Maintaining Human Intelligence in an AI World

Intro 0:01

Welcome to the She Said Privacy/He Said Security podcast. Like any good marriage, we will debate, evaluate, and sometimes quarrel about how privacy and security impact business in the 21st century.

Jodi Daniels 0:21

Hi, Jodi Daniels here. I’m the founder and CEO of Red Clover Advisors, a certified women’s privacy consultancy. I’m a privacy consultant and certified informational privacy professional, providing practical privacy advice to overwhelmed companies.

Justin Daniels 0:37

Hi, I am Justin Daniels. I am a shareholder and corporate M and A and tech transaction lawyer at the law firm Baker Donelson, advising companies in the deployment and scaling of technology. Since data is critical to every transaction, I help clients make informed business decisions while managing data privacy and cybersecurity risk. And when needed, I lead the legal cyber data breach response brigade.

Jodi Daniels 1:01

This episode is brought to you by No One Can Hear When You’re a Claw Red Clover Advisors. We have Elise Houlik, who is the Chief Privacy Officer at Intuit, where she leads the company’s global privacy and responsible data innovation and protection strategy, ensuring data is used to safely power innovation across into its ecosystem of financial technology products. Her team is deeply engaged with the business on all matters related to product development, data innovation and governance, and information security. Elise, welcome to the show.

Elise Houlik 1:36

Thank you. Hi, great to be here. Hopefully, you can hear me as well.

Jodi Daniels 1:40

Yes, can totally hear you. All is good.

Elise Houlik 1:43

Excellent.

Jodi Daniels 1:45

All right. Why are you laughing?

Justin Daniels 1:48

Laughing at you, I find to be entertaining.

Unknown Speaker 1:51

I mean, okay, it’s your death. So hard to start. I can’t speak.

Justin Daniels 1:58

Not enough espresso for you today. No. All right, Elise. Not

Elise Houlik 2:01

enough espresso in the world on a Monday. Like, let’s just hang in here.

Justin Daniels 2:06

So, Elise, why don’t you tell us a little bit about your career journey?

Elise Houlik 2:10

Yeah, absolutely. I think the thing I love about my career is the thing that I have excellent, smart, competent, amazing attorney peers and colleagues in the world hate, but I love it. Which is, it’s all happening right now. I could not have imagined when I went to law school back in the 1900 s that privacy and data and everything that I’m dealing with today-it wasn’t a thing. I didn’t even have a cell phone. Like that’s how old I feel, and so to have that journey of learning about privacy along the way, I’ve had a great career where I’ve had early days, even as a paralegal in a law firm, we were working spam cases and running things down for internet clients, and then that grew into my time over at Fannie Mae, as you mentioned, where I was leading up the the legal support for cybersecurity and privacy, and then really evolving more into the tech sector. So I’ve had a lot of time in recent days working on building systems while the law is being built, help the law get shaped, create a new space where we could use information in ways we never thought of before, and it’s exciting. I really enjoy it.

Jodi Daniels 3:26

I definitely feel like we’re doing the whole build the airplane while we’re flying it situation. Ounce, it’s certainly an interesting time. So speaking about where we are in AI, it is changing what employers should look for when hiring privacy professionals. So, in your view, how how is AI changing this landscape?

Elise Houlik 3:48

You know, you’re you’re catching me in in full swing recruiting right now. We’re we’re expanding too much EM, which is a blessing. I don’t take that for granted in the tech sector. What I can say is, I have seen over time the legal profession at least initially start to catch up with tech. I would I’ll be the first to say I will not project on others that when I was starting as a lawyer, I was formally an English major. Communication was my skill, writing, engaging with people, doing it all from my big legal brain, you know, and so the tech was another team, and they implemented their things. And over time, to be successful in my career, I had to get much more curious about technology. That was something that I thought personally was going to help me be a better lawyer. Now, today with AI, everyone has to do it. They just have to do it, right? It is the same as the transition from remember we used to have computers on a desk, and then we moved to laptops, and then laptops had stuff in the cloud. Remember how everyone freaked out about the cloud? Like things have evolved so quickly, and I find more and more people engaging. Really, really successfully with AI when they’re super open and curious about it, and they find ways to seamlessly fit it into what they do day to day, whether their background is in tech or not, because that’s the whole point of the AI. It gives everybody access to the ability to use that and think and have these tools available for your day to day work.

