a pedestrian in Seattle is very dangerous. Very dangerous. Now, there's proof, he's a Marine, so he's fine, he's like up and at 'em, but it changed his ability to use technology. He cannot use a computer, won't use a smartphone. His interaction with life is through voice, whether that's voice on his phone, the Fire TV was a huge thing for us because now he could really navigate his entire world. I'll tell you, he was in an ICU bed for 6 months. 6 months, which is crazy at the age of like 68. He was in an ICU bed, he looked up at the sky, I brought in a beta device for Alexa, and I said, "Dad, you can ask it anything." And he was like, "Really?" And I'm like, "Yeah, anything, anything you wanna listen to, anything." And he remembered, even though he had a traumatic brain injury, he remembered the call sign for his high school favorite classical radio station in New York, and we were in Seattle. And he said those call names, numbers, whatever, right? WKRP, okay, I may have just dated myself, but whatever they are. But like, I literally got goosebumps. I was like, and what happened to him in that moment when it actually started playing that station, and thankfully that station was still playing classical, he relaxed for the first time since the accident. And I was like, this is the future. I will go all in. Here's the kicker, though. I have no experience, did not at that time have any experience in artificial intelligence. Lucky for me, the people at Amazon didn't care. They saw in me— and I love the idea of this for any of you who know someone who's interested in data science or interested in data engineering or any data profession who's like, but I'm an anthropologist, or I'm a geneticist, or I'm a whatever. I actually uncovered— Through my journey, and I just show you this to give you some— the end of my journey is that because I knew nothing, I asked different questions than everyone else in the room. I asked better questions now in hindsight because I didn't just say, "Oh, that's cool." I mean, I was on Jeff Bezos's pet project. We've all seen his latest, but his pet project back then, right? And it was awesome. I actually got an email from Jeff Bezos because my skill failed his brother-in-law. It wasn't pretty, but I still, I have that email saved. It was just a question mark, but still, it was very meaningful. But after that, I started to get excited about talking out loud about the journey that I went on. I built over 100 skills for Amazon Alexa. It just so happened nobody building for Alexa in 2014 cared about mindfulness or kindness. Now it's pretty common, but back then, no. So I got this really great opportunity to be first to market. At one point, I had a t-shirt that said 10% of Alexa, 'cause I had written 10% of the skills that were on that platform. Yeah, there was less than 1,000, but still, it was a thing. I own it. I'll teach you that, like, own your successes and be willing to talk about them. But I wanted to become, you know, basically an example and show people you don't have to be classically trained, though I honor and respect those that have even a high school degree, 'cause I don't, I didn't wanna finish high school. But I honor and respect the academic world because that's where Alexa came from. I then went to Microsoft and got an opportunity to work on the Microsoft AI team in its infancy. I was hired into research and development to help with the cultural shift that was necessary to take 17 academic projects and productize them. If any of you have been on this journey before, they need someone like me. They need to look in someone's eyes and go, it's gonna be okay. Our lives are gonna be okay. So that's what I ended up doing. I ended up being part of that, of course being a technologist and being not necessarily academically trained, but certainly I had built lots of code, I was a critical part of their testing, but most importantly, I just helped them realize that productization was the way the world was gonna see their work, and it became what's now today called Cognitive Services. I wanna share with you right now, it's only a minute, one, a very quick video, but it was the video that launched the day I started my journey at Microsoft, and to this day, I just, I'm like, that's me. So I wanna share it with you, and then I'll take you on a journey through open data science. So let's take a look. Today, right now, you have more power at your fingertips than entire generations that came before you. Think about that. That's what technology really is. It's possibility, it's adaptability, it's capability. But in the end, it's only a tool. What's a hammer without a person who swings it. It's not about what technology can do. It's about what you can do with it. You're the voice and it's the microphone. When you're the artist, it's the paintbrush. We are living in the future we always dreamed of. We have mixed reality that changes how we see the world and AI empowering us to change the world we see. You have more power at your fingertips than entire generations that came before you. So here's the question: What will you do with it? Every time I'm like, oh my gosh. When I first saw this video though, I saw it from the perspective of somebody who wanted to get into AI and who was going to be the artist, the musician. I went to Abbey Road and was the first ever to hold a hackathon in Studio One ever. Like, even to this day, I'm like, gosh, I hope I didn't make the gods mad. But it was incredible because we combined, of course, data science with music. I