Nicholas Rhodes – How AI in the Workplace Can Improve Employee Experiences

Connect with Nicholas:

• 

Connect with Dwight:

• 
• dwight.brown@salary.com

Connect with David:

• 
• 
• david.turetsky@salary.com

Podcast Manager, Karissa Harris:

• HRDataLabs@salary.com

 

Production by Affogato Media

Resources:

• Consulting Practice
• HR Data Labs Season 7 Episode 24
• HR Data Labs Season 7 Episode 22

• “[When using generative AI], no longer am I utilizing the operating system of the computer, I’m using literally a conversational operation system.”
• “AI is really scary right now to a lot of folks . . . but it’s free right now, and these tools are not going to be free forever. ”

Announcer: 0:02

Here's an experiment for you. Take passionate experts in human resource technology. Invite cross industry experts from inside and outside HR. Mix in what's happening in people analytics today. Give them the technology to connect, hit record for their discussions into a beaker. Mix thoroughly. And voila, you get the HR data Labs podcast, where we explore the impact of data and analytics to your business. We may get passionate and even irreverent, that count on each episode challenging and enhancing your understanding of the way people data can be used to solve real world problems. Now, here's your host, David Turetsky.

 

David Turetsky: 0:46

Hello, and welcome to the HR Data Labs podcast. I'm your host, David Turetsky alongside my co host, trusted friend, partner, I could come up with a couple other names, Dwight. Dwight Brown from Salary.com!

 

Dwight Brown: 0:57

I think a lot of people could come up with the other names for me.

 

David Turetsky: 1:01

Yeah but this is a PG rated show so we can't

 

Dwight Brown: 1:04

Yeah, exactly. I gotta, I gotta keep those secret.

 

David Turetsky: 1:07

Yes, yes, it is a secret. Oh, maybe we'll cover that on the 25th episode. Today we have with us a pretty brilliant guy and you'll hear why in just a minute. Nicholas Rhodes, he's the founder and creative director of OutSnapped. Hello, Nicholas! How are you?

 

Nicholas Rhodes: 1:23

Hello. Hello! I am well. I'm I'm coming to you from New York City. We have bright blue skies with big blue, uh big white fluffy clouds. So I'm in a good mood today.

 

Dwight Brown: 1:33

Wow.

 

David Turetsky: 1:33

There you go. In fact, I was just I was just there yesterday, dropping off some boxes for my kiddo to pack up their dorm at the School of Visual Arts

 

Nicholas Rhodes: 1:44

Oh cool!

 

David Turetsky: 1:44

on 24th Street. And they'll be coming. I'll be picking them up. Maybe in two weeks? Oh, my goodness. That's incredible.

 

Nicholas Rhodes: 1:51

Yeah, this year has flown.

 

David Turetsky: 1:53

It really has. But Nicholas, tell us a little bit about you!

 

Nicholas Rhodes: 1:57

Yeah! Where do you want me to start?

 

David Turetsky: 2:01

Well, you can start with birth, but that might be a little bit long for this 30 minute podcast.

 

Nicholas Rhodes: 2:06

I don't remember much from that day. So I guess well, I'll give you this short history. So I went to school in Boston and studied what they called New Media at the time, which was early 2000s. Back in the time when the internet was black webpages with red text and we thought it was super, super cool. So I did four years at Emerson learning what we called New Media, like I said, and at the time I graduated, the incoming freshmen seemed to know more about new media than I did. Also, after four years of looking for missing semicolons and commas, I decided that I wanted to really focus on print design, and not so much on the technical aspects of the internet anymore. So up to date, and we'll kind of jump back and forth in time, I've kept up just enough with technology and coding to what I call or I should say, as my dev team called to be very annoying, because I know what's possible but I can't do any of it myself anymore.

 

Dwight Brown: 3:07

Know enough to be dangerous.

 

Nicholas Rhodes: 3:08

Yeah, exactly. I know enough to be a nuisance. Going into a weekend when I'm like, I don't know, it seems like you could have done that. And so after school, I was a double major in new media and photography. After school, I decided that photography was going to always be my hobby. And I was going to really dive headfirst into editorial design. And so it was a really interesting time, especially, I mean, everywhere, but in New York City, I came in as a 21 year old, eager to work in print design, while every other 21 year old and everyone was eager to work on the internet. And in my head, I was like that internet, I've been there done that. You know, I remember literally a day where the entire class was devoted to watching a movie buffer, like a one minute movie buffer and thinking to myself, like how dumb this idea was. And now obviously, we know the Internet has come more than full circle on that moment.

