Broadband: With Jessica Rosenworcel
#922 minutes

Victoria Espinel, an expert on the intersection of technology, innovation, and public policy joined Commissioner Rosenworcel for an in-depth discussion about her career as a lawyer, professor, and trade negotiator. She also discussed her time as President Obama's advisor on intellectual property, and her work at the helm of BSA | The Software Alliance. On the podcast, listeners will hear Commissioner Rosenworcel and Ms. Espinel discuss the growing impact of software on our civic and commercial lives, how we can build unbiased artificial intelligence, and what the future looks like for the use and deployment of AI.

Transcript: 

MS. ROSENWORCEL: Welcome to Broadband Conversations. The podcast where I get to talk to some amazing women from across the technology, innovation and media industries. You get to hear what they're working on, what's on their minds and what they think about the future of our digital lives. I'm Jessica Rosenworcel. I'm a commissioner at the Federal Communications Commission.

And please join me in welcoming my guest Victoria Espinel. She's the President and CEO of BSA, the Software Alliance. I'm going to let her tell you a little more about her very distinguished background, but I want to give you a few highlights.

She served for more than a decade in the White House for both Republican and Democratic administrations. She was President Obama's advisor on intellectual property. And before that she was the first ever U.S. trade negotiator for intellectual property and innovation at the U.S. Trade Representative. She is a founding co-sponsor of Girls Who Code's Washington, D.C. Summer Program and she chairs the World Economic Forum's Global Future Council on the Digital Economy and Society.

Welcome, Victoria. I am so glad you are here with me today.

MS. ESPINEL: Thank you. I'm really excited to be here.

MS. ROSENWORCEL: Ah, so let's start at the beginning. You've had such a long and impressive career in Washington, and I want to get a sense of how you got to where you are today.

MS. ESPINEL: So there's been a fair amount of serendipity in my career and a fair amount of starting over actually.

MS. ROSENWORCEL: Oh, I like that. I like that you're starting there by talking about the starting over.

MS. ESPINEL: But in terms of where it started, I went to law school here in Washington, D.C. at Georgetown. And when I went to law school I wanted to be an immigration lawyer. My father immigrated to the United States from Colombia with $5 in his pocket and 20 words of English. And I went into law school wanting to work on immigration. But I realized pretty quickly that while I still feel very strongly that our immigration system could be improved, that wasn't what I wanted to do with my career.

I have this very vivid memory of being a first-year summer and knowing that I wanted to do something different than the immigration law firm that I was working in that summer, but not knowing exactly what that was. So what I did was I got my hands on something that probably doesn't even exist today, but it was as directory of every government agency and sub-agency in Washington, D.C. And I literally started cold calling down -

MS. ROSENWORCEL: Oh, my goodness.

MS. ESPINEL: - the list of agencies with no filter in terms of issue or subject matter, calling office after office to see if they were interested in having any free legal work.

MS. ROSENWORCEL: Did you just come up with that yourself or did somebody give you that idea and pushed that book in your direction?

MS. ESPINEL: No, I came up with that myself. So I probably made 100 phone calls and no one was interested in having a free summer intern. And one of the moment of serendipity that really led to the rest of my career I called the Chief Counsel's Office of the National Telecommunications Information Administration.

MS. ROSENWORCEL: I know it well.

MS. ESPINEL: And it happened that the woman who normally answered the phone wasn't at her desk, so one of the lawyers picked up the phone and I think was amused to hear this student on the other side of the phone saying I'm willing to come in, I'm willing to do whatever you have to do, you don't need to pay me. And he said sure. Come on Monday.

And it happened that I was there when a woman named Jean Pruitt (phonetic) was chief counsel at the NTIA. It happened to be the summer that they were negotiating the financial syndication rules, so there was a lot going on. But most importantly it was a moment in time when there was all this conversation about the fact that it might some day be possible to send text or maybe even a photograph over telephone lines and what that convergence was going to do to the cable industry, the entertainment industry, the telecommunications industry. And I thought this is fascinating. And I left law school wanting to become a telecommunications lawyer. I actually never became a telecommunications lawyer, but that experience at NTIA did in many ways lead to the rest of my career.

MS. ROSENWORCEL: Ah. All right. So then how did you wind up negotiating on trade policy?

MS. ESPINEL: So I - immediately after law school I actually went to New York to do structured finance because I wanted to live in New York City more than I wanted to do anything in particular, and that was the way to get to New York. But I knew I didn't want to do that long term, so I left my job at a big law firm in New York and I moved to London to go to the London School of Economics to get a degree in telecommunications law.

But what I found out when I got to London - you know, quit my job, took all my savings, put it into this move, I show up for my first day of LSE to register for telecommunications and am informed that the telecommunications professor is on sabbatical for the entire year and they are not offering telecommunications law.