went to the Metropolitan Museum of Art, held the first ever hackathon with MIT data scientists and the curators at the museum to try and solve their problems. So when I first saw this video, I was like, oh yeah, I can change the world. I see I'm the user. But what ended up happening was I became the paintbrush. I became, right, the tool that someone could use. You are the tools, right? So an artist can't do what they do if we don't give them these resources that we're building, these models that we're building. I honestly believe that most of the work, if you are in the work of data science, It is a bit locked up right now. I have developers working in a company that has a data science org, and they've never used a data science model. And their company invests millions in it. And I'm like, how do we— how do we get those people to get together? Because that's where the magic is. And I found that out through hackathons. But the magic wasn't just in me having an idea of how to build a tool or even a developer having an idea of a tool they want. It was only in the delivery of an app that put the two together where a customer could consume it, and that's really why I came here. I mean, we have today more capability than ever. You're all part of this journey, right? I remember when I first, I remember in high school I was not super popular. I was a nerd, I guess, is what they called them back then, but it's funny because now I'm like nerds are, Actually cool, right? Like, they want you, all of you, whether you're good talking on stage or not, there are lots of people who just want to hear what you have to say because this is your time. And so I'm here to kind of share with you a little bit about how or what my vision is for how we get more of our work into the hands of people who can then make it available to customers, to students, to people like my son, to people like my dad. So of course, you know, this is an incredibly powerful time, but I always have to remind us, you know, good old Marvel quote maybe. But with great power comes— Great responsibility. Great— okay, we are family, good. Phew, got that out of the way. You never know how that's going to go. But yes, with great responsibility, and I do know that that actually comes from other sources like Voltaire, but— But Uncle Ben said it pretty well and made it cool for me to say in a boardroom that I could go to a boardroom, which is what I do today. I speak to boards and executives about the required investment that they need to make in responsible AI. I'm not building responsible AI. I rarely get the opportunity to pull a chain on a model that should not go to market. As a matter of fact, I was on a team where I knew It should not go to market, and I didn't say anything. And of course, you know, when you have those moments, you always look back and go, I should have said something, and I didn't. Now I do, but now I'm not in that position anymore. I'm not on those teams anymore. Now I'm leading those teams, but I can tell executives the cost of those types of decisions. And I'm seeing a lot of traction, actually. So the good news is with great power comes great responsibility, but that responsibility not only lies with you doing the work, but you leading the work, you inspiring others to think differently about the solutions that they're building and who it will impact. I now ask two questions when I go into a project for an assessment. One is the Jeff Bezos question: who is this going to be awesome for, right? The 1% of the 1% who want to talk to their kitchen. That's going to be amazing. But then, who could it be awesome for? The kids in, you know, that need speech therapy, that a device like this can help them articulate better. Or the guy in a, you know, with a spinal injury that can't do anything but look up at the ceiling and he gets to call his favorite radio station. We get to think differently and we get to think about it first, not in retrospect, which is what I learned on the Alexa team. I always like to mention this because Jeff Bezos, Jim Whitehurst, who was the CEO at Red Hat, many of the CEOs that I've worked for gave me this quote. And Jeff Bezos tweeted it out like 3 years ago and was like, I have this on my fridge and I encourage my kids to read it every day. It was when I— no, I shouldn't say that. But I was like, oh, he's a good guy. He's a good guy. So let me share this with you. I rarely read things on stage, but I think it sets the stage for the last kind of downward slope of the things I want to share with you. So what is success? You can read it along with me. To laugh often and much. To win the respect of intelligent people and the affection of children. To earn the appreciation of honest critics and endure the betrayal of false friends. To appreciate the beauty, to find the best in others. "to leave the world a bit better, whether by a healthy child, a garden patch, or a redeemed social condition. To know that one life has breathed easier because you have lived, this is to have succeeded." I love this quote because senior leaders at tech companies embraced it, and I was like, "That gives me hope in humanity." But I also love it because the person that made this popular was Ralph Waldo Emerson. A known poet laureate, metaphysician, big fan. I then found out because someone nice enough pulled me aside when I got off stage one year and said, "You do know a girl wrote that, right?" And so Bessie Anderson Stanley actually wrote that, and it inspired me every single day to always dig a little