 

David Turetsky: 4:03

Oh, yeah.

 

Nicholas Rhodes: 4:04

And so that was a big learning lesson for me though, that just because I couldn't see the promise in something at the moment, I started to keep an eye on it and play with it and tinker with it. So I worked in print design, it was really, really cool, like I said, because I got this opportunity to really jump ahead of where I should have been as a 21 year old because no one else was vying for these jobs in magazines. And so I had a really good run going from a production intern up to a art director within the first handful of yours. At the same time I started as my hobby, photographing parties and nightlife and music, you know, the up and coming 20 Something with not a lot of spare money. I was getting entry into concerts and parties by saying I would take photos in exchange for going to the event so it's getting photo passes and media passes to things. And that spiraled into what at the time they called a personal brand, fastforward to now, it's influencer. But I had a blog that was operating, essentially, in my off hours from my day job, I would go out to parties and take photos and post them online starting in like 2006 ish. And at the height of that website, we were doing close to 150,000 pageviews a day, from all the major like fashion centers in the world and other major market cities. So this was like now we think of kind of the most of the world is kind of being homogenous in style, and concept and music, because all these things pass borders so quickly. But back then people were tuned in to see what was happening in New York from all around the world. So that was really awesome. So as a result of that, I transitioned from my role as an art director at the magazine, I was working with at the time, Radar, RIP, and became the managing editor of the website, which was once again in role that I was probably too young to be doing, because I did not know how to manage anyone. But I did know how to run a website, because while everybody else I was working with was sleeping, I was getting 100 plus thousand pageviews a day. So I came in and was working with these folks, and was a really interesting time because we were in the process of moving from kind of website for a magazine having a subscribe button on it, essentially, to subscribe to the official print magazine. And then started running content on those sites themselves, especially in sort of the gossip world that I was working on. Radar magazine was what I like to call a very smart People or US Weekly. And you can see who was reading it, who got the jokes, like we would always joke in the focus groups who was very smart. So we would start the story essentially, where US Weekly and People left off and continue with the more interesting parts of from those things. So long story short, that went out of business in 2008. Magazines literally are still not hiring and I had to figure out how to make a living. So I started working as a professional photographer, and with that website that I had met figured out how to monetize it. I did a lot of really amazing things, eventually realizing that in the world I was living in, the only way I was going to be able to make rent was if I started throwing events myself. So so many of the things that I have learned, I've sort of in my head invented digital marketing, because there was no one to teach it at the time. There was no one to teach me at least at the time, so I learned about AB testing and things like that, and creating products to literally sell tickets to future events.

 

David Turetsky: 7:29

No, it's cool.

 

Nicholas Rhodes: 7:30

Yeah!

 

Dwight Brown: 7:30

Awesome.

 

Nicholas Rhodes: 7:31

Fast forwarding to where we are right now, I could not expand on that particular project, Nikki Digital because I was one person. So I started figuring out ways to expand that brand, which then spawned OutSnapped, which is the company that had all the learnings from starting a business and all of those other things that were part of it. So now OutSnapped focuses primarily on marketing experiences, internal events for companies, we used to be just an events business but now we do a lot of longer term marketing campaigns. We've had full two three year campaigns internally with companies like Cisco, for their onboarding programs through HR where people can take selfies together, and document their onboarding experiences and things like that.

 

David Turetsky: 8:18

That's great.

 

Dwight Brown: 8:18

Cool.

 

David Turetsky: 8:19

Well, now that we know so much more about you, Nicholas, we need to ask, What's one fun thing that no one knows about Nicholas Rhodes?

 

Nicholas Rhodes: 8:27

So this is the thing that eventually usually slips out, like a year two or three have a friendship with someone we'll jump to it today.

 

David Turetsky: 8:35

We're fast forwarding.

