Breaking Barriers in AI - Inspiring Change and Empowering Careers with Polly Allen

Christine Li  0:01  
Welcome back to the Make Time for Success podcast. This is episode number 151. 

This episode today is centered around the topic of artificial intelligence. And because we are talking about artificial intelligence, you're gonna find that I was struggling a bit inside the interview to really find words to describe the questions that were coming to mind but thankfully for me and for you, my special guest in this episode is Polly Allen. Polly is someone whose whole mission is to demystify artificial intelligence so that everyone including lay people like me can share in the economic boom and and all the benefits that artificial intelligence will bring us. She has over 20 years experience developing software, building and leading software teams and most recently leading data science and engineering teams as a principal product manager for Alexa AI at Amazon, where she led the development and launch of the first generative AI answers on Alexa using technology similar to Chat GPT. She's been a leader in the application of machine learning. And she is passionate about diversity and inclusion in the artificial intelligence space, and she wants to empower more people to understand and participate in the field. She also has founded as of November 2022, the world's first AI focused career accelerator called AI Career Boost in general, she is just a really clear and wise and wonderful person to learn about AI from and I learned a ton from this interview conversation with her. And I can't wait to share this conversation with you now.

Hi, I'm Dr. Christine Li, and I'm a psychologist and a procrastination coach. I've helped 1000s of people move past procrastination and overwhelm so they can begin working to their potential. In this podcast, you're going to learn a powerful strategies for getting your mind, body and energy to work together. So that you can focus on what's really important, and accomplish the goals you want to achieve. When you start living within your full power, you're going to see how being productive can be easy, and how you can create success on demand. Welcome to the Make Time for Success podcast. 

Hello, everyone, today is going to be a very cool session. I have Polly Allen with me as our special guest. She is the CEO and founder of AI Career Boost, we're going to hear from her all that that entails. And I want you to know she really is an expert in the field of artificial intelligence, she's played a role in developing many of the systems that we are familiar with and some of us are not so familiar with, when it comes to artificial intelligence and really the this boom that we seem to be in when it comes to new tools that we can use to help us write essays that we might be stuck on and explore different travel itineraries. Whatever we've got going, we can use AI for. I'm excited to have Polly really give us her insights on AI as a tool and how you all can begin to use it. Welcome to the show, Polly. 

Polly Allen  4:02  
Thanks so much for having me. 

Christine Li  4:04  
So Polly and I met at a conference recently in real life. And as soon as I heard that she had aI as her expertise, I thought immediately, how can I get this woman on the May time for success podcast, and she fortunately was also really gung ho on the idea. So Polly, could you describe for us how you got connected to the AI section of the universe to begin with?

Polly Allen  4:33  
Sounds good. Sounds good for sure. So my background is as a software engineer, I did my degree. And both my undergrad and Master's in Computer Science many, many moons ago, over 20 years ago. Even in my master's degree, I had been interested in AI. And at that point, we were in the middle of what's been called an AI winter so it was very not cool at the time. And very different from the kinds of AI systems that we're building today. I spent 10 years as a software developer coding systems before realizing I'm pretty extroverted for a software developer, and I'm really interested to on what we should build and why. So I moved over to the business side as a product manager, which are the people who kind of put the business case together and but also help things get built. And my most recent role in 2019, I joined Alexa AI. So the team that builds Alexa and Amazon, and we were focused on have expanding the ways that Alexa could answer you. So instead of having a hard coded answer for a particular question, we launched the very first generative AI answers on Alexa, which we're using a similar technology to chat GPT. So people are surprised sometimes to learn that it's been around for a few years, it's not very, very brand new. And it's been really exciting to see this area of AI in particular, explode into public consciousness in the last six months, really, really amazing just to see the attention around it, and see all the new ways that people are using it, which is really, it's really mind blowing. So in 2020, to the end of the year, I actually left Amazon, really, because I found that I was pretty sure AI was going to become a very big deal in society, I just didn't know how quickly. And I had found that there wasn't a lot of diversity in the rooms where people are making decisions about these systems. As I'd been promoted at Amazon, I every time I went into a new meeting, there were fewer and fewer women, I was often the only woman in a room 2030 men making decisions about systems that affect 10s of millions, hundreds of millions of people. And it really struck me that this wasn't the way for the best AI to really, you know, come to light. And it really wasn't fair either, that these really high paid and high power jobs are really seen as this realm of like, only the techies can do this, and only a certain subsection of the population. So I founded AI career booths to increase access to careers in AI, and also diversity and leadership in AI companies. And it's been a really exciting ride. Since I launched it in November, I've been super excited with the folks who are really passionate about AI using it new ways, and tried to get into careers in this space. It's been a it's been absolutely amazing with the folks that I've come in contact with. Through this, yourself included, Christine,

Christine Li  7:37  
Thank you so much. And that is wonderful. So not only are you a pioneer developer and involved behind the scenes in really important and specific ways, you're also starting of movement, a new movement within the movement, which I love, which is really focused on diversity and equity, which is just so important for society, I believe and not just the field of AI. 

