Janeen Uzzell: Full Interview

Hey Alexa, play Machine Yearning.

Here's a sample of Machine Learning by Fingerspit.

No, Alexa, play Machine Yearning podcast.

You should try to mumble less.

Probably, I don't know.

M-A-C-H-I-N-E-Y. M-A-C-H-I-N-E-Y. AI, AI, AI.

It's Machine Yearning from Assist. It's a podcast where we think and dream about the future of AI, the talking internet, and how we're reshaping our culture. When you're a scrappy new podcast covering the third wave of the internet, life isn't about being pitched by big PR firms and their arms folded on the COVID of Forbes CEOs. When you want to find out what's really going on with AI and machine learning, You talk to data scientists and researchers and startup leaders. So this week and for the next couple episodes, we're taking you to the VOICE Summit, the largest single gathering of the conversational design ecosystem. 2,000 people swarmed the New Jersey Institute of Technology July 24th to the 26th. The team at MoDev crushed it and they took great care of us. While we were at VOICE, we spoke to tons of technologists and entrepreneurs. But we also met people who are living at the front edge of creating a more inclusive culture in technology product design. We're not giving away family secrets when we point out that the US technology sector is not covering itself in glory by creating diverse, inclusive workplace cultures. Enter Janine Yuzel. Janine is a New Jersey native who is on a mission to use her influence and voice to lead significant work that changes lives, communities, and the world. Most recently, Janine was the head of Women in Technology for GE. There she worked across the 300,000 employee community, leading a culture shift to accelerate the number of women within GE's technical female workforce. Janine spoke with conference organizer and friend of the pod, Janice Mandel. In this conversation, they take a half step back from voice and dig into the process of human-centric design on a global scale, how you design for kids, and how you build and lead inclusive, diverse teams so you can attract and serve a diverse customer base. You've been in meetings where these topics come up. It's hard. We get it. Today's a chance to be a hero and part of the solution. Take a second and share this one around with your boss and every single person in your group. The stage is set. So now, Jeanine Yuzel from The Voice Summit. Let's go.

As an emerging market leader, my job is to really understand the global market and to put our technology in those spaces or to understand what technologies can work well there.

You did how many years with GE?

I was at GE for almost 2 decades in a number of roles from operations. I worked in global health and then ended my time there as the head of women in technology. So wonderful experience. I was leading a team and we were focused on ultrasound. More specifically, when you have a baby. Yeah, ultrasound. So ultrasound is a really large, bulky, heavy product, and it needs a lot of consumables, what you actually need to make it work, and the power and electricity that it generates.

You said it was like the size of a washing machine.

I call it the washing machine. And so it's a mini washing machine, but a washing machine nonetheless. And we were simplifying this product to take it into markets like Indonesia and Bangladesh and sub-Saharan Africa. And the result was our handheld ultrasound, which is a product called vScan. It's a wonderful, wonderful product. It can be solar powered. It can be used, um, very non-traditional spaces and it does very, very basic ultrasound technology in a handheld, in a handheld device. Now, you know, when we were putting this product together, we definitely were thinking about the end user. You couldn't tell me we weren't. I lived in India, worked there, been to Africa, watched the midwives work. We know how to design this product. At the end of the day, we are creating job opportunities for non-traditional task-shifted users. We were helping support the United Nations Millennium Development Goals. We were going to generate incremental revenue for the company.

All good. All good. You had your white hat on.

I was ready, like, give me my finish line and run over the edge. Until it just— none of that happened. We had certainly designed. What they needed. This product gave them what they needed: solar-powered, handheld, easy to use. It just wasn't what they wanted. Some of the most basic things, for example, the product is white. It gets dirty too fast. We're not proud of it. It looks dingy and old. We don't want it to be unattractive because we're businesswomen and we're trying to build our company and the other women don't like the way it looks. We really don't want it to be handheld. "We don't have any pockets to put it in because we wear saris or very traditional clothing and women didn't have pockets." So we're thinking, you know, you put on your lab coat and you put it in your pocket. But in Indonesia, which has some of the largest population, over 100,000 midwives that go through their midwife program, they don't have pockets. They're wearing a traditional garment that doesn't have that. So they're like, "I need a handle. I need something to carry it with." So some of these things, were so easy and so simple, but cost us in the end because now you're talking about going back to marketing, going back to the design team, having to recreate a product and redistribute it. And fortunately, it did work for us, but every company won't have the resources or the opportunity to do that. So we've got to catch this up front and solve it early on.