 

Nicholas Rhodes: 8:37

Yeah, as a kid, I was a magician, but like, obsessively. It was all I cared about, like, legitimately all I cared about. And when we would travel around, the first thing I would do when I went somewhere with my family was open the yellow pages to see if there was a magic shop. And there are so few of them around the country that occasionally I would really, really score big. There is an old New York Times article floating around so I can't say no one knows about it. But I was lucky enough to be on Broadway for a little bit opening another magician show, sort of like the kid, the kid opener to get the crowd warmed up, which was an amazing experience. And so much of what I have learned about interacting with people and public speaking really came from that as a child and having people take, you know, take interest me.

 

David Turetsky: 9:28

That's a wonderful experience.

 

Dwight Brown: 9:29

Yeah, that's awesome.

 

Nicholas Rhodes: 9:29

Yeah, it was really, really cool. Luckily, they have lost the photo. It's no longer attached in the New York Times article. You have to wait until you really know me to see the photo of me in a top hat.

 

David Turetsky: 9:41

There you go. We will we will probably hopefully one day ask you for that.

 

Nicholas Rhodes: 9:45

Yeah.

 

David Turetsky: 9:46

So our topic for today is one that will pique the interest of a lot of people inside and outside the world of HR. And that is how AI in the workplace can improve employee experiences. So our first question for you, Nicholas is: you're a creative director, you're a photographer, you've been offering lots of experience to corporations and their employees for many years. Give us an overview of what you think generative AI is, and what it will really mean for an HR professional.

 

Nicholas Rhodes: 10:22

Yeah, well, okay. Well, I want to be clear, I'm not an HR professional. And one of the reasons I

 

David Turetsky: 10:28

And you don't play one on television, right?

 

Nicholas Rhodes: 10:30

Yet! And one of the reasons I had to keep starting my own companies is because HR professionals Google me and see photos of me dancing on top of things and swinging from polls with famous musicians, but um,

 

David Turetsky: 10:42

So they see you as a risk is what you're saying.

 

Nicholas Rhodes: 10:44

Yeah, I don't get I don't get past HR's AI is essentially the issue. Right? So what is generative AI? Generative AI is so extraordinarily vast that it's hard to really even comprehend. So one of the examples that I like to use about where generative or AI in general are going right now is, we think of an operating system on our phones or computers in this old fashioned way, right? So when I used to pull up a calculator app on my computer, I now just do that in chat GPT, I'm already talking to chat GPT. And I literally just say, I have this event, it starts on this day. And this day, and this many hours per day, tell me how many hours total, right? So no longer am I utilizing the operating system of the computer, I'm using, literally a conversational operating system in which, in the future, I can do any number of things, even right now I can say make me an illustration of XYZ. And that's how we're using it at my company, mostly, aside from interacting with our clients, we are using it as a product to make generative AI visual. So that's kind of the broad scope of generative AI. I'm happy to dive in on any of the smaller particles of it. But that's like, the broad scope is that this thing is going to do, everything that we do is you're gonna get in your car, and that's going to drive you somewhere.

 

David Turetsky: 12:02

But you have to learn how to ask it in order for it to be effective. Because if you

 

Nicholas Rhodes: 12:06

Yes!

 

Dwight Brown: 12:10

Go to LA first? say, I want to go, let's use, because you're in New York City, and I love my New York City. If you're in New York City, and you say, hey, I want you to take me to Jersey City. And you don't

 

David Turetsky: 12:29

Well, I was gonna say that doesn't involve say I'd like to use the Holland or Lincoln tunnels to do so. Is actually putting your car on the water. So yeah, you have to it going to try and find a route across the across the water? learn how to ask it the right question, right.

 

Nicholas Rhodes: 12:43

100%. And I think this is also like a good point to pause and think about generative AI to everybody feels like this brand new thing. But directions is a really great example of how we've been using it. I don't know if Mapquest technically was AI, but it was definitely figuring out how to get you somewhere. To answer your question, like Waze, anytime I try and outsmart Waze I regret it for the entire drive, right?

 

David Turetsky: 13:08

Sure.

 

Nicholas Rhodes: 13:08

So it's already doing things like that. But yes, it is about asking the questions. And as the new interaction systems and as the AI gets smarter, it'll be less about asking the question because it can have enough smarts or reasoning to respond to you and say, I don't understand what you mean. Do you mean x? Or do you mean y? And we're already starting to see things like that with Chat GPT. It will give you two answers and say which of these is what you want?