Polly Allen  8:01  
Right. I love it. That's right, starting a movement starting a movement. 

Christine Li  8:06  
Can I ask you about the word generative? Yes. Before I ask you about the people that you're working with? Could you explain exactly how that's different from what was happening before? And what that gives to us? What that gives your experience and how you thought about it as you were working on developing that?

Polly Allen  8:27  
Yeah, really, really great question. So AI drives a lot of the systems that we're used to seeing day to day already, right? When you think about everything from your social media feed, how does it decide what to show you on Netflix, those content recommendations, that's all artificial intelligence, right. And a lot of what we call AI today is referring specifically to the branch of AI that looks for patterns in data, and then makes a prediction or a decision based on that, right. And so generative AI is the first time that we've been using to actually generate new content, whether that's images, right, and people are often familiar with Dolly to or mid journey, where you know, six months ago, 12 months ago, people were blown away by being able to just describe an image and have it generated pretty well. And then this new boom of generative text AI has been exciting, because not only can you describe what you want it to write, but it's even able to generate things like code or translate something. So this combination of like not just being able to make a single prediction, you know, a number or classify something as a yes or no, or give a set of recommendations, but actually, like, generate new content in a way that seems pretty in line with human creativity sometimes. That's the piece that's really new and exciting. 

Christine Li  9:56  
Wow, wow. I am just reflecting on how I was struggling at the beginning at the top of this episode to find words. And I'm thinking that it might actually be because I'm just like, What am I describing? Where are we headed into? And I'm wondering if you had that kind of experience as a developer as a as a person who was curious about where could we take this? How did you come up with ideas? How did you nurture your own ability to create in this way? Hmm,

Polly Allen  10:27  
That's a really great point. When I started working with this, the capabilities of the models we were working with were pretty limited. So we were looking to do specific things with them. And while they were a little bit capable of them, and that was mind blowing at the time, they were still not great, right? It wasn't something that you were going to just be like, oh, I need to write a letter to my congressman, I'm going to start using this technology. It wasn't there yet, by a long way. What I found has kind of nurtured my creativity around it has been anytime I have a task involving writing, I just give it a try, honestly. And it's not only helps kind of me get more creative and get more ideas about things I can try with it. It's also really given me a sense of like a more grounded idea of what it's good for, and where it's got a ways to go. Yes, I've definitely found that the more people have played with it, the less afraid they are, that it's going to turn into Skynet and take over the universe and do all sorts of scary things that you know, sometimes makes its way into the headlines. Definitely playing around with it gives you a much better sense of its capabilities, but also its limitations.

Christine Li  11:39  
Yes, thank you for that, I will share a story. It's a cute story, I think of me. I've been trying to get people to at least experiment with it once. And I'm actually surprised at how many people have not experimented with it. So I want to make the assumption that many of our listeners today have not opened it up, are really the uninitiated in terms of artificial intelligence, and I'll tell you the cute story. It was with my aunt. And she was kind of out of the clear blue sky, describing her suspicions and fears and skepticism about the security or safety of using something perhaps that was so intelligent. And I said, okay, here, give me your phone. And I opened up the app. And I said, let me know something that I can help you with. And she said, Well, now that you say that, her husband was unfortunately in a car accident, and they were having trouble with an insurance company. And she was wanting to write a letter to them, I asked her quickly about what details needed to be involved, I quickly typed it in. And within 30 seconds, a I had generated an extensive, detailed, professional sounding impactful letter that would have taken me maybe three hours to write because it was that involved. I was surprised at that. I turned the phone to her. And she said, Give that to me. Like how do we get on this, you know, it was an instant transformation of a convert. And it was really fun. And I think that is just the miracle of it, and the greatness of it. And the fact that it can help you for just an every day task and actually save you from having to do the thinking I joke that I no longer think as much because that's the truth is maybe not so much a joke. Because if I have something to write like you, Polly, I'm going to turn to chat GPT and see what it has for me, I will never use it just straight out as what it turns out. But sometimes I'll use 90 95% of it. It's really, really helpful. So I just wanted to put that piece in in case our listeners really don't know what we're talking about quite yet. And are curious what they could do today after, right? This episode is over.