Jodi Daniels 5:20

I’m curious. When you’re talking and interviewing, are there clues that help you? This person is embracing it versus another person who might not.

Elise Houlik 5:31

Yeah, I think, at least in my experience in the recruiting context, it’s been very interesting because I think there’s just a little bit of hesitation in acknowledging that AI is happening even in the interview, and I would far rather a person say, “Hey, I used AI to prep for this interview. I’m going to have a chat assistant running on the screen while I talk to you. I hope you don’t mind. Like, let’s go. Let’s have the conversation. I’m here. I’m still present. I think that the challenge is when you see people looking to the side and reading, and then acting like, “Oh, I’ve got this, and I’ve made this up, and all. And it’s like it’s fine. Use the assist. I want to know that you can because you’re going to do it when you work here. So let’s just be super open about it and go.

Jodi Daniels 6:18

I appreciate you sharing.

Justin Daniels 6:20

So, Elise, I have a follow-up question because we’re talking about employers, but I want to change the context and add a ask a similar question. So, obviously, at a company like Intuit, you probably have relationships with all kinds of outside firms who you get support on for various things, and one of the things I have found since I have really embraced AI in my practice, is the billable hour simply doesn’t work anymore because I find consistently I can do certain kinds of projects and do it in a shorter amount of time with AI supporting my decision, not substituting for it. That the billable hour just doesn’t work anymore because it measures how long it took to do something, not the actual quality of the deliverable, and more importantly, how important was it to the particular strategic goal of that engagement? And I would love to hear your thoughts on how you see that relationship evolve, because AI, the current billing model, I just I don’t think it works anymore.

Elise Houlik 7:29

You know, I would say we were probably considering that before AI even came along, right? There are certain projects. You know, part of the value of really engaging with a terrific outside counsel is hopefully they’ve invested time and they understand your particular needs, your issues, the things that are going to arise in your business day to day. But at the end of the day, typically large firms at least support many many people who have the same question. And so once you’ve answered that question and you have confidence in that answer, the exercise is really in tailoring that answer for a particular client. It’s not a blank piece of paper every time. So we we were already looking at non traditional sort of flat rate or different type of models to do it. Now with AI, I think where I see the difference more is sometimes we don’t need to go to you because we have the extra capacity to have AI kickstart that analysis, and then look at it with our judgment and our expertise, and customize it and whatnot. There is still always going to be a need for outside counsel and and strong support, but you know there’s more and more we can accomplish in house because AI is a sort of extra teammate, and it gives us the capacity to run down nuances and then pull facts together and get things going. So I do think the legal profession in general is going to have a moment where they have to think about certainly if they want to stick to the billable hour model, how are they at least again acknowledging, hey, we have these tools too, we use them, they make us more efficient, and so here’s your, you know, final bill.

Jodi Daniels 9:05

As we think about AI and privacy work, and you mentioned it, and Justin, you did too. Around, we still need to be able to think. So it’s really helpful to gather facts and and collate through information. In your mind, how do we make sure that professionals are still able to have that critical thinking and judgment, and even question AI? Did you get this right? Do I agree with you? What are you missing in this analysis?

Elise Houlik 9:32

Yeah, I think for me, my hope-I think I’m right-but my hope is that you still don’t lose the you know, and intuit we call it the AI, right? So the human intelligence part of whatever it is you’re doing, you have AI, you have AI. There you go. Now you’ve got it. And and the HI part is massively important because again, AI is a tool at the end of the day, the same as you know a cell phone is. Or another tech tech piece of equipment that you’re using to get through your day, it is going to get you to a place where you, the human, have the ability to spark. Oh, okay. Let me look at this. Let me question this. Let me do this. And I think for the legal profession in particular, and I think about even as I’m hiring, that’s what lawyers will be bringing to the table. It will be the AI. It will be the creativity, the critical thinking, all of the ways we’ve been trained to look at something and say, “Is that quite right? Is that quite right? We just aim it at the AI output. I think any anyone who leans too far into the AI spectrum, you end up getting a lot of what’s the term slop right back right sounds very generic sounds very you know there’s a certain way of phrasing there’s a cadence and stuff like that and we want to make sure that not only are we looking at the accuracy of the information and how we use it but the quality of the content right we want to show up like a human being when we’re supporting the customers that are using us, who are also human beings.

