S5 E28 | Recruiting Community: Lindsey Zuloaga and Transparency in AI
Chris Hoyt, CXR
So I think that that takes us to the question everybody’s dying to know Lindsey, are you? There’s two kinds of people in this world. You’re either you’re either cake or your pie. So, Lindsey, are you cake person? Are you a pie person?
Lindsey Zuloaga, Hirevue 0:15
Oh my gosh, I can’t have gluten so technically really neither. I have been known to ask for cheese cake, which is really pie. On my birthday.
Chris Hoyt, CXR 0:29
It is really pie.
Lindsey Zuloaga, Hirevue 0:30
I just had this the other day. So as a kid, I wanted my mom to make her really good cheesecake for me on my birthday rather than the traditional cake.
Chris Hoyt, CXR 0:40
Oh, man. Now is it a regular cheesecake? Or is that like a super fancy cheesecake?
Lindsey Zuloaga, Hirevue 0:44
It’s pretty it’s pretty standard. But delicious.
Chris Hoyt, CXR 0:48
Crushed is the graham cracker crust.
Lindsey Zuloaga, Hirevue 0:50
Yeah, so now now I know I gotta get the gluten free version but but still sometimes do.
Chris Hoyt, CXR 0:56
Oh man. I’m definitely pie although I had cobbler last night, and this has nothing to do with anything. I don’t think. I don’t think that’s either cake nor pies, or is everything cake or pie?
Lindsey Zuloaga, Hirevue 1:07
It seems physics question.
Chris Hoyt, CXR 1:13
That’s why we bring smart people on the show to answer real questions that matter. Alright, are you ready to get going?
Lindsey Zuloaga, Hirevue 1:20
Chris Hoyt, CXR 1:21
Okay, here we go.
Welcome to the CXR channel, our premier podcast for Talent Acquisition and Talent Management. listen in as the CXR community discusses a wide range of topics focused on attracting, engaging and retaining the best talent. We’re glad you’re here.
Chris Hoyt, CXR 1:53
All right, good day, everybody. I’m Chris Hoyt fan of pie and your host for today’s recruiting community podcast episode. If you’re new here, that’s okay. We actually hope you’ll make yourself at home and jump right in if listening live on Twitter, LinkedIn, YouTube, or Facebook. You can do that by just saying hello in the chat stream, putting your social or professional profiles in there, or even asking today’s guest any question that you’d like, I’ll do my best to include that here with an overlay and a shout out for you. So if you’re watching it, you’re gonna see that on the screen, and if you’re listening, we’ll say your name. So don’t be shy. Also, this show so that you are reminded is powered by love. That’s right, love. It means we don’t have any sponsors or people that actually pay or contribute in a monetary way to the show. We just sit down once a week. And we actually talk to people that our community members and leaders have said they’d like to hear from or learn more about, so please hit the like and subscribe buttons, and visit our community online at CXR.works to literally see how you can help determine who should be next, and how you might even be able to join us for a chat or two. Now, I am happy to introduce a new friend to today’s episode. She has an intellectual force to be reckoned with on the topic of data science and AI, and was a recent CXR community panelists where we discussed with about 40 companies or so the importance of transparency and ethics within the realm of AI. And she’s a new CXR talent solutions community member herself. Please welcome to the show, Dr. Lindsey Zuloaga, there we are. Hello, Lindsey. How are you?
Lindsey Zuloaga, Hirevue 3:25
Great. Thanks for having me.
Chris Hoyt, CXR 3:27
We’re really excited to have you on but before before we jump in, I’ve got three or four questions I want to ask you. We’ve covered pi, we’ve covered K, thanks for being a pie person, by the way. Why don’t you give us sort of an escalator pitch about who is Lindsey? And why why should we be paying attention to what Lindsey has to say?
Lindsey Zuloaga, Hirevue 3:46
About me, not HireVue?
Chris Hoyt, CXR 3:49
Just about just about you, like get a little bit of background, what you do at HireVue would be helpful. Yeah.