 

Dwight Brown: 13:34

Right.

 

Nicholas Rhodes: 13:35

Yeah. And it's learning from those interactions when you tell it which one you want.

 

Dwight Brown: 13:39

Yeah, and I think the, you know, I've, I've really started to integrate a lot of the AI, the generative AI and to even my work processes, both life and work processes. And it, it has been interesting going through that learning process. And to your point, David, the fact that you have to know how to ask the questions, and sometimes it's peeling back the layers of the onion, as you're trying to get it to do something, it'll, it'll kick something out. And sometimes it just gets stuck completely wrong. And, you know, the fear out there is that AI is going to take over the world and nobody will need human beings. And that's just not correct. I mean, you've need those human beings to interact with it, to tell it what to do, or ask it what to do and to recognize where it's

 

David Turetsky: 14:29

Well, I think one of the frustrations that I wrong. have, Dwight, is that even the you know, Nicholas, used the the context of Waze and I use Apple Maps, right? I find it very frustrating to ask Apple Maps. Hey, listen, I want to go home. I am from going from New York City from the School of Visual Arts on 24th and 1st. I want to go home but first I want to stop by a bagel store so I can feed my kids. But even just the user experience of being able to go and do that is very frustrating, because it asks, Oh you want bagel place in New York City. No, no, no. Do one in Yonkers or outside of Manhattan. Okay, do you want one in the Bronx? Yeah, actually, the Bronx bagels are good. But no, no, I want one further. You see, we're still not at a place though yet where it can read my mind. Where it knows kind of, No, I want to kind of escape the city first, before I try and get a bagel. I want to be able to park in a parking lot, and not have to worry about parking on the street. So what I'm talking about in terms of generative AI is almost like a smart tool learning you an assistant who learns you, learns how you like to do things, and then enables you to do them by asking them something very simple and they already know the context and all the assumptions behind it.

 

Nicholas Rhodes: 15:45

Yeah. So there's, well I'll have to see if it still exists, but there's a really good bagel place in Dobbs Ferry, which should be on your way for next time. So remind me to look that up.

 

David Turetsky: 15:53

Yeah please do. Yeah, that'd be great.

 

Nicholas Rhodes: 15:55

And I won't take a dig at Apple Maps either while we're here, but

 

David Turetsky: 16:00

It's all fair.

 

Nicholas Rhodes: 16:01

The thing that I think is really interesting is, since these are conversational operating systems, to do what you're talking about is already available within chat GPT or Gemini. Right? Chat GPT has something that I think it's a feature that I personally think they should have come up with a better name for. Those are called GPTs within chat GPT. And you can conversationally program them to do things. So for example, our team, we have conversationally programmed one of these through iteration, to draft emails based on responses and things like that. So a lot of the times we can plug in a bunch of information, cut and paste. Or another really good example of this is I've created one that I put together, before I hop on a meeting for people, I scour the web for mentions of them and then I take their face, their sorry, their LinkedIn profile, the email that we've had to date, and drop it into this tool that I've created. And anyone on my team can use it via the same link. And it will put together a meeting briefing about the person based on what's on their LinkedIn, it'll tell me if we have people in common all of those things. So I can sit down and instead of doing the four hours of research before a meeting, which no one has time to do anymore, I can read a 10 minute thing that my GPT put together for me, and then go and do the further research on those things that I want to do. Once again, AI is not fully replacing me or an assistant, because I still have to go through and do those things. And my assistant is the person who's putting that information into the GPT and giving me the output, right?

 

Announcer: 17:43

Like what you hear so far? Make sure you never miss a show by clicking subscribe. This podcast is made possible by Salary.com. Now, back to the show.

 

David Turetsky: 17:54

Let's transition this though to well, so what does it impact? Or how does it impact the world of HR, because what we've been talking about is really kind of the how it can personally impact us, what normal stuff, you know, outside of the world of work. And well, what you just mentioned would be really useful, because it's a really cool tool for you to be prepared for your next meeting. But how does it help HR?