bit deeper, to take a cool story like Alexa being cool for the 1% of the 1% and scrape a little bit deeper. Who could I serve? How can I serve? How do I make this product more inclusive? How do I make sure the data represents more of who will possibly use this, not just who I think will use it today? It changed a lot about how I think about problems. And so I just think it's just interesting, like how many things do we presume and really are told to us as fact And we believe it. And actually it was just someone's opinion or someone's fancy rewriting of something that was really said quite a long time before. So something to always think about. Be introspective. Ask questions. So let's get into this. I have, you know, I only have a few minutes on stage, so I have to like tell you all about my philosophical beliefs for this world. One of them is inclusive teams. So I've done a lot of work on inclusive engineering. And what that means to me is that I, as a team leader, as someone who builds teams, I wait for the right people. I don't just hire who's readily available. I wait. And I have been told by CEOs to my face, waiting costs us, or it's basically at a cost we are not willing to pay. And that says a lot. I quit those companies. Some of you might look at my LinkedIn and be like, "Wow, there's some volatility there." Yep, 'cause I have no tolerance anymore. I'm like, if you tell me that, I'm out, sorry. Like, you don't get my TikTok anymore. So inclusive teams, it is what makes good products. Could you, yes, throw billions and billions of dollars into a hole, like Alexa was at first, and then hope that it works out? Sure. However, if we ask slightly different questions at the beginning, if we create, and you don't necessarily have to hire a team of every single person be different, but you need to hire someone. I was VP of engineering at NPR, and I did a talk like this to all of internal NPR, and I said, "I'm interested in inclusion," and for the very first time, this woman who was in audio engineering came up to me and she goes, I've never heard someone talk about accessibility and inclusion here. And I'm deaf. And she had cochlear implants. And she was like, this is the first time I felt like I was heard. Another thing that is really disturbing to me is when I look around, I've never worked with someone in a wheelchair, ever. And I've built— I've hired hundreds of people and worked with thousands of people. And I've never worked with somebody who's in a wheelchair. Or who has any significant disability at all. I've never even interviewed them. So it makes me wonder, right, how can I build a product that serves more people if they're not even in the pipeline of people that we're asking questions of? So the good news is, is you don't, if you have no power in this area or you have a CEO who's like, I'm not giving you money to wait for the right team, these people exist. As a matter of fact, you know them, you might be them. So I encourage you to speak up and more importantly be the one asking the question. Oh dear, I hit the wrong button. It's bound to happen. I get all excited. So the last thing I want to spend some time doing is really talk to you about my passion for this bridge. The bridge between your work or our work as women in analytics anywhere in the data profession and the world being able to consume it. There's lots of data today that says— actually, I'll probably go through some of it, right? It's super awesome to use data in your apps. It yields lots of benefits. We know this. It's why we're in this industry. It serves customers better because as an organization, we know more about them. We can meet their needs better. Talk to Netflix, right? Amazon. Even the creepy stuff that Amazon does, they actually told us they were gonna do it and we as a society were like, "Yeah, that sounds awesome." And only now are we like, "Oh wait," right? Like Alexa, we were super transparent in that initial launch. We had 2 lines of our privacy statement, "We save everything." And everyone was like, "Okay, but I could talk to my kitchen, that's awesome," right? So, like, it's interesting now that we fall back and we're like, oh my gosh, privacy, I'm not buying that device. I'm like, and your phone. Anyway, don't get me started. But we know now that there is a significant competitive advantage in understanding our customer data. But again, with great power, our customers' data comes great responsibility, asking better questions about how we're going to use it and how transparent we're going to be. But I would caution you before you judge teams that launch products because many times I've sat on a stage like this and I've asked this question. If I could alleviate all airport traffic forever, you'd never have to wait, you'd drive up, drop off, go, you'd be happy. The only thing I need to do is surveil you. That's it. I just need to surveil you. I've asked this on stages at like, you know, Not like us, normal people, you know, like people who are not in technology. 