Lindsey Zuloaga, Hirevue 3:54
Yeah. So I always start with my pretty far back. I’ve always loved math. And I didn’t really understand why math was important until I took physics. So physics was a huge kind of force in my life. I had a really great teacher in high school who really turned me on to this is how math is used in the real world. And I thought that was really cool. So I was the first person in my family to go to college. I was very intimidated by studying physics. I it was what I wanted to do, but I was very nervous about, you know, kind of succeeding at that. But I kept chugging along, finished. My undergrad, ended up going to grad school ended up getting a PhD and doing research and physics. And when I when I started my PhD, I really thought I wanted to work with my hands. I didn’t want to sit at a computer all day. So I was an experimental physicist and aligning lasers and working with nanoparticles. But much to my surprise what I enjoyed the most was sitting at my computer and analyzing data.
Chris Hoyt, CXR 4:57
You liked that more than lasers?
Lindsey Zuloaga, Hirevue 5:01
Yeah, there’s a little more. When something’s wrong with your with your computer program, you can actually figure out why most of the time. But in the real world, there’s so many variables. And I had a lot of holes burned in my shirts from lasers. But um, a lot of time spent in dark rooms. But I Yeah, so when I transitioned, I did, I did a postdoc as well, in Germany. And then when I transitioned into industry, I was kind of, you know, in this weird position where in some ways, I was overqualified for a lot of entry level jobs, because I had a PhD. And I’ve been working in that academic world for years, but also under qualified because I’ve never had a job before. I think I really underestimated the value of connections, I kind of thought, you know, I might resume or CV which you’d say in academia, speaks for itself. And I, you know, have been doing pretty well in academia, and it’s very competitive. So I should be fine in the real world, as I as I put it, but when I started applying for jobs, I was shocked at kind of how broken the whole system is. And I know a lot of people who are in this position as well, where they tell you, it’s really important to curate your cover letter and your resume to that particular job. And so you go into some applicant tracking system, you upload those things, you’ve already spent hours working on that. And then you have to manually re enter every part of your resume, again, into the system. It’s just painstaking. You press submit, and often you never hear anything ever again. So that was a pretty awful experience. And I know so many people that went through that who are now in good jobs, and they’re successful, and they’re doing really well. And so you have to think how many people pass them up? Or did they just never get seen by by these companies that kind of missed an opportunity and finding talent?
Chris Hoyt, CXR 6:59
Did you graduate yesterday? Because there’s a lot of companies that’s still going on?
Lindsey Zuloaga, Hirevue 7:02
Yeah, no, no, well, the company wasn’t too long ago. I like to think it was more recent than it really was. Like yesterday, but no, no, so I think that, you know, that experience, I got a, I got into data science at a really good time when it was still pretty new. And there wasn’t data science majors. So a lot of people from different STEM backgrounds coming in, you know, using data in the real world, or writing code to analyze data, all that stuff that I like, and, and so it’s kind of an opportunity when I got into it, and then you know, learn about what HireVue is doing. And it’s really excited because it tied to my own experience. And, and so, so yeah, HireVue is a talent experience platform. We are kind of known for the asynchronous video interview, which are our founder had that idea in 2004. When he he he had the idea to ship everyone webcams so they could record themselves taking an interview.
Chris Hoyt, CXR 8:04
I remember that. So I was in this space, when that was still going. You know who else was in the space Gerry Crispin was in the space say, Hello, Gerry just popped by.
Lindsey Zuloaga, Hirevue 8:12
Chris Hoyt, CXR 8:16
So I remember when this was like, unbelievable. Nobody, nobody could believe that an organization was shipping cameras. Right? Well, I know, we’re totally off topic. But we’re shipping webcams out so that people could, you know, make make a better sort of pitch for themselves and sort of get to the front of the pile.
Lindsey Zuloaga, Hirevue 8:35
I love that. That was brilliant. But at the same time, nobody had the experience of talking to, to dead air. And so and so the whole skill set of being able to just talk to yourself and answer questions, and still be you know, engaged in whatever was a skill, I think that a new generation had to learn.