 

Nicholas Rhodes: 18:17

Yeah, well, I think I joked about it earlier. But I mean, this is goes back to 2015, 16, 17, HR has already been using these sorts of initial screeners that are AI based to determine whether or not they should even have a conversation with someone. Right? So very similar kind of tool that I just mentioned, for different use cases, going through all these resumes, and just determining whether or not it's worth talking to someone. I think, you know, like I said, and this goes back to why humans are still very relevant. I think that I would have been a very valuable asset to a team. But I couldn't get past those AI bots, right? To even talk to a human. But another really cool way you could use these things is for, you know, imagination workshops, too. I think that these tools, since they can do so many vast things. It's a really great opportunity to lead workshops internally with your team. Sales kickoffs, team building experiences, the ability to imagine what you would look like as a better leader, you know, if you're working on a CEO workshop or something like that. So there's a lot of opportunity. And I think one of the things that makes these conversations so tricky is that I'm a creative. So my imagination starts spinning and I'm not super data oriented, where I know HR is or I shouldn't say I'm not data oriented. I'm not looking at the same data metrics on a regular basis. So one of the ways that I think it could work really well for HR in this use case, for example, if you're at a career fair. You could have a interactive AI experience that asks prospective, we do a lot of these for schools and prospective students that they could be first, prospective employees in as well, ask a lot of questions about where they are in their educational career to see when they will be hireable, what their, how they see themselves as a leader, what traits they want to learn and improve on. And those answers to the madlib could actually give a visual output that's a takeaway for the individual. But it also will give the HR person a list and a CSV of all the information that was collected, which could tell you about your prospective hires as well.

 

David Turetsky: 20:26

And it's scouring the web!

 

Nicholas Rhodes: 20:28

Yeah, it could.

 

David Turetsky: 20:30

Which might be good. Or it might be really bad for that person, if they've been a very frequent poster on TikTok and have drank a lot of alcohol and then posted those, which we're really not supposed to look at in the world of hiring, you know, we're really not supposed to scour the web for those things. And I guess the question is, so where does HR have to draw the line on those things? So that we know what we can do and what we can't do? Or is it really kind of I hate to use the term wild west right now. But is it kind of like open?

 

Nicholas Rhodes: 21:00

Yeah, I think I think it is Wild West. And I, you know, I would love to say that HR would never do this. But if you have two candidates, and they're equal on paper, except one of them, there's an explicit photo on the internet. No, you shouldn't be taking that into account. But as a human, there's no way not to, right? So in theory, if you were to program that into AI, AI could not have the human emotion that's connected with that, that we have. So in theory, it would do the reverse there where you're actually making a better educated decision, because you have no human perspective on it. To me, that seems silly, though. Because the human perspective is how you became the HR professional and therefore, you know, what makes you good at your job.

 

David Turetsky: 21:46

But but as the HR professional, we know, the rules that we have to follow, though.

 

Nicholas Rhodes: 21:49

Sure.

 

David Turetsky: 21:49

And we need to tell the AI that there are certain rules, like we have to have certain blinders on for certain things.

 

Nicholas Rhodes: 21:56

Sure.

 

David Turetsky: 21:56

And then, you know, if a background check comes back, and someone has a criminal history that they haven't disclosed, well, that's a bad thing. And we need to, you know, we need to take specific actions on that based on what we're allowed to do, not just on what we want to do, or what we think is the right thing to do.

 

Dwight Brown: 22:11

And one of the things to take into account with this is the error rate. That's, that's out there. And the, you know, I, I googled myself, for lack of better term, on Chat GPT. And it was interesting what it came up with, apparently, I'm a journalist over I don't know, somewhere out east and there were a couple other facts that I must have been in a total blackout for. But

 

David Turetsky: 22:40

It may be because you hit your head when you're doing some.

 

Dwight Brown: 22:43

Yeah, exactly a botched launch on extreme sports. But but it you know, the HR professional has to has to be able to weed through that as well. You need that human interaction there that, and I don't know, I don't even know how that works exactly. But

 

Nicholas Rhodes: 23:01

Yeah, there's there's an up and coming photographer somewhere in the UK with the same name as me, which has been giving my Googlers a real run for their money. And I keep getting alerts of these very cute photos of squirrels that are being published around the internet.

 

David Turetsky: 23:15

And like Nicholas, are you sure those aren't you?