So how do you do your research then? Are there some ways that people can be mindful of— Making sure they're inclusive?

I am a very strong proponent of going as grassroots as possible for research. I was very non-traditional. I was in the field with the midwives, you know, working with them, trying to, you know, to understand this most. But what I could say to people overall is you have to ensure that your sample size is broad and you have to think out of the box. And so, you know, if you're a financial company and you're designing a financial app to teach or to ease the process of investing, I'd go talk to the people that don't have any money. Why are you going to sit around and talk to people that have tons of money, call their financial advisors or know how to use their apps, know exactly what they want to do? I'd go to the folks that want to aspire to be that and don't do traditional banking because then they're going to say something like, "Well, when I click on this, I want it to give me a small definition. I forgot what that means. I don't remember all my investing terminology," or, "I just want to use something that's that's entry-level, something that allows me to invest a dollar a week or save my change at the end of the month that will— and I don't make enough money to be in your organization or to be in that group of sample size. But I'd get the data from all of those people because now you're going to design a product that's going to market and expand to people that think that they can't even be a part of your continuum. I'd also, again, I like to talk to media, whether it's magazines or schools or programs, for anyone that I think could be a potential end user, live with one of the end users, know someone that has a grandmother or an aunt that could be an end user, all of those folks are important to content. And the more conversations you have with people, the more you'll be able to learn where your resources are, and they come from the most non-traditional spaces. And so, data is everywhere. It is everywhere. It's on the bus, the subway ride that you're taking home. It's just a matter of being open to the uniqueness of where it can come from.

Now, I know that fintech gave us a jumpstart, right? They figured out how to make things available the way people live, run, you know, into— make banking available just, you know, so they didn't have to make the dangerous run home.

Correct.

Right? And they can just send it home. We have a dangerous situation coming up because there are no laws right now that talk about employment and what these algorithms will control. How are we going to do that?

Well, right now, you know, from a legislative perspective, there's some things that are always thought about, right? Age, race, some of the basic discriminatories. But there are communities of people that even I can't think of them all. But for example, people that work the midnight shift or some of the non-traditional users that I talked about earlier. When we're not covering or thinking about them, then obviously legislation needs to go much broader. The opportunity to put more seats at the table is absolutely where we need to go. Also need to include ethicists. They need to have legal teams at the table. This table has to go broad and just when when you think it should just be my programmer, my coder, and the person that understands how voice technology or any technology should work is exactly when we're missing it. So, we've got to come together. There are some organizations and guidelines and teams that have been put in place by like IEEE and other organizations to say, let's convene and ensure that this table is broader. I think psychologists need to be at the table. That's important. There's a level of thinking that's very different as we're starting to tap into the homes and the lifestyles of people with this type of technology. Where's the privacy? How are we managing this differently? I think we'll start to see it in all sorts of ways as people are starting to leverage social media and all of these apps to run for office or to determine where they stand politically or socially or economically. Having all of these people at the table is going to matter tremendously as we get deeper into this.

That's interesting. It's a big job. You know, Steve Case and his, uh, is it "The Third Wave"? He wrote a book recently, and he said, you know, we're past the time where you can make an app and sit in your dorm room and everybody will get it and that's it. We're now messing with the very fundamental foundations of our society, and that's a big ecosystem.

It's huge, because we don't know— I mean, things go everywhere all of the time now. I was reading where an author was saying that The disparity or the circumstance that we're leading to could be circa, you know, just the challenges of civil rights and some of the problems that we've had because we're at the mercy of the thoughts of the programmers. And again, we know that some of those, while they're not intentional, some of them could be, you know, some of the racism, sexism, otherwise. We're teaching machining how to do this now.

Right.

Machining is now going to become bias to end users. And machining, we're not even talking human beings. So now we've got humanity and the machining all turning against members of society, possibly just based on the programming of one group of people. And while it could seem like, "Is that possible?" It absolutely is, which is why—

Huge.

I said earlier, there is proof that some racism is stemmed just from programming. Wow.

So you got a big job ahead of you.