So I quickly decided to pivot and I signed up for an Internet law class. And in the Internet law class they were talking a lot about copyright. And so I thought that sounded interesting and so I learned copyright. And I got out of LSE with a dual degree: one in Internet and copyright law and one in the Great Gatsby and whether or not the Great Gatsby was an accurate depiction of a prohibition-era gangster.

MS. ROSENWORCEL: I'm getting the serendipity part as you talk.

MS. ESPINEL: (Laughter.)

MS. ROSENWORCEL: I really am.

MS. ESPINEL: So then I had a decision to make about where my career was going to go from there and ultimately I ended up going to Covington & Burling and working on copyright law issues in Eastern Europe, Western Europe and the Middle East and Africa over the next few years. It was an absolutely fascinating experience, particularly being in Eastern Europe as they were coming out of the Communist regime and putting together their legal system. It was a fascinating time to be there.

MS. ROSENWORCEL: It's starting with a blank slate, right?

MS. ESPINEL: It was really incredible. Part of what was really incredible was meeting and training all these very young lawyers coming out of law school, so essentially teaching international copyright law in order for them to be our local counsel and how - and then watching many of them go on to be preeminent lawyers and scholars in their country, but bringing together that group of young, energetic, ambitious people was really exciting. And doing it in the context of them also thinking about how their country was rebuilding itself was fascinating. And I have -

MS. ROSENWORCEL: Wow.

MS. ESPINEL: - many stories from that.

But when I was at Covington I got a phone call from USTR saying that they -

MS. ROSENWORCEL: The U.S. Trade Representative, right?

MS. ESPINEL: Yes. When I was at Covington I got a phone call from the U.S. Trade Representative saying that they were looking for someone who knew something about intellectual property to come to USTR. And I was very torn, because I loved living in London. The job I had was fascinating. But at the - I decided I would go to USTR for two years and then leave Washington again and come back to London, but when I got to USTR I found that I absolutely loved it.

I loved representing the United States Government. I thought the work that USTR was doing was very important. I was there with Bob Zoellick and then Rob Portman and Susan Schwab at a time when the United States was negotiating many trade agreements, and it was a fascinating experience. And my first experience in public service, which I think is such an important aspect of any young lawyer's career to the extent they have the possibility to do it. So I had a great experience at USTR.

I left USTR to become a law professor and taught international trade and intellectual property. And then I left that because President Obama asked me to come to the White House and be his advisor on intellectual property and start a new office inside the White House to coordinate the agencies on intellectual property, which was a fascinating and extremely gratifying experience in many ways. And then I left the White House to take the fantastic job that I have today.

MS. ROSENWORCEL: Ah, what a story. You're crossing continents, you know? This is neat.

So tell me not just about BSA and the Software Alliance, but why software is so important for modern life.

MS. ESPINEL: So part of the reason that I took the job that I have today is because I don't think there's any industry that is going to have a bigger impact on our economy, but also on our society. We represent the software industry around the world. We are a global organization. We have offices around the world and we do policy in real time around the world and our issues include privacy, cyber security, artificial intelligence, digital trade, intellectual property, data being kind of the underpinning of many of the issues that we're working on.

But the reason that this is important is twofold: One is jobs and the economy. So just in the United States the software industry creates over 102 million jobs. We -

MS. ROSENWORCEL: Wow.

MS. ESPINEL: - contribute over 1 trillion to the GDP. And it's not just jobs in the software industry. I think one thing that people don't always think about is the fact that software creates jobs in every single sector. So at this point every single sector uses software, is dependent on software, is a software sector in many ways. So health care, financial, manufacturing, all of these sectors are sectors that are creating software jobs and where people need to have the training and the skills to do those jobs.

MS. ROSENWORCEL: Yes, I can't think of anything in the economy that is not impacted by software today and won't be even more impacted by its growth in the future.

MS. ESPINEL: Exactly. And then on top of that the impact that software has on society and how we live our lives is enormous and really important. And so there were two main reasons that I took the job that I have today, but one of them was to be a part of that conversation because of the enormous impact that software is going to have on our society.

MS. ROSENWORCEL: So tell me, what does a typical day look like for you with all of these issues that you cover?

MS. ESPINEL: So my days tend to move around. They generally kick off around 7:15 a.m. and I generally - if I'm in town I try to be home at 7:00 p.m. And then like many working parents work a little bit from home once the children are in bed.

MS. ROSENWORCEL: I'm totally familiar with that.

MS. ESPINEL: But, you know, within those hours it's a variety of things. So part of that is talking to our companies. You know, our companies are our number one priority. I spend a lot of time working with the team on the policy issues that we work on, which are evolving quickly and -

MS. ROSENWORCEL: Very fast.