Chris Hoyt, CXR 8:58
Oh, yeah, well, that’s a good call out right. really struggling to talk to the camera. I don’t know. We’re all pros, especially given the last two years but yeah,
Lindsey Zuloaga, Hirevue 9:06
Yeah, it is. It is a different I took a higher view in my experience of applying for jobs as well. And I remember being pretty thrown off and just kind of what is this? Right? And, and I think now it’s more and more people have experienced with HireVue or just, you know, talking to a camera recording themselves and all that the technology is lends itself really well to doing that.
I still want a friendly avatar to be asking you those questions. I just say, you know, far from that, but you could say well, would you like to have some semi human thing, you know, talk you and make you feel a little more comfortable?
We’ve had the discussion of it’s been hotly debated. In kind of, you know, we don’t want to be in the uncanny valley, as they say in AI Have you got some creepy? But then you have these deep fakes that can add ask questions. Yeah, I think
Chris Hoyt, CXR 10:07
You can create your own slit that like they could pick what celebrity gets to interview them. So I’d like a Tom Cruise deep fake. All right like Emily Blunt, deep fake to interview. My interview with Emily Blunt would probably then be like four hours long because we just chatted up it’d be fine.
Lindsey Zuloaga, Hirevue 10:23
I’m sure she would love you.
Chris Hoyt, CXR 10:24
Yeah, of course, the deep fake Emily would love. Love. Okay, so on to track Lindsey. So you are quite literally a scientist. And we have had quite a few conversations around AI and the topic of transparency within that space, right and ethical AI. But let let me ask you why you think you sort of feel that transparency is so important when we talk about AI within the world of talent within the world of interviews and hiring decisions. What Why does Lindsey feel like it’s such a big deal?
Lindsey Zuloaga, Hirevue 11:00
Yeah, I think. So it ties back to my experience somewhat right with kind of, you know, it’s a black box, when you when you apply through an ATS and you never hear anything back, you don’t even hear back? No, you didn’t get the job, or here’s how you did or anything like that. And so, too, for the candidate to understand what’s going on. I also, you know, I’ve been in this space for about six years with HireVue. And the change I’ve seen in that time is huge, kind of going from a little more secretive, to protecting our IP to really opening up everything that we do, because we want to make sure people don’t assume the worst, which was often happening, right? It’s creepy or using a video, what are they looking at? Could they be not hiring me because of my skin color? Because my eyebrow moved a little weird things like that. We those kinds of questions and concerns really have pushed us more and more to open up. And be very clear. And it’s been a good because it’s it’s pushed us to also question certain things that we’ve done. Now, you know, like, if it’s concerning, there are use of video as a good example, it was really not offering a lot of predictive power, and it was causing more concern than it was worth. So now all of our models are just based on language. So there’s been kind of this healthy kind of, you know, a lot of communication between us and our expert advisory board and academics and candidates and audits that we’ve done, that have kind of formed our direction, and going, you know, really transparent more lately with our explainability statement. So this is a document who the short version is I think it’s about 20 pages long, but we there’s a longer version as well. But we kind of walk through, you know, how we approach AI, how we use it in our system, when you certain interviews that are recorded, are scored. And here’s how we train those algorithms. Here’s the data that we’re using, here’s how we test for bias. Here’s how we mitigate bias, if we find it. All of those things are outlined for anyone to read candidates customers, you know, any anyone interested in the system? So, so we hope it answers a lot of questions that people have, or concerns that they may have.
Chris Hoyt, CXR 13:25
Yeah, and I think it’s a I mean, we sort of say tongue in cheek when we when we say it’s the short version, it is the short version for real and 20 something pages is no joke, it is worth the read though, to sit down and go through. And for those who are listening, if you’re watching live, we’ve got the URL up on the screen. But for those who may just be listening, it’s just CXR.works/documents/HireVue-explainability-statement, and there’s dashes between each of those, but we will put it in the show notes to selection will be a real link. But you can check that out. It is worth going through and sort of exploring that document and the work that’s gone in on it. And I imagine that this is the first iteration of this right, I suspect we’re gonna see maybe a couple of versions of this as this grows, and the knowledge and awareness increases.