90% of them were willing to give up that data. Pshh! Couldn't raise their hands fast enough. And my very next word is like, no, like, no, don't do that. But the world, you could see this on my TikTok, I have a TikTok account and I do deepfakes. People believe things that aren't necessarily true. And so it's our job in the building to protect them. At least it's my job. That's part of what I do. I keep hitting the red button. So I like to remind you of how vast this work is. I got to, I was honored to work in the area of data science in nonprofits, which was really cool, but it's everywhere. Nonprofits, also, you know, worldwide, not just product organizations like Coca-Cola and Peloton or Netflix, right, but in organizations that are really trying to do something good, like the— I shouldn't mention them because it's funny now. I say it and people are like, oh, I don't like them. But there's organizations out there, we know them, that are doing good work. But healthcare, telco, right? We saw incredible things happen over the last 2 years just with COVID alone. And the only way that that really happened was that we released our need to hold on to that data. And that, I think, was a critical moment for us in our industry. I tried many times to get organizations that had movie data, right? No critical lives are in jeopardy. No PII, right? Movie data. I'm like, you have movie data and you have movie data. Let's put that in GitHub and you both could have better access. They said no. They're like, no, but that's my movie data. No, no, the model's yours, right? What you do with it's yours, but like, let's share. And COVID gave us a reason, a worldwide compassionate reason to share. And I'm hoping that we stay on that track. These are just some of the areas that I have worked in. I did a study of about 1,000 companies that I, you know, at Microsoft and Amazon and, oh, over there. And these literally I said, If you're going to work on data projects over the next, mm, I think I said 6 months and 5 years, what would you do? These were the Post-its that were put on the wall. This is where we're going to spend our money. And as a matter of fact, the 6-month boxes were things like customer intelligence, intelligent search, business process automation, chatbots. That was like the short-term goal. But the long-term goal was like predictive maintenance, edge surveillance, right? Some of these really interesting use cases that we're starting to see. Autonomous driving, of course. So companies know that this is the right thing to do. And I'm already seeing companies build data science teams that would never have done that 10 years ago. Like, I won't tell you the companies, but I'm like, really? Kindbar, you have a— Data, that's cool. That's awesome. But like, there's a lot. Every company now realizes that they— oh, and I spent the first before AI part of my career telling everyone to just store their data. Magic will come. Just put it somewhere. Hadoop. Do we remember those days? Like, just put it on a server in a cloud. It's super cheap. Just save everything. I am pained every day by that work that I did. But everything, just save it all, put it all in the cloud, and a magical unicorn called a data professional will come and save your life later. Little did I know there'd only be like 6 of us that could solve that problem for the hundreds and thousands of companies that now need this work. But companies know this. This data is very well known. So here's what I'm here to basically share with you. Obviously, I have a shirt that says open data science. The world of open source technology Because in the data world, we don't tend to think like devs, or we don't tend to associate ourselves with being a developer, because that's just— it's kind of a different world. We're not writing code in a traditional way. As a matter of fact, most data scientists I talk to, as soon as I start talking about code pipelines, yeah, they kind of glaze over, and they're like, yeah, that's not me. Good call. So what we could— what I decided to do was think about how could we make this more accessible. And so that's how I got ended up at Red Hat, actually, is I got very passionate about a project called Open Data Hub. Here are things that I spend my life doing that I really wish I didn't. Setting up an environment to build, I don't know, Jupyter Notebooks. I spend a lot of unnecessary time just making my Jupyter Notebook work. Like just getting it to work from the time I open it to the time I can actually get that first thing to run, I feel like I shouldn't have to do. But what's better yet is I want some of my stuff to be able to run faster. I want to be able to provision cloud services. I know they're out there. I know I have a company. It has a team of people that do cloud. I'd like my stuff to run there and go fast. I'd like to be able to— provision 200 nodes to support it, but I just wanna do, just me, just to test out if the idea that I'm working on works. And up until recently, that was difficult to do. So what a open, a group of us, right, got together and was like, what if we created pre-provisioned environments for the most common workloads for data scientists so you could just press a button and poof, they exist? So that was the first problem that we tried to solve. And again, this is the same problem we solved at the Metropolitan Museum of Art. Their biggest pain was just something they do every day, like lack of automation. They had to tag things manually. I think it was like 6,000 years of art manually tagged by curators. And they were like, sometimes that work isn't awesome. Sometimes it's fun when we find a new thing or we get into an area we're interested in, but most of the time, It's just rote tagging on things I don't really understand anyway. And that's where data science really came in to help them. So in looking at this, you could see— I really wish I would have put pigtails on that little image right there. But I will. You'll see it. I'll post it on the Twitter. But the data scientists, right, what