MS. ESPINEL: - complicated and important. And then I spend a lot of time talking to governments. So here in the United States, but also in Brussels, Germany, U.K., Japan, India, Singapore, Seoul, Beijing, all the markets where we operate. So one of the consequences of my job is that I tend to be on planes quite a bit because our issues are not -

MS. ROSENWORCEL: They're worldwide.

MS. ESPINEL: Our issues are worldwide. And what happens in Europe has an impact on what happens in the United States. What happens in the United States has an impact on what's happening in China or Japan. They're very interconnected.

MS. ROSENWORCEL: So you mentioned data, cyber security and artificial intelligence, and it's that last one I want to ask you a little bit about. I'm struck when I read about AI, that on the one had you have this ominous notion of robots that are coming in our future. And on the other hand you have this enthusiasm for the efficiencies it will bring and the growth it will offer for our economy and the change that we'll see as a result.

And sometimes I just wonder between those challenges and opportunities what does it really look like? What do you think the future of AI involves?

MS. ESPINEL: So I think there are different kinds of AI, and that is one of the things that gets lost sometimes in the discussion. And the type of AI that our companies create is mostly focused on helping people make better decisions. So not making those decisions for them, but giving them more information so they have a better ability to make a more informed decision. And so whether that's doctors working patients or farmers trying to determine how much water and fertilizer they need to maximize their crop, or manufacturers who are trying to respond to customer demand and redesign their products, a lot of what our companies do is try to give those people the tools that they need in order to make better decisions.

But I think there are definitely going to be challenges that we're going to see coming in the future from AI. I also think we need to have more conversations and discussions about how to maximize the societal benefit of AI. And I know those conversations are starting to happen -

MS. ROSENWORCEL: So what would those conversations -

MS. ESPINEL: - but I think that's really important.

MS. ROSENWORCEL: Yes, what would those conversations look like?

MS. ESPINEL: So the way I think about it, I think there are two things to focus on: One is how you build AI, but the second is how you use AI. And right now I think there's more discussion around how you build it and create it and less on how you use it. And both parts are important.

So to take them in order: In terms of how you build the AI, I think it's incredibly important that it be as accurate and as unbiased as possible. And that I think really comes down to three things: It comes down to people, data and results. And what I mean by that is I think in terms of people I think it is extremely important that as AI is being created and trained it is being done so by a group of people that have a divergent set of perspectives.

MS. ROSENWORCEL: Right.

MS. ESPINEL: And there are lots of different things that need to happen to make sure that that is true, but I think making sure that that wide set of perspectives is there at the creation is incredibly important.

The second thing is data, making sure that the data that is going in is as accurate and comprehensive as possible. And so we are in Washington, this is probably a place to mention that is a place where I think the government could do more to help. I think the government has lots of data and making that available to the private sector and the public sector in a very open accessible way is -

MS. ROSENWORCEL: It's so true.

MS. ESPINEL: - incredibly important.

MS. ROSENWORCEL: I mean, we have amazing data sets at this agency that are not in machine-readable format. And to me there's so much loss in that because it's now only about what we can do with it, but what if we were to put it out to a broader population and let them, you know, crunch through it and look at it and identify what patterns that could change our policy making here? And I think it's a lost opportunity that we don't make more data open and available as to the government.

MS. ESPINEL: I completely agree with that, and I know that there has been legislation that's been passed to try to improve that. I know it's something that the U.K. government and governments around the world are looking at how to do better. It is partly so that other organizations can look at the data, discern patterns from it and then come up with new solutions, but it is also because those that are already creating solutions in AI, the more data that they have -

MS. ROSENWORCEL: The better those solutions are.

MS. ESPINEL: - the better the solutions will be.

MS. ROSENWORCEL: Right.

MS. ESPINEL: - and the less possibility of bias or skewed data sets or incomplete data sets. So it's - I think the issue of open government data is incredibly important when we're trying to achieve AI that is as unbiased and as accurate as possible.

MS. ROSENWORCEL: Is that difficult to pursue worldwide, or do you sense that there's a worldwide movement towards that?

MS. ESPINEL: I think - yes, it is difficult, but I think it's more because of the lack of understanding than because of opposition to the concept. So one of the things that we do is talk to Governments around the world to try to help them understand why this is important. And I think once governments understand it, they tend to be very responsive to it. So it's more of an education process.

But then the results are also important. And so - and this really is something that falls on the companies to ensure that the results that are coming out of AI also are as un-skewed as possible, and to the extent that they are seeing problems there to go back and try to fix those problems as quickly as possible.

MS. ROSENWORCEL: So do you mean skewed as a function of incomplete data sets or because there's bias associated with their collection?

MS. ESPINEL: Or because there's bias because of incomplete data sets. Often those two -

MS. ROSENWORCEL: They're related?

MS. ESPINEL: - elements are related.

MS. ROSENWORCEL: Yes.