Well, I want all of us to take it on together. And I'm telling you, it's the responsibility, again, of every technologist, every person that decides that they want to work in this space and be a part of the growth of this business. It's the responsibility of the leadership, the people who lead the teams, the people who contribute to the content. The lawyers, the ethicists, the psychologists, all of us, because all of us are gonna be affected by it. It's our world. So it's no way that you could say, "Oh, I don't use that, so it's not gonna faze me. I don't need to be involved." You're involved. Facial recognition, you're involved. Anywhere you go and your photo is being scanned, you are now a part of this. Don't wait until you have a negative experience to decide that you wanna get involved.

Machine learning.

Janine Yuzel speaking the plain truth. Every one of us has a stake in the future of AI, and it starts with us building teams that are representative of the people who will be using the products and services. How is your company or organization doing? How do you know? When was the last time you asked? We'll have more from Janine and our guest interviewer, Janice Mandel from Modev, from the VOICE Summit in just a minute. Quick plug: we have some great conversations from VOICE coming up in the next few weeks. Will Hall from The Rain Agency, Kathy Pearl from Google, and Patricia Scanlan who has built a natural language dataset from over a million samples of children under 12. The use cases of that dataset are kind of boggling. Got an idea for a guest you'd like to hear on machine learning? Let us know. DM us @Assist on Twitter. And now more with Janine Yuzel.

Today at your session, you were talking about that you're an evangelist. For getting the bias out before we really go full forward on AI.

Right.

What does that mean? Where are you taking us?

The term evangelist, I love that we get to come up with these sleek, sexy kind of names in the tech world. But it just, you know, really means that it's a powerful, powerful thing to me. It matters so much. And we're going to baptize everybody and help convert them. Where we're taking it is this. Diversity and inclusion is not a strategy or an idea. It needs to be the way that we do things always. And it only becomes that when it's your regular practice. It's like taking a shower, brushing your teeth. You just kind of do these things all the time. The value and the, just the criticality of having this in our world of technology is so important. And we see the broader and the faster we go with AI, the more we're starting to see the flaws because we know that Our technology is subject to the software, and the software is subject to the programmer. And so, every one of us that are either hands on the keyboard or in the product design rooms or leading teams that do this work need to be responsible or feel a burden to ensure that there's no bias in what we're designing.

So, that's where you want to get to. How are we going to get there?

Well, that's the perfect world, right? We get there, practice, practice, practice, and over time. Because bias is inherent in all of us. We could say, "Oh, I don't have any unconscious bias." Well, that's why it's unconscious, because it's there and you just don't know. And it's kind of just the way we were raised or brought up, the environment that we're brought in. But to be able to build and design and to think beyond race, gender, age, sexual preference, culture. It's not even something— I'm not putting that burden all on you. It's not on one person. That's why we need diverse teams. That's why we need to have people that can help us to focus, to make sure that we're thinking a lot wider. Now, in terms of how long it'll take us to get there, it's going to be ever-evolving. This is something that will always be a practice and something that's a part of the continuum of technology. What I'm saying is that we need to pay attention pay attention to it and get it together. Because we're missing out. The customer is missing out. One, they're not getting a full product. We're also— there are tons of customers that we're missing by not having the breadth of product, market share. All of this affects investors, shareholders, product design, the advancement of this. It's all going to be impacted.

So they're all part of this. Absolutely. Is this something their employers can help us do? Where does it start?

I would say for sure, when I think about employers and when I think about your hiring practices, some very tactical processes. Where are you getting your talent from? Where are you recruiting? When's the last time you refreshed the schools that you recruit at and the schools that you invest in? Because maybe some of the folks that you're looking for are going to very different schools. Maybe they're getting educated in different spaces. Everyone's not going to Schools 1 through 10, but there's still some great technologists out there that have broad thinking. Another thing that's very tactical is, what do your job descriptions say? Sometimes we write a job description and it just sounds so dull and so dry, and the creativity that's required to make it sound much more poetic so that people are like, "Yeah, I really wanna do this work. I want to be a part of this team and do this work." What are we doing with the managers and the leaders on our team? What type of training do they have and are they ready to lead the new crop of technologists that are graduating from the schools. So, you know, just some of the practical ways that we do business. But it's also the responsibility of, again, the leaders of these teams, the actual technologists themselves, to want to think beyond their perspective. And again, we're not asking you to do it yourself. We're asking you to work in these mission-based teams as you design these types of products. You have to reach out and get the help. You know, it literally has to be intentional. And like I said, if I'm not thinking about something right now, it's because it's just not my norm and I don't know to think about it. But I have to figure out how to, you know, is my team reflective so that we can understand how to do it differently.