Justin Daniels 11:06

So, Alicia, I kind of wanted to ask a follow up to you. So, in my own experience using AI, it does certain things well: drafting me contract clauses, being a negotiation kind of tool partner. But sometimes when you get into regulatory issues on various degrees, I find it to be a whole lot less reliable. And I’d love to hear your perspective from privacy lens when you look at you know the the regulations across privacy and then the AI that touches it. Where do you see from a privacy lens where you think AI is very reliable and a helpful tool in other instances where you need to tread a little bit more carefully, and the AI that you mentioned is really important.

Elise Houlik 11:50

Yeah, I think you know, as as you both are probably very well versed, the privacy landscape is still super super fragmented. There are wildly weird nuances depending on what state, what country, what region you’re in, and so we found it more helpful to be an initial aggregating point, right? Combing, resourcing, getting to models that are actually trained on the regulations, and what we do is, well, we’re building out our own compliance program once the law is passed and final. We’re pulling that in to other models that the attorneys use and stuff, and using the final actual model thing, not somebody’s summary of the model, not an early draft, but getting those things in. We have a sense of where our program is built around existing compliance, and so what the AI does is help us hunt for those nuances? What is this law different than this one? How and why, and does it add requirements? Does it create a situation where maybe we have to rearchitect a system, that kind of thing? But we’re constantly looking at the output. I wouldn’t say that we’re right at the point where we’re going to run that and we’re going to spit out a report on laws and regulations, and just say, “Yep, I’m sure it’s fine, without looking very critically and making sure that the the way that it’s referencing the law is the right section. You know, double checking work is still kind of part of our bread and butter right now until it gets a little bit more sophisticated.

Jodi Daniels 13:20

I also remember the era of no cell phones.

Elise Houlik 13:24

Thank you. I’m not alone,

Jodi Daniels 13:26

which which makes me feel I’m in a certain generation. So as we think about this next generation of privacy professionals, and we’re talking about how to use AI, what are your thoughts or concerns on the over reliance on AI for this next generation that’s really really tech savvy, and how do we how do we prevent them from from too much use of it? It’s it’s the fine balance and making sure that there’s still those critical thinking skills.

Elise Houlik 13:55

Yeah, I think you know, for me, I was really thinking about it more from the perspective of even my own career journey. You know, again, I I was based in in the humanities, right? English major, writing, reading, history, like all of those kind of things were my educational training before I went to law school. Helped me carry through law school, and now I’ve got a profession in more in the tech space. When I was early days looking for people to help me run my data programs and my privacy programs, I was always looking for attorneys who had sort of a non traditional background. I would have put myself in the more traditional bucket, but I’m looking for people who have like a math background, a computer science background, and then somewhere along their journey, they decided to become a lawyer and come because those were my my on ramps to the conversations with the tech teams. Now with this tool link, you know, and as we think several years down the road, I think today there’s a still probably a mixed bag. But as the the AI gets more sophisticated. I wonder if we’re going to come back to people who have more of that humanities training a little bit because everybody can use the tech tools. We have we all know how to speak with one another in the tech space, and then especially if you’re supporting like a data program, a privacy program, that kind of thing. And so, who has been trained to think critically in whatever you know realm that is, I think is going to be really, really important. And you know, my hope is that we all, to some degree, are going to become more and more dependent on this, but that we don’t lose the human side of this. I am the person using the tool; it’s not living my life for me. So, so how do you try to find people who are trained to look at things, just as you said before, and say, “Wait, is that right? Do I agree with that? Why do I think that? You know, that’s the skill that I want to see people really developing over time to run parallel with using the tool.

Jodi Daniels 15:53

That makes a lot of sense. Yes, mr. Thinker.

Justin Daniels 15:58

I’m just thinking about what Elise said, because we’re talking, and I saw an article today, Elise, where the whole online education system is really at risk because AI agents can come in and take the courses, and you know, since we’re talking a lot about how this next generation grapples with AI. You know, what do you think we’re going to do to ensure that people maintain their critical thinking skills? Because, you know, you like you talked about. I I remember when there was minutes for cell phones before the cell phone. How many minutes? Yes, no, this was going to be.