Lindsey Zuloaga, Hirevue 14:07
Yeah, it’s a living document. So as things change and grow, we will update it. And you know, we also have some audits out there. If you if you Google around, you can find a couple of audit reports that we’ve shared publicly as well. And I think that’s a great step. Audits are interesting, because no one knows really what an AI audit is. And we there we did a few that kind of approach it from different lenses, but those are interesting.
Chris Hoyt, CXR 14:35
Yeah. Yeah. I love that. So let me ask you, what do you Lindsey, what do you think are some of the ways people from your space from our space are actually making an impact or a difference outside of obviously the explainability statement and the tremendous amount of work that’s gone into that but but making a difference on the topic of transparency and ethics in artificial intelligence?
Lindsey Zuloaga, Hirevue 15:01
Yeah, I think, you know, five or six years ago, the that topic was just not even a part of the data science toolbox, really. So it is changed entirely where that is, that’s a part of data science education. There, it’s, you know, it’s actually not as simple as it might sound like. There are even mathematically many different definitions of fairness and bias. And it’s, it’s impossible to satisfy all of them at the same time with most real world data. So you kind of have to pick which ones are important to you. Outside of just the math, there are many different ways of looking at fairness, our audit with Kathy O’Neill is a good example of kind of looking at all the different stakeholders that interact with the system and concerns that they might have. That’s not just mathematical, it’s, you know, kind of the process and functional, kind of bigger picture or holistic picture. But we’ve seen, you know, this become a really big part of the conversation in obviously, a lot of realms that hiring is a big one. I it it is, you know, AI is being used at many different places in the funnel, and I’m kind of talking about the assessment piece. But even just how do you source candidates? And that sourcing is an interesting area, I think, because when we assess candidates, they’ve applied for a job. And so there are there are guidelines from the EEOC. And there are many, you know, precedents that have been around for decades around kind of how we define fairness in that sense. But when you’re talking about before, application, it’s kind of a different world. And it’s, you know, you have like, kind of like programmatic advertising type, almost like targeted marketing type issues. Where are we saying that if someone looked at handbags, we’re never going to show them a software engineer role? Like, who is even regulating that? Right. Like, that’s, that’s just like a totally new question. So I think, you know, there are a lot of players in this space. And we had I sat down recently with you, with Athena from HiredScore, and Barb from Sapia gives two really great examples, Sapia have put out a really great document on how they approach their fare framework. I know that that we have, we have a couple of players in the space. Also from pi metrics, people that have like, done audits, or are testifying for things like this New York City Law, we’re seeing a lot of a lot of talk in the regulatory space. So I think it’s very important that we are a part of that conversation. And we know the realities of this area of the kind of the whole area. And I think New York City is an interesting point. There are some things that lawmakers will write into a law with good intentions, but we need to weigh in. In this case, there’s a temporary waiting period.
Chris Hoyt, CXR 17:59
Yeah, and I just I’m gonna say what you’re being so nice about some of our lawmakers are just dumb. Like when it comes to the topic of AI, right, or it comes to the topic of even hiring and employment, some of them are just dumb. And so I think it’s super important to get brilliant people in the space together like you or Athena, Barb, like Frida and many others to come together and say, Look, this is what this actually means. Let us help to educate you. And that’s why we were excited to see the HireVue team during this talent solutions community piece that we put together, and this is an effort for those who maybe haven’t heard, yeah, Gerry’s way more elegant than I am. But the idea here is that the talent solutions, communities, establishing a baseline on this topic and a few others sort of pulling the industry together to sort of help define these challenging issues right together and getting momentum. There’s a live meeting coming up. I think in Chicago, we’re doing that. Love to say live meeting again, it feels so good. But we’re we’re bringing those guys together. And in no particular order. We’re talking about obviously HireVue and some brilliant folks we’ve pulled in Seekout, Oracle, iCIMS, Hiredscore, The Mom Project, Gr8 People, Sparc Start. We’ve got Higher ed, College Recruiter, Shaker Advertising, Phenom, Biometrics, Social talent, Aspen Technology, Cielo, Paradox Compa, and Outcast. And I think there’s a partridge in a tree that we’ve added to that as well that we’re excited to come it’s about 21 companies that have come together, we’re being incredibly particular with it, but it’s co founder level that are doing exactly what you’re talking about. Lindsey is pushing to raise awareness and drive those legislature, that legislature we’re in the right direction, right to help provide some guidance and a resource and a hub of information and experts. I don’t know, Gerry, if you want you want to add to that, that I missed some,