we want to do is take all this work you are doing and make it available to a developer. And we want to do that super fast. And I'm going to actually give you this kind of concept as a— I don't know— gift. And then I'd like to see if there's any questions about anything that I've talked about as well. But a data scientist typically builds their thing and is kind of like, that's cool. And that's it. A developer often would love to have a model predict rather than them write decision rules. Which is what they're doing today, right? So, we have this perfect moment that we can come in, and I'll tell you a quick story about Alexa. In the early days of Alexa, there was no pipeline. There were people. People monitored the unit tests, the functional tests, and production, and things broke in every environment, and I think, oh my gosh, I remember us trying to automate builds, and it was like 0 of 420 automated builds Zero. Like, not one worked. I almost was like, we should be proud of that. Like, how— that's hard. Right? But what— but we, of course, evolved. I hired a software engineer and said, come build our pipeline. And they did. And magic happened. We stopped worrying about deployment. Which I shouldn't be worrying about anyway. Right? It gave us freedom to focus on what we should do.— or I shouldn't say should, but what we want to focus on. So Open Data Hub is a project, you can go take a look at it, but there's lots of open source communities. Most importantly, datasets are open. I'm sure you know this, but more datasets become available every day. Metropolitan Museum of Art, they open sourced 400,000 of their images, and you can use them, which is pretty cool. Music is open source. Like, there's huge collection libraries of music. But really, it's anything. Health data, dog data, banana data. You'd be shocked what's out there. But GitHub is your friend. And for those of you just starting on this journey or know someone who wants to, start there. I'm a big fan of learning by doing. Start with those types of processes. So this is my— philosophy. I've shared with you a little bit about who I am. I feel like I know you, and you probably— maybe I don't know you, but I feel like I do. I feel like we're friends. Hopefully we'll get to be connected. But I think it's important we recognize that the world, in order for us to do the things we want to do— and I should say do the things I want to do, which is improve the lives of people, make people happier, more content, focused on things they love— we need to to think about open source. We need to think about taking what we're doing and making it available for other people to learn on and build on. Of course, we need to do that with an open mind. Mindset is key. So, if you— I mean, if you're here, I would imagine you're open-minded because we're in a pandemic that's not over yet. So, yay, good, we're here. But we should really be aware of the power of having an open mind. The power of an open mind is asking questions no one else is willing to ask. And being willing to stand up for people who have no voice. And then finally, Noelle, like, have an open heart. Be willing to be emotional. I swear I had to not— like, last night I'm like, "All right, when you get up there, don't you cry. Don't you cry." But have an open heart. Be willing to be yourself. And maybe yourself is not emotional. There's plenty of us, right? Maybe yourself, it isn't that way, but be whole. Be wholly you because I have proven the craziness that is Noelle, it's okay. As long as I am showing people what broke and how I'm trying to be better, it's like Mother Teresa. If I can't feed 100, just feed 1. If I could change 1 person and make them get into a field they never thought they'd get into or productize something that was only ever a pet project in an academic environment, That's a win for me, and it's a win for all of us. So thank you very much. I want to give us a couple minutes, maybe take a couple questions, but thank you for having me. So we do have time for a couple questions if there are any. I know, it's over. Oh, the selfie! Okay, while we're— if anyone has a question, just raise your hand. Oh good, there's one. And while we do that, I'm just gonna take a quick selfie. Say hi! It's terrible because when you— I do this to my kids, they say hi and then all of our pictures are like— but if you look at my socials, it's all that way. Okay, carry on. Hi! Is this on? Hey, so my question is around hiring diverse teams and things like that. As you mentioned, sometimes the pipeline is part of the issues. So do you have any tips for maybe alternative means or how we reach out to make sure we're capturing that? Because I definitely see a lot of the same types of people and I would like to diversify. Yes, thank you, great question. So the question really is around like, where do we find these people? Because they're out there. First thing I would encourage you to do, one, is be— is change the wording. Make the wording— do not copy and paste your job description from the poor fellow who wrote it before you. Don't ask me how I know that. But it's easy to do because you want to get it out there, but create inclusive wording because I've heard so many women be like, oh yeah, see that? I can't do that. They're not actually asking for that. They don't even know what that means. Like, just apply because what you do matters. You could do this. So first thing, Create more inclusive wording in your job descriptions and fight for that so that people don't immediately say