 

Nicholas Rhodes: 23:19

Pretty sure. I took Photo 101 in school right on the Boston Commons. And there's actually a rule that you are not allowed to take photos of squirrels because there's so many squirrels there. Apparently, pre that rule just everyone would come back with just tons of photos of squirrels.

 

David Turetsky: 23:35

The squirrel sued.

 

Dwight Brown: 23:37

Yeah, exactly! Copyright.

 

Nicholas Rhodes: 23:38

They want the royalties.

 

David Turetsky: 23:39

We want all the nuts. It's got to be in nuts. But But if we think about this in the context

 

Nicholas Rhodes: 23:43

Yeah. of HR, though, other than, you know, other than recruiting, you know, is there other kinds of data that Gen AI can capture that would be useful or helpful in the AR world? Sorry, in the HR world, not the AR world.

 

Dwight Brown: 24:02

It'll be AR pretty soon.

 

David Turetsky: 24:03

Augmented reality.

 

Nicholas Rhodes: 24:05

We're all gonna be on our headset.

 

David Turetsky: 24:07

Dude Dwight has one!

 

Dwight Brown: 24:08

Not anymore.

 

David Turetsky: 24:09

He's got one!

 

Dwight Brown: 24:10

Not anymore!

 

David Turetsky: 24:11

You got rid of it?

 

Dwight Brown: 24:11

I got rid of it.

 

David Turetsky: 24:12

Dude! You should have sold it to me!

 

Dwight Brown: 24:15

Well, I just returned it. I didn't even sell it.

 

David Turetsky: 24:18

Oh. He had the Apple one.

 

Nicholas Rhodes: 24:19

Yeah. Gotcha. Did you see like a day two or three after that came out there was a guy driving the Cyber Truck?

 

Dwight Brown: 24:26

Yes!

 

Nicholas Rhodes: 24:27

Amazing. And everyone looks at it like what's wrong with the world today? So I think, yes, data mining is going to be huge for AI. Obviously, it's a really obvious one. I think one of the other ways that it can really work is activity, or strengthening relationships on teams, building and creating conversation amongst people. And I think that, for me, the data side of it is really interesting because it can look at data over a long period of time and see things that we as humans can't see. We work a lot with visual AI and we've had a few kind of sticky moments where some things come back and you know, it's kind of a minefield, because AI is based on society, right? So it has all the inherent biases that we as a society have, for better or worse, we've had photos come back where people look of a different race, perhaps than you would perceive them to be. And the person in this case happened to be Native American. And we didn't see that and then, but they were very surprised that they were like, no one ever knows it but I'm like, 70% Cherokee, something like that. And it was really interesting, because AI saw whatever was in her face

 

Dwight Brown: 25:41

Interesting.

 

David Turetsky: 25:42

Yeah. Well, from an HR perspective, but it's structure, and was like, Oh, I know that this what this person is or who this person is, what their, you know, their identity is. And that's a really interesting kind of thing to flip flop on too. really kind of scary as well, because it kind of outs people, I was using air quotes everyone. Outs people as to what, what their background is and that's not necessarily a good thing in the world of HR.

 

Nicholas Rhodes: 26:13

No, no, it's not. And I think it's an interesting thing, because in this case, it was actually a moment that I was like, oh, no, but the person actually felt very seen in a way that they had never really been seen by their peers before. So it does go both ways. But yes, I understand what you're saying from that perspective. For sure. I means\, it's, it's, there's so many parts of AI that are going to be perceived as landmines until we figure out how to use them, right. And I think it's like any tool like, the wrong person using a backhoe can do a lot of damage. And with AI, you're the operator, right? And you get to use it in a responsible way.

 

David Turetsky: 26:51

But the problem with the AI versus a backhoe is the data is really important in the world of AI, whereas data for a backhoe is kind of what which lever you pushing, right?

 

Nicholas Rhodes: 27:03

Sure.

 

David Turetsky: 27:04

Whereas the and this has been one of the issues with how some AI have been trained, especially when it comes to different ethnicities and different races. And so the data we have has been unbelievably biased from against or for one particular type of group. And therefore, when we talk generative AI in the context of HR, we always get, you know, I hate to use the word defensive in this case, but we get defensive about it. Because Dwight and I both share this because Dwight's a data guy, and I'm an HR data guy, and we share this kind of thought process that HR data sucks, or kind of does suck, and therefore, how do you really get the most benefit out of it? Unless you go through the exercise of cleaning up the stuff you're gonna leverage first.