Lindsey Zuloaga, Hirevue 19:50
You know, just just one observation in terms of the conversation we’re having is that is that when we look at all of these issues that around transparency a lot Out of them or, in the past were designed around what we would do to touch the candidate, if you will. And now, a lot of this has to do with how the candidate perceives what we’re doing to them. And, and so it goes back from a, from an academic point of view to that the set of validity is that because when you when you learn in academia, about validity, you learn that there’s various kinds of validity from face validity up to predict it. And the scientists love the idea of the predictive validation approaches. But fundamentally, what’s what’s coming to the fore with transparency is face validity. It is the perception of the candidate about what is happening to them. And their perception of that being fair or not. And that’s why we have so many different approaches to all of this. But in the end, it’s the candidate who’s going to perceive recruiting as something that is a fair competitive practice that they can engage in, that doesn’t screw them over. And and if we can get to that level, because of that transparency, we are literally moving the industry and the industry should certainly have a strong voice, in addition to all the other folks who want to weigh in on this.
Chris Hoyt, CXR 21:28
Yeah. So Lindsey, let me ask you, What would be a good resource? I guess, for people who who perhaps want to dive in deeper, I obviously start with the explainability statement, right to work your way sort of through that. But is there anywhere, podcasts book anything you might recommend to, let’s say, a talent leader, or recruiter who just wants to learn a little bit more about the topic of, of ethical AI, or making AI a better person in the hiring process? Anything you recommend?
Lindsey Zuloaga, Hirevue 22:01
Well, obviously, your community is a good one. Shout out. I think, you know, we we participated with the Data Trust Alliance. And that’s a that’s something that I think, within the last year, they’ve come out with kind of guideline framework at you’re looking at vendors in this space, what questions do you want to ask? There’s a lot of I know, as Athena mentioned, the World Economic Forum has a council on AI. I also think just kind of keeping an eye high level on the regulatory legislative space, there’s a lot going on, and it is going to be really hard to keep up with all the individual states. But it looks like that’s the way it’s going to go. So you’re going to see a lot of the same themes that probably are coming out of Europe that a lot of states will adopt. So it’s already starting to happen and a lot to keep up.
Chris Hoyt, CXR 23:01
It’s a good it’s a good call out because we do see a lot of from a data privacy standpoint and fair and we’re seeing a lot of it sort of developing in in Europe, and specifically out in the UK and sort of come over slowly but surely, into our practices here. So it is a good call it to keep an eye on on across the pond, so
Lindsey Zuloaga, Hirevue 23:20
to speak. And Lindsey, have you come across PIA I think it’s a group. That vizier has been heavily engaged with Lexi Martin, who.
Chris Hoyt, CXR 23:33
We haven’t actually pulled that up. Gerry, if you want to talk about it, I’ll pull it up from the LinkedIn
Lindsey Zuloaga, Hirevue 23:41
Yeah, I think it’s another effort of again, our industry trying to do something in a collaborative basis, specifically on some of the technology issues. And I just, I just connected Frida with Lexi Martin, who is heading up that group. And she’s been a long term PhD, who’s very, very much an advocate in this area. So I’d be happy to connect you as well.
Great. Sounds great.
Chris Hoyt, CXR 24:15
Yeah, here we go. I just pulled up her profile, share it, she’s on I’ll put this on here in the chat real quick in LinkedIn, anybody’s interested.
Lindsey Zuloaga, Hirevue 24:26
So yeah, finding ways that we can find more collaboration with folks who can agree that we should be moving in a particular direction, I think lends kind of a catalyst field to to building a voice for our industry.