Machine learning.

Janine Yuzel with specific real-world practical insight on how to nurture a culture that's grounded in inclusion—

Machine Yearning!

At Machine Yearning we're focused on the intersection of culture business technology it's not huge event so sharing this pod to your friends colleagues would be amazing thank you very much for listening I'm Sean O'Leary from DataScientist.com see ya next week at machine learning we were focused on the intersections of cultural businesses and technologies that's not a huge vent just share the podcast with our friend or colleague who might enjoy hearing about what happens behind closed doors here at Google Research London thanks again everyone bye-bye now foreign speaking Thank You rating us on iTunes, or even writing a review. It makes a difference because it helps other people who are interested in a healthier, more sustainable approach to business discover the pod. So make sure you subscribe to Machine Learning so you don't miss the upcoming interviews from The Voice Summit. Now, more with Janine Yuzel.

So what's at stake here? I mean, entering into an era where AI's involved, Does this have a multiplier effect? Are we at risk of rolling back civil rights?

What I will say to that is, we certainly believe that the bias in coding and in technology can have an influence on racism. And so, we could certainly be shining a brighter light in areas that we want to dim down. We definitely— we definitely want them to go away. Mm-hmm. But we're starting to shine a brighter light on that. And we're seeing some of that now, because technology is everywhere. And AI is going to happen. It's happening now. The question just becomes, will we be the responsible community of people to ensure that it's a fair practice for everyone? And when we don't do that, then the end user or even a person that is not an end user but just is on the other end of that bias receives the penalty for the arrogance, the ignorance, the lack of knowledge of other people or the programmers, the coders that are in the room. So we're going to, we will experience that. We'll experience it when certain things in life require us to have access to a technology that I may not have access to. Um, we'll experience it when a technology is responding based on a bias in the way that it's been coded. We'll experience it in things like facial recognition and not having a broad enough bandwidth of diversity for faces. And facial recognition will presume that I look exactly like someone else that has done something inappropriate. And then I am the one that's blamed for it. So there's, there's so many ways that it will have an influence on us in a negative way, but that doesn't have to be the case. This technology is being built and designed in advance. So for the good of our society, and there's so many great things that we can do to make it better. We just need to ensure that that's the practice.

Well, we learned something from the last go-round, did we not? As we started to.

True. —True. —And we also learned that there is no digital divide. You know, there's a presumption that, particularly in the world of minorities, that there's a digital divide and that we are not able to code or understand technology, so therefore, the world doesn't have to design for us, and that's not true. We hosted a hackathon about a year ago where we had students just from historically Black colleges and universities, and this is when I worked for GE. We hosted this hackathon, And the goal was to prove that there was no digital divide in diverse communities and that we could have these students come in. Half of the students were actual coders and the other half were technology students like engineers that may not be coders, like a mechanical engineer or someone. And in 36 hours, learning how to use our technology platform that you coded on, they came up with the most amazing solutions, 3 solutions, one in health, one in phone technology, and the other one was working with air trafficking and control. But the uniqueness of all of these three solutions were that they were all for adjudicated environments because that was the way that those students thought because they come from those types of spaces. So they were designing solutions around Black Lives Matter and things like that because that's what their experience was. So we need people like that in the room. Right. And that's why it matters. So we, we learned the hard way, but what we know is that there is no digital divide. Technology is at the hands of everyone. My mom is She's 86 years old, and she knows how to use her phone, her smartphone. She knows how to take a photo and send it. She knows how to use her Bible app. And, you know, there's a few things she really doesn't know how to use, but, you know, you give her a sheet of paper and a pen, and she'll write it down, and she's teaching herself. She knows how to use her smart TV and, you know, her Apple TV. So she's learning, and she's like, "I just want to know how to watch Netflix." So she wants to learn how to do that. Well, that's— so there's no divide.

That's how we got you, I think. Just raise your rice.

Right.

Machine learning.