Elise Houlik 16:44

Yes.

Justin Daniels 16:45

Right. But where I’m heading with this, Elise, and I’d love your perspective is probably similar to you. I spent the time and the effort to do a lot of that grunt work, which wasn’t a lot of fun. But the byproduct of doing that is as I got exposed to a lot of things that helped me kind of learn what seems reasonable, and if you don’t have that and you go right to the AI, what do you fill the gap? Because in privacy, especially as you pointed out, and I’ve seen myself, is the the laws are very fragmented. But now you add on these AI laws that impact privacy, then you have the intersection between privacy and cybersecurity, and you’re watching it all play out very similarly with all the states coming up with laws because our federal government can’t get their act together. That in and of itself is a really tough road to hoe. But now you add on this other layer of these people coming up who may work with you or for you don’t have this base layer of of knowledge and critical thinking. To me, it’s very concerning.

Elise Houlik 17:51

Yeah, I think look, the burden is going to be on those of us who manage people and manage teams and want to grow them. We really have to lean in, right? So you’re right. The way I I came up and the way I learned because I was the one doing it all, right? The the only reason I have a career in privacy is because there was a project where I was working that had to do with privacy. No one wanted to do it, and I was like, I’ll do it. And now I have a career, right? Like it all worked out. I’m very happy for that. But I had a willingness to engage on the topic, but I had to be in the room. I had to take the minutes. I had to collate them. I had to get you know everything as a junior person. That is how I learned. I think I have no problem with AI taking the notes. Honestly, at this point, there’s so much fast conversation. Somebody hand writing the notes or typing them along the way, and I may miss something, but I still want them in the room. I have to bring the junior people along. I have to have them observe the conversation when the notes come out. I want them to be the one to say, “Did this catch everything? You heard it firsthand with your ears. Now look at it and be critical and help me prepare, and then let’s have a conversation about what you just heard. And I think that’s the thing that, as managers, we’re going to have to increasingly lean into to bring folks along. We can use the tooling as shortcuts for the the administrative aspect for it, but the development conversations, the exposure that you have to give to people in rooms where, if you wanted to, you could let AI kind of cut them out of the conversation. We don’t need you. We’ve got a notetaker. We don’t need you because somebody else has already prepared these clauses and whatnot. We’re going to have to lean in and say, “Hey, this has been prepared. Now your job is to look at it and tell me, do you think anything’s wrong with it? Do you think this is great? You know the other parties because we’ve had other conversations before, are they going to receive this well? Are they not? You know, it’s a lot more work in a different way.

Justin Daniels 19:48

So, at least another question for you. Just this has sparked so much interesting conversation today. And Jody, why don’t you give your two cents as well? Is you know as. Part of what you’re talking about training these junior people, I’m just curious. How often do you or do you instruct your people to argue with AI? Because I do it all the time, particularly when I use it for negotiation. Because what it does is it allows you to identify how it’s thinking, where it’s going wrong, and it helps you identify the limitations. So I’m just curious. How often did the two of you get, you know, tell the AI you’re wrong? Explain why it’s wrong, and what do you learn from the response?

Jodi Daniels 20:28

Well, I’ll start. I’m not bashful from my. I’m a pretty direct person already, and I have no issue telling you that that was terrible. That was wrong. Try again. So I do it fairly often.

Elise Houlik 20:42

I look the the beauty of the AI is at least as of today, it doesn’t have feelings. You know, sometimes you you know you can tell when you’re sitting across from a human being who’s tried really really hard on something and they just miss the mark. Those can be rough conversations because you’re like, oh, you’re not going to get it. The AI, I’m like, this isn’t true. Go fix this. Get better. Don’t deliver me answers that don’t have support. Why do you know? You can get right in there and and tell it. And I think that’s again part of the job, not only of learning how to interact with that tool, but it makes the tool better. Over time, the tool will look and find credible sources and ways to do it, and won’t make leaps that you don’t want. But you’ve got to be very vigilant and very direct to Jodi’s point and say, “Look, this isn’t the result I wanted. Go back and do it again.