no. But the second thing is that there are really great groups just like this one dedicated for Latinas, African Americans, there's what's called AfroTech, there is, there's so many different groups. There's an organization called DevColor for mid to senior level engineers for women and people of color in general. Find those areas and I will say this, I've said this 100 times. Become part of those communities. Just like there are men in this audience, thank you for coming, right? You don't have to be brown to be in DevColor or be in Black Engineers. That's not a requirement. The only requirement is that you are part of the solution. So that's, I think, a critical part, is get into those communities, befriend those people. That's why I talk everywhere. I have been, and I'm heterosexual, But I went and did a keynote for Lesbians Who Tech. And I was literally painted with that brush. People made an assumption about my sexuality from then on. And I was like, but I mean, I guess it didn't matter. But I would do it again and again and again because it created visibility for those individuals to have a role in tech. And that it shouldn't really matter. But get in those communities. Be an ally. Join up. I mean, I think we do need to be a little bit better in some of our community— communities about embracing people. But, but that's what I would suggest. What was the other advice I gave to someone recently? Oh, become more flexible. So we are in a pandemic, of course, but the world of the future is flexible. Try to think about, do they really need to live here? Do you really need to see the whites of their eyes? And if you're not looking at the whites of their eyes of everyone else on your team, Don't do it for a person of color. Again, don't make me cry, just sometimes we set unconsciously different standards, and that's hard. So yeah, thank you, great question. Any last question? All right. Oh, one more, yay! Thank you, I literally had to be practicing like, what if nobody asks a question? Hello. Hi. So I have a question. Can you please collaborate on one of your slides? You say that in 2030, the demand is going to be higher than the supply for that. I'm not understanding well. Yes. Oh, there it is. This one, right? Oh, maybe. But that's true today. Today, there are not enough people with the right data skills to do the work that companies have on their backlog or have in their roadmap. And what that means is that there aren't enough data scientists who can build models. There aren't enough even AI engineers who can augment existing models. We also— it's a secret. Why is what we do a secret? People don't actually understand data science. And I think I felt, at least as a new member of the data science community in 2013, it was a bit of a pedestal. Like everyone I talked to was kind of like, oh, How many PhDs do you have? I'm like, I didn't graduate from high school. And they're like, oh, so no. Right? So I think— but that's different. I think now we're much— at least I'm trying to create a much more available and accessible world. But a lot of companies don't realize— they don't realize that they probably have people in-house that they could reskill. But they don't have enough data scientists to actually build models from scratch. I am a huge proponent of not building from scratch, taking an open model anyone Anyone know ONNX, right? A huge fan. If you don't know it, O-N-N-X, but it's basically a platform that lets you separate inference and training, but most importantly, you get to pick different languages and you get to increase your flexibility. If somebody knows AWS versus Azure, I mean, it just, it increases accessibility. So that's the biggest thing is that, yeah, I hope, was that your question? Like, yeah. So if you're in data science, chances are your salary will go up. That's good. Because supply is definitely going down. The other sad thing— and I'll just tell you from a Latina— the percentage of people of color and Latinas is going down. And it's hard to imagine because we're already in the single digits. But it was like 3% 2 years ago. It's 1% this year. And this is based on the data of pipeline, meaning people that are going to— get educated on data science. Because let's face it, data science isn't something you can just learn at home from a book. At least I haven't been able to really. The models I've built, I would not put in production. But it does create— we have to create a way. And the way I've solved it— ooh, good. OK, last point. If you don't know it— oh, gosh. OK, if you don't know it, This is like when the big hook comes or the song from the Oscars starts playing. But if you don't know it, MIT has a platform for educating middle schoolers. It's called RAISE. Are you familiar with it? Yes, I am such a huge fan. They have no video. So this is what I'm doing now. AI Leadership Institute, what we're doing is I'm creating videos for classrooms to use. So I'll be like the face of the middle school. But here's the funny thing. RAISE, which is for middle schoolers and younger high schoolers, it's like intro to AI, what is machine learning, that kind of stuff, most importantly, ethics and responsibility with AI. The cool thing about that is that actually everyone could use those classes. Everyone, everyone, right? Like it's meant for middle schoolers, but let's face it, if you can teach a middle schooler, you could teach anybody. So yes, so take a look at it. I can leave you with that, RAISE. At mit.edu. It's an amazing set of resources. Thanks for having me.