I have 11 nieces and nephews. The youngest ones are 3 and 4, so they don't even understand that you can't swipe on the television to move things around. Jace, he stands there swishing the TV and he just says, "Broke, broke," because he thinks the TV's broken. My great-niece Kennedy thinks Alexa is her friend. "Hi Alexa, how are we? I'm sad." Or if Alexa doesn't have a conversation with her, then it's frustrating to her, because Alexa answers a question, but she's not just like, "Good morning Kennedy, how are you?" And so that's confusing. I have a few challenges there. One, well, let me say the things I think are great. My godson, who has a lot to practice with spelling and things, thinks Alexa's a toy, and so therefore he likes to practice, and he can practice spelling with her more. So practicing and things like that, I love that. I'm concerned with what voices they'll use in the future. I don't know because I don't work for the company, but what voices are in the pipeline? Will it always sound like the current voice that it sounds like now? Will it be a male influence? Will the accent ever sound like my great-grandmother or my grandmother? What if I'm from New Orleans or from the Southern Geechee Islands? Will it sound like that? Will it understand that? Will Alexa know what collard greens are? In my family, you know, we eat soul food. Or is she gonna tell me how to cook kale, you know, and think it's It's the same thing. So those are the things that I always test when I'm doing some of the research. But it concerns me that this will become the resource to raise our children if we don't continue to activate dialogue. It also concerns me that the younger children have less ability to make eye contact in their communication because they don't have to look at this machine to talk. So when I'm calling your name and I'm talking to you and you don't turn around and acknowledge me. There's a skillset that's missing there as a future leader. You're not— I don't believe the future's going to be just the net where you are sequestered in this house and you never leave and you work from your computer and all your screens. This is not a reality that I would want for any of the children that are a part of my life. So I love technology, love it, love it, love it, value it, and want to see it be the best that it can be. But I also know that it's So important for us to maintain some of the basics. I never thought I'd be old enough to be saying, "Back in the old days, we used to," and this matters, you know, but it really does. And so I think we're going to see the good and the bad. And I also heard during one of the keynotes yesterday where from an elderly perspective, there's research that senior citizens are saying it's just nice to be able to hear a voice if you're alone and by yourself all the time. Right. That was something I had never thought of. Never thought that you could get lonely and just say, "Hey Alexa, and maybe play me a song or say something to me," and she respond. And that could be a source of comfort, which is why it's so important to have psychology and sociology at these tables as well, because we are going to lose something if we don't integrate and do this together.

One of the presentations earlier talked about making sure that you're not using this to replace replace that human contact. It's a supplement to human contact with the elderly. It's not a license to ignore them. And you know what?

Ultimately, this has got to be intentional. This practice of diversity and wanting to ensure that we try— I understand we're not going to get it right every time. It just needs to be intentional. It has to be something that we make a priority.

And how do we keep it there?

You have to be intentional about it. When it's your intention to do certain things, it becomes a part— it's your lifestyle. It is my intention to be a good person, to do what is right. I'm intentional about it. Do I do it every time? No, but I'm intentional about it. And then when I get it wrong, it's a priority for me to figure out how to do it right again. It's intentional for me to make my family a priority. And when I can't, I have to go back and course correct. So it's really all about intention.

And do you have people in your life that hold you to it? Absolutely.

We have to hold each other accountable to that. And that's where these teams come around to say, "I know it's our intention to be this way, but we are missing it here. Let's go back and figure out where we're missing it and grab it." So it needs to become a part of our work culture.

Absolutely. And our values. Yep. Awesome. Janine, I can't thank you enough.

Oh, thank you so much. It's been such a pleasure to be here. Such a pleasure. Such a pleasure. Thank you. Thank you so much. Thanks for making this topic a priority. I appreciate that.

All right.

Thanks for listening. And thanks to Janine Yuzel and our friend of the pod, guest interviewer Janice Mandel from MoDev. Next week, Will Hall from the Rain Agency, Brooke Hawkins from Nuance, and Joseph Hapley, who heads Aether, the marketing agency he founded in Washington, D.C. All from The Voice Summit 2018. Get in touch on Twitter @assist. DMs are open. Tell us who you think should appear on the podcast. Please subscribe so you never miss an episode and share this with someone who cares about how we make sense of these changing times. Machine Learning by Assist is made by Paul Chufo and Michael Alcesser for Limina House. Have a great day.