Yeah. And I think most people in the industry are broadly supportive of the regulation that’s being brought it brought to the fore. I mean, it’s, it’s all it’s kind of saying, Hey, if you’re using AI in this space, you still need to be following the rules. And most of us are hopefully so But then then there’s weighing in on kind of the details like, like that that aren’t really, I mentioned the 10 day waiting period, if you make people wait 10 days, they might not get the job, right. Like there’s a lot of things that that are kind of throwing sand in the gears but in the wrong place. And so I think it’s really important for us to be active and making comments and kind of educating legislators on that.
Chris Hoyt, CXR 25:23
Cool. Yeah, there you go through that on there. So so there’s it’s people intelligence.org Is we can find out more if you’re watching. If you’re watching on the video stream, instead of just listening, we had put Lexi Martin’s LinkedIn profile up there, she’s been a guest on the show as well, you can listen to past episodes there, Lindsey, for the sake of time, I’m going to ask you the question that we ask everybody. Before we close out the show, if you were going to write a book on this topic, what would the title of that book be?
Lindsey Zuloaga, Hirevue 25:53
I think I would say because I’ve been on both sides of this hiring, this hiring interaction, I think of it as I would. I would name my book Needles in the Haystack. There’s not just one meal, but there are many, how are you going to find find those people, right, and we talked about how things how things used to be and sometimes people think it’s still this way, where you go down and you get an application, you fill it out, and you apply for a job that red brick and mortar place and, and it is just that world is so far gone. Every every position has hundreds or 1000s of applicants. And that is just the reality of what we have to deal with. So many of these solutions aren’t perfect, but we there is no perfect solution. So how do you kind of you know, what’s best for those needles in the haystack? And what’s best the person who’s trying to find the needles to kind of get rid of of people who aren’t a good fit and find and find the right fit? is kind of the way I think about it a lot.
Chris Hoyt, CXR 27:01
I like it. I love it, actually. So now now I’m really going to put you on the spot who gets the first signed copy.
Well, since you asked, I guess it’s you. This whole idea. Career was all because you asked that question.
Man. I’m just glad I could contribute in some way. I’m just excited about that. Nobody ever gives me the copy. We’ve been asking that question for a year. Nobody gets. Thank you, Lindsey. Appreciate you. All right. Well, speaking of appreciate you, we are so grateful for your time today. Thank you for cutting out of your busy schedule to just give us a couple of minutes and share your insights. I hope that some folks connect with you directly online. And that this has been official. I really look forward to seeing you at some of the TSC events coming up.
Lindsey Zuloaga, Hirevue 27:47
Yes, thank you so much. It was fun chatting with people.
Chris Hoyt, CXR 27:50
All right, good stuff. I’m gonna put you both back in the greenroom. So hang out, please don’t go anywhere. And there we go. And I’ve got my little list here. I’m gonna share with you coming up at CXR.works/events. We have an employer branding meeting, that is tomorrow actually on August 3 for our members. We are also doing a solution spotlight on Crosschq with CEO Mike Fitzsimmons. Now that’s on the fourth. What is that, that’s two days away. And we’ve got Charlie Franklin, the CEO of Compa who’s going to be on the ninth so those will be a lot of fun for our members sit down and kind of talk about what’s what with both of those services and solutions providers. We are bringing in also Andrew Flowers. Now he is the lead labor economist at Appcast. He’s also the director of research over Recruitonomics. He’s going to be on the fourth and he’s gonna be talking about what the rest of this year looks like through his eyes and what we should be watching out for. Also, you can find all this also at CXR.works/podcast. I had we have an interview with our friend Madeline Laurano. She’ll be back on the ninth now she’s going to share with us what recruiting Tech Trends she feels we should sort of be paying attention to and really keeping our eye on. And also ahead, we’ve got Sarah White, Athena Karp, Lena Mach Bock And Dave Smallwood. So again, go to CXR.works and check out the events page make sure you don’t miss any of these awesome awesome lineups. With that I want to say thank you to everybody and we are going to see you next week Ciao.
Thanks for listening to the CXR channel please subscribe to CXR on your favorite podcast resource and leave us a review while you’re at it. Learn more about CXR at our website CXR.works facebook.com and twitter.com/careerxroads and on Instagram @careerXroads. We’ll catch you next time