Justin Daniels 21:34

That’s interesting because when I argue with it, I was like, “If I were a less experienced lawyer and I just accepted this, and we gave your wrong answer to the client, they would likely fire us. What do you think would happen here? And then, oh, I lay into it pretty good.

Elise Houlik 21:46

There you go.

Jodi Daniels 21:47

I think you use more words than shocking. Really?

Jodi Daniels 21:51

I’m just really short. Bad. Nope. This

Justin Daniels 21:55

oh, I

Jodi Daniels 21:55

not what I asked for. Because

Justin Daniels 21:58

I think what you find out. Oh, I pattern match these words. Oh, I use these rules from this other law, and I’m like, well, why did you do that? And what it does for me, at least, is when I watch where AI goes wrong, it helps me hone prompts because now I say you have to support everything you got. You have to show me what clauses, and the better your prompting gets, because to me, that’s the skill people have to learn right now. Maybe that changes; it probably will. But for now, how good your prompts are is important. And what I find, at least in our profession, is a lot of people are very like surface level prompts because they want to do it quickly, not realizing the real gold is honing that skill and having some really good prompts because you’ll get much better output.

Elise Houlik 22:40

Yeah, I completely agree. I think you know, not that long ago. Let’s go back. Let’s say spring of you know 2023, when Chat GPT kind of really hit the scene and people were starting to use it. There was, there was a lot of like, just do AI, just do it, and it’s like, well, you know, I’m gonna pull it up and I can ask it, you know, how to plan a party or whatever it was, you know, out there for. But over time, you know, AI AI can be a very big time saver. It also can suck away your time if you’re not using it effectively and being targeted and thinking about why is it that I’m pulling AI into this experience right now? What’s it going to bring to the table? Why do I need to use this tool versus that tool? Because there’s a wide array right now, and they’re not all the same. Having that understanding first, and then coming as you say with a very specific intentional prompt is going to be the way to navigate through that. Because there’s only going to be more tools available as time goes on.

Jodi Daniels 23:38

If you were to think about someone listening who realizes they need a little bit more skills in this area, what would you recommend for them to get started?

Elise Houlik 23:50

You know, again, the the great thing about this day and age is it’s all out there. It’s all available now. As a privacy professional, I feel as though I need to say when you’re using public models to be very thoughtful about what personal information you might want to put into it, because that is going to train those models, and that might be a choice that you’re willing to make. But you should go in intentionally, right? So that’s my TED talk. We’ll put that to the side. Getting in there and playing with it. There is almost no software that’s available today. It doesn’t have some form of a copilot, right? So if you’re a Microsoft Office user, you can use theirs. Gemini within Google, like all of these things, have assistants that can help you get started to at least make it feel very accessible. I think there are any number of courses. If you’re a professional in the legal space, you know, IAPP has a whole training on AI and what it does and the governance. If you really want to get into the weeds, but if you’re really just trying to use the tool and be proficient in it, use the tool and be proficient in it. Right, just start right because it’s never going to come to you. You’ve got to go to it right now, and more and more you’ll see. I think in our daily lives, it’s going to be surfaced, you know, even more than we’re seeing today, and so it probably will not be something people can avoid for very long. But it’s not as intimidating as you might think, and quite frankly, the the the negative results or the incorrect results and all come from a place of like, wow, that was a great question, you know, and and right, like, so you don’t have to worry, you know, the AI is not going to shame you. If anything, it’s going to boost you up, you know. Try it out, use the tools as often as you can, and find out the differences between them again, because to me, that’s very impactful. Not every every tool is met for every purpose.

Jodi Daniels 25:42

Very, very true. These different tools for different purposes.

Elise Houlik 25:45

Yeah.

Justin Daniels 25:46

Yes, Copilot sits at the bottom of my list.

Elise Houlik 25:48

It is. It is available.

Justin Daniels 25:50

Yes, it is. Though you’re right. What should we ask next? I’ve been going off. Yes, I know we have our questions, but I’ve asked so few. We

Jodi Daniels 26:00

always have. We always. We have to ask every guest our same.

Justin Daniels 26:03

Yes, I know.