Polly Allen  14:18  
Anything to do with writing, I have to admit, I was really skeptical at the beginning, probably because I'd used it when it was so bad, right? That it sort of feels like well, if this thing isn't 100% reliable? Is this going to you know, take more time for me to check it over than it is to have just written it myself in the first place. Right? And unfortunately, we're having the very first studies come out where people are doing this, you know, in the lab or in controlled settings, and it's showing like, No, you can get the same quality when you're doing office writing tasks, about 37% Faster, on average in the studies that are coming out And just for context, like when ng when the whole factories switched to steam power and the industrial revolution, they would see an increase efficiency of about 25%. Just to give you an idea of like, this is powerful 37% Doesn't sound like that much, it's a lot. It's a lot. So do check everything do spend the time, but the amount of time and effort it can, it can save you on things that perhaps are not just the most enjoyable, you know, waste a lot of your time today. It's really powerful what it can do.

Christine Li  15:33  
You just sent a shiver down my spine with those percentages, because we all know, I think we all should know how dramatic an event the Industrial Revolution was, and the impact and the societal shifts that happened because of that. And the lifestyle shifts that happened because of that. Yeah. Right. Okay. So let's now move towards the people that you're working with now, and what your vision is for what can be possible for women looking for career advancement, career development, career expansion, and making a difference in how AI continues to be brought out to the world?

Polly Allen  16:21  
Yeah, 100%. So it's interesting, having come from a much more technical background, where I didn't know how to program when I wanted to get into AI more in depth in 2018 2019. All the courses about how to work with AI and learn about AI, started with step one learn to program. So I was lucky that that didn't intimidate me and I could dive in. What I thought was crazy was I was working in a business role as a product manager. And really, those courses weren't at all what helped me be good at that job. Right? After I left Amazon, I had a bunch of conversations with some top leaders in AI and ask them, hey, for our product manager or project manager, anyone working with developers in this space? Do you really think they need a computer science degree? And often at the beginning of the conversation, they would say, Oh, yes, you know, it's really helpful. But then if I pressed them and said, like, why though, because you don't need them to write code, they don't have to read code, you know, they, they just need a common language with the developers. By the end of the conversation, every single one had been like, you know what you're right. And actually, my best product manager was a social sciences major, right. And it really is the all of those roles that have more to do with people and relationships and, and thinking about the benefits that that software delivers. All of those aren't necessarily best filled with people with with technical backgrounds. And we can definitely increase the diversity by finding the folks who do that part well, and then just giving them the tools to collaborate and problem solve and work with these more technical teams. But it can be a little more top down explaining the concepts, instead of them having to glean at bottom up by like, going back to get a master's in data science, or something like that, right. So the courses I teach as part of AI career boost, really focus on getting folks the the essential terms and understanding they need without having any technical background, and there's no coding required at all. In my most recent six weeks program, we're focusing on giving folks the skills they need to lead and succeed as AI product leaders. And as part of it, we actually build a generative AI prototype, so it's no code tools. And the idea is you can actually build a tool, whatever tool you want, just using tools like chat GPT, and some no code tools. So it's been never been more accessible to start building your own products. You can kind of just ask chat GPT at like, a recommendation engine inside of Netflix and tell me what what movies I'd like from this set of movies. And while they're not as robust and high end as the existing production AI systems. It's absolutely a wonderful way to build a prototype, get something working, show it around, maybe get some funding if you're a startup founder. So that's what we've been focusing on is really increasing people's ability to be a part of organizations that are building AI tools and products by focusing on leadership roles and getting startups up to speed and stuff as well.

Christine Li  19:48  
That is fantastic. Is there some sort of exclusivity about who gets to do what I'm thinking? You know, you're, you're very positive, you're very much alive. Either you're very much creating a movement. I'm wondering if there's pushback from people who are maybe saying, no, no, you really need this, you really need that it's just us. Are you sensing that? With the work that you're doing?

Polly Allen  20:14  
I think I don't even know if it's necessarily intentional. But I do think like, I called it AI career boost a little bit, because I thought, it feels like there's some gatekeeping going on around these careers and muted, some people just need a little boost over that gate. Right? Let's, let's provide that. There's definitely some folks who get a little defensive saying like, well, you can't have people who don't know, anything technical, doing XYZ. But once you explain, like, I'm not talking about getting people to be data scientists, or engineers, right, that those definitely have, you know, different requirements, but the business functions that help and enable them and work around them, that even lead those initiatives that they don't, they don't necessarily have to have that level of technical depth and detail to do those jobs really well. Right. So that's, that's really the goal. And honestly, I think most of the time, there hasn't been that much pushback, it's actually really that folks that are so educated, and have been in, you know, this soup of acronyms for you know, years, have kind of lost that ability to explain this stuff and English. So, so that's really where I hope I can provide like a translation layer for people and give them the tools that they need at the right level.

Christine Li  21:37  
I love it. You're the perfect person for this, I believe. Thank you. I have at least two more questions. One is, now I'm gonna, I'm gonna because I am not a robot and not artificially intelligent. You're proving it? Yeah, exactly. This is a real interview. Okay, so my thought was that typically, in these podcasts, interviews, I start with people's stories and the obstacles and the internal obstacles that have happened. And I think we kind of managed to skip by that. But I want to return to that. Because you left, what I imagined was a very comfortable, high level, interesting job, and a stable company to go off on your own. Could you describe that experience? And what helped you to make the leap to say, You know what, this is the time, this is the person, this is the space, and I better do this now. Really, really great question. For sure.