MS. ESPINEL: Which again is why having data sets that are as comprehensive as possible are really - is really important.

So that's all about how to create it. And we have a lot of thoughts and we have recommendations on what companies can do to try to ensure that the AI that they're training and creating is as unbiased as possible. But then there's a whole conversation about how we use AI. Not just how we build it, but how we actually then deploy it to try to make sure that we're - it's being used in a way that will broaden inclusion, not just exacerbate discrimination, but actually actively and affirmatively broaden inclusion. And I think there is enormous potential for AI to do that.

We're already seeing some of that in apps that are being used to help people that are visually impaired understand what is happening in their surroundings in real time. We're seeing that happen with apps that could help position who - with autism to be able to recognize facial expressions, which will completely change the way that they interact with society and interact in the workforce. And that's just the tip of the iceberg.

But I think it's incredibly important that industry and government be thinking really affirmatively about how to use AI to broaden inclusion so that we ensure that what we're creating is accurate, as unbiased as possible and then what is it going to be used for? It will affirmatively broaden inclusion in our society.

MS. ROSENWORCEL: Right, and do it now at the start, in the dawn of the AI age.

MS. ESPINEL: Absolutely. Absolutely. this is the moment to do it.

MS. ROSENWORCEL: Oh, sounds good. I like the optimism, too.

So if you've listened before, you know what's coming next. I want to ask you a few questions I ask everyone at the end of the conversation.

And this one involves going back some years. What was the first thing you recall doing on the Internet?

MS. ESPINEL: So the first thing that was meaningful was being in London and being away from my family and sending them long email descriptions of what I would see as I walked around London and got to know my new surroundings.

MS. ROSENWORCEL: Oh, that's fantastic.

MS. ESPINEL: And that was very important to me as a way of connecting with them.

MS. ROSENWORCEL: And that they could immediately receive it was so novel at the time.

MS. ESPINEL: Yes. Yes, it was - I mean, there are so many other ways to stay connected, but at the time that was - the speed and the detailed descriptions as opposed to, you know, a phone call or ways of communicating, it was really meaningful to me.

MS. ROSENWORCEL: It's hard to remember how extraordinary that was when that first came about. So what was, more mundane, the last thing you did on the Internet?

MS. ESPINEL: So in the taxi on the way over here I did a few things: I looked up the weather because it's more humid out there than I had realized it was going to be.

MS. ROSENWORCEL: I hear you.

MS. ESPINEL: I sent quite a few emails and texts to people in my office. I read an article in The Economist. And I paid someone using a mobile payment app.

MS. ROSENWORCEL: That is some serious multitasking in the back seat of a car.

MS. ESPINEL: (Laughter.) Oh, and I looked up directions to the doctor's appointment that my child is going to when this podcast ends.

MS. ROSENWORCEL: Believe me, I understand. I understand. You're doing a lot of things at once.

MS. ESPINEL: (Laughter.)

MS. ROSENWORCEL: All right. Now this is bigger picture: What do you think the future of the Internet and digital life should look like?

MS. ESPINEL: So I think with anything that has a big - as big an impact on society as the Internet does and that software does, I think it's really important that we; and by we I mean the industry, but the government, be thinking collectively about how to maximize the societal benefit. Not just the economic benefit, but the societal benefit. And I am optimistic that if that is done thoughtfully and truly collectively that the future of the Internet and the future of our society as a whole will be bright.

I think there are a lot of ways that have been demonstrated recently - how do I put this? I think the impact of the Internet on politics has played out very publicly in ways that are very entrusting, and clearly the opportunity for the Internet, like any tool, to be used for divisiveness exists. And that was always going to be the case.

And the way that anything, any tool or communication from the printing press to the Internet can be used as a tool for divisiveness. But I think the scope and the reach and the ability of the Internet to put power in people's hands is very different than other communication tools that have been abused in the past. And I think the transparence that that can lead to is very different than communication tools of the past.

And so it is my hope but also my expectation that because recent events have put a spotlight on ways that information is being abused and being manipulated, but I think companies are also responding to that and responding to it in good faith and with a real will and effort to try to address that. I'm optimistic that the negative effects of it can be ameliorated and the really positive effects of putting power in people's hands and the transparency that go with that will be maximized. But that is - it is going to be - I don't want to minimize the difficulty of figuring out how to do that in a way that is inclusive, that doesn't lead to inadvertent censorship and that is effective. That's a really difficult challenge to figure out, but I am also highly optimistic that it will be figure out, probably not tomorrow, but I think there are enough - I think it's going to be difficult to figure out, but I'm also optimistic that it will be.

MS. ROSENWORCEL: Yes, there's no minimizing the difficulty, but I like your optimism.

So that wraps up another episode of Broadband Conversations. Thank you for being here, Victoria. And thanks to everyone for listening. Take care.