Jodi Daniels 26:04

It wouldn’t be consistent at almost five years in.

Justin Daniels 26:07

Yes, I’m aware.

Jodi Daniels 26:09

So or six years.

Elise Houlik 26:09

So don’t derail us. Come on, can

Jodi Daniels 26:10

be six years. We’re closer to six than five.

Justin Daniels 26:15

Yes, just like I’m about to be closer.

Jodi Daniels 26:17

Oh yeah, no, we’re not going to talk about that.

Justin Daniels 26:18

Yeah. Okay. Oh no.

Elise Houlik 26:19

Oh dear. I’m

Justin Daniels 26:19

sorry. So Elise with all that you know about privacy, and we didn’t talk a lot about security today, is do you have a best tip you would like to share with our audience?

Elise Houlik 26:32

Ah, you know, if thematically with what I said earlier, don’t lose the HI in all of this. Don’t, don’t, you know, there’s, there is a there is a path where you can go and let the the AI bots run and their agents run and move on with your lives and those those tools are going to get better. The kind of errors and mistakes that are surfaced today are going to decrease over time because these these tools are going to get smarter and and better at their jobs and stuff, but how sad if we lost the ability to be a part of this conversation. Like, I really think part of what we bring to the table as humans look at it this way: we’re the ones who created the AI, right? What else could we possibly do? So please make sure that you’re endeavoring to engage and and really, you know, think about why it is you’re even using them in the first place, and how you can be made better, not replaced by it. Because I think there’s a huge opportunity for us as humans to let this run shotgun, and then we create something even cooler down the road.

Jodi Daniels 27:37

And Elise, when you are not working in privacy, AI, security. What do you like to do for fun?

Elise Houlik 27:45

I, I am. You know, I, I was an insane travel junkie. I would say before COVID, and COVID, I think has as it did for everyone. Everyone took a bit of a break from everything, and now I’m kind of getting back into it. So I’m lucky. I have a good group of friends, and every now and then, someone you know starts the text chain. You know what we should do? We should go to Japan, and you’re like, “Yes, that is a great idea. Let’s go. You know, and I’ve I’ve done Machu Picchu, and I’ve I’m Kilimanjaro, and I’ve also sat around at some lovely resorts and just had a nice cocktail. So I just like to get out and see the world as much as possible when I have the time.

Jodi Daniels 28:27

That sounds really fun. I hope you have a good one coming up, or that you’re in the midst of planning.

Elise Houlik 28:31

Yes, always, always, always try to plan for the next one. That’s the best part of any trip is guessing where the next one’s going to be while you’re in it. Very different.

Jodi Daniels 28:39

Dustin does that very often. We’re on vacation, so what should we do next summer? Exactly, we’re at the beginning. Well, Elise, we’re so grateful that you came to join us today. And if people would like to connect or learn more, where should they go?

Elise Houlik 28:54

I am on LinkedIn. I’m pretty confident I’m the only Elise Houlik on LinkedIn. I’m going to declare it. If you find another one, you let me know. But I’m right there. And you know, when I’m when I’m not doing awesome conversations like this, I’m I’m hanging out at Intuit, trying to make sure our products and services deliver for our customers all day, every day, even with AI and AI and all of the things that are going to come next.

Jodi Daniels 29:22

I really love the HI. I’ve not heard that before. You need to talk. No,

Justin Daniels 29:26

it’s funny because HI before in media circles were known as human interest stories. They call them HI.

Jodi Daniels 29:34

Oh, I just heard human interest

Justin Daniels 29:36

stories. But that’s but she’s Elise’s repurposed. This is

Elise Houlik 29:40

I’m I’m I’m taking it away. Yeah. Well,

Jodi Daniels 29:42

I mean, we all need acronyms because it wouldn’t be oh the universe. We didn’t have any of our own. Correct. I

Justin Daniels 29:48

given. Yeah.

Jodi Daniels 29:50

Amazing. Well, again, thank you so very much for joining us. We really appreciate it.

Elise Houlik 29:54

Absolutely. Thanks a lot, guys.

Outro 30:00

Thanks for listening to the She Said Privacy/He Said Security podcast. If you haven’t already, be sure to click subscribe to get future episodes and check us out on LinkedIn. See you next time.