Polly Allen  22:40  
I mean, I think, like you said, it was a stable company. And I remember in November 2022 was right around, things started to look really scary for big tech. It was interesting. I left in November. And I think it was three days later, they had the layoffs at Amazon, Alexa, and all these folks came to me go, Oh, were you laid off? I was like, no, no, I left them. And it wasn't a mutual breakup it was. But I had really gotten to a point where, week after week, I was finding it was grinding at me a little more that, you know, consistently. I'm the only woman in the room consistently. I bring up concerns about things and they get kind of dismissed. And it seemed sometimes at least, like there was some gender bias behind which concerns get listened to, and which ones don't. Right. I would see candidates coming from non traditional backgrounds, not getting interviews or getting dismissed early. And it had just become that thing where you sort of get download the Sunday scaries. It wasn't really scary, but I just was realizing I wasn't looking forward to going to work on Monday anymore. So even though it was a little bit scary to take a leap at that, at that moment, sometimes you just know, you're like, No, I'm just this part of my journey is done. Right. And it also just seemed, whatever it was that called to me, it was like, No, it's going to be too late if you start later, right. And I still am like, Oh, it's a little creepy. To find out later that, that it was really, really timely that I laughed and almost immediately chat GPT kind of blew up in the public consciousness, right. So I do think there's a little bit of of knowing yourself and trusting your gut that goes into it. I think you've got to listen to those inner voices that are telling you that are whispering at you. That it's when it's time to make the leap.

Christine Li  24:41  
I love that. I love that piece of advice. Everyone remember that? Listen to your inner knowing. Always right. That little voice really does have your back. That's true. Okay, now the second question, which I do remember is for the women who are listening to us what can kind of first steps might you have for them and sort of guidelines or Yeah, things that they can start off with?

Polly Allen  25:10  
Totally. So I do have a free AI learning guide. So it's I think you're you are going to give the URL, right? Yes,

Christine Li  25:17  
the URL is make time for success. podcast.com/ai Ai, again, make time for success. podcast.com/a i Go ahead.

Polly Allen  25:29  
Perfect. Yeah. So it's, it's a list of resources, places to start, no technical background required. My very favorite is a YouTube series called Making Friends with machine learning. And it's done by Cassie Korsakoff, who is the Chief decision scientist at Google. She did this course for Google employees internally. So business users who weren't necessarily technical, and it's a really entertaining course, she talks about wine, it's really, really fun. She makes it really, really entertaining. It's about six hours. Overall, if you don't have time for the whole thing, even just the first hour or two, really gives you a sense of what's going on under the hood. And again, can help demystify and make you a little less afraid of using these systems and kind of getting the idea of how they work.

Christine Li  26:16  
I literally think if there was anything that was going to make me afraid I would call Polly and asked her to make me not afraid if you're in perfect for this. So Polly, I'm so glad you came on the show. Thank you so much for describing what you've been through what you're creating and where people can go with it next. I really appreciate your time and energy.

Polly Allen  26:35  
Oh, thank you so much, Christine. Yeah, I hope I hope your your listeners will join me on the journey.

Christine Li  26:40  
Yes, yes. Yes. Could you tell our listeners speaking of that, how they can stay in touch with you.

Polly Allen  26:45  
100%. So please feel free to join my mailing list. So that's an AI career boost.com/interested. I also have descriptions of the courses that we have available up on the website there.

Christine Li  26:58  
All right. Woohoo. All right, here we go for a whole new world. Follow Polly for more. Try out chat GBT. If you haven't listened to that inner voice because that's the original intelligence.

Polly Allen  27:13  
You always have that the non artificial intelligence,

Christine Li  27:16  
right, the genuine intelligence that we all have. And let's all help each other rise. Thank you so much, Paula. You're amazing.

Polly Allen  27:23  
Thank you, Christine. Amazing. 

Christine Li  27:25  
All right, everyone.I will see you next week. Thank you so much for listening to the make time for success podcast. Bye bye. 

Thank you for listening to this episode of the Make Time for Success podcast. If you enjoyed what you've heard, you can subscribe to make sure you get notified of upcoming episodes. You can also visit our website maketimeforsuccesspodcast.com for past episodes, show notes and all the resources we mentioned on the show. Feel free to connect with me over on Instagram too. You can find me there under the name procrastination coach. Send me a DM and let me know what your thoughts are about the episodes you've been listening to. And let me know any topics that you might like me to talk about on the show. I'd love to hear all about how you're making time for success. We'll talk to you soon!

Transcribed by https://otter.ai