 

Dwight Brown: 27:50

All right, you could use AI in that cleaning process?

 

Nicholas Rhodes: 27:54

Yeah.

 

David Turetsky: 27:54

But then you got inherent bias potentially built in.

 

Dwight Brown: 27:57

Right, exactly. So it's no longer clean? Dirtying?

 

David Turetsky: 28:03

Well, it's clean. But what is it clean to? What's the standard use? What are the assumptions used, you're using and and what are we doing to that? Or do we are we getting to a layer of abstraction that doesn't matter anymore?

 

Dwight Brown: 28:13

Right?

 

Nicholas Rhodes: 28:14

Well, I think it does matter. I think whether Hey, are you listening to this and thinking to yourself, Man, I or not we'll be able to fix it is the question, right? I think there's so many questions that are still remains to be seen. It is for the lack of better term, the wild wild west, you're wish I could talk to David about this? Well, you're in luck. We right. It's so nascent. And if you look at what we've seen change in literally it's been a year and two months, three months since Chat GPT was publicly available and what it's done between now and then these things are going to be wildly different. And I do think, you know, there are people who are talking about it. The government, I think, should be doing a little bit more than they are right now. But I think we're about to see a sea change with that with starting with the TikTok ban. And hopefully it starts and continues the conversation around privacy in general and that's going to go into AI as well. Versus just singling out a single app, because I think as much as I'd like to point fingers around, our American companies are doing very similar things, right? And that's where a lot of the data that we're talking about right now is actually coming from, from those social networks. have a special offer for listeners of the HR Data Labs podcast, a free half hour call with me about any of the topics we cover on the podcast or whatever is on your mind. Go to Salary.com/HRDLconsulting to schedule your FREE 30 minute call today.

 

David Turetsky: 29:42

So let's talk in our third question about what can HR professional do to get started in this world you know, do they put their toe in and get Gemini or Bard or whatever the name is? To download or or just you know, wait until the their particular flavor of AI comes available? Where should they go? Yeah,

 

Nicholas Rhodes: 30:04

I think it's really interesting. I, I had a really interesting conversation with a good friend of mine who works at a company which will not be disclosed for the sake of this, but they, as an entire company have been given the task of exploring AI, right? So everyone is sort of being challenged to figure out how AI can work in their workplace. And, interestingly, their workplace blocks access to both Gemini and Chat GPT.

 

David Turetsky: 30:32

Wow!

 

Dwight Brown: 30:32

Really?

 

David Turetsky: 30:32

Figure it out! Except.

 

Nicholas Rhodes: 30:34

Yeah, exactly. So they're being asked to do this on their own time, as well, which is really interesting. So their argument is that they don't want any company information going into Chat GPT or those places, because once again, once that data goes in, it doesn't really ever come back out. No, it may not spit it out exactly the same, but the learnings from it never come out. But I think what I suggest people doing is, AI is really scary right now to a lot of folks. And when you jump in and use Bard, or sorry, Gemini, which is another funny name choice, considering there are so many other Geminis out there right now. But anyhow, if you hop in to Gemini, it's three right now. And these tools are not going to be free forever Chat GPT, you can use not the latest model, but a pretty great model to play with it that's free currently, too. And these companies are losing oodles of money so that you can play with these tools for free right now too. The compute power is not cheap on these and you can go in and literally just have a conversation with it. Ask it to tell you jokes, ask it to write new lyrics to songs, just to see the power of the tool. And then innately you're gonna start asking questions like, you know, the Turing test, for those who aren't familiar is just the to pass the test, the machine has to be able to fool you that it's a human.

 

David Turetsky: 31:53

Right.

 

Nicholas Rhodes: 31:53

And with Chat GPT, it can get pretty close. But one of the things that you'll probably do, if you've never played with this, is you as a human will not be able to help yourself quizzing this AI on how human it really is, right? And you'll hit barriers, and you'll hit walls. And it can be kind of a choose your own adventure game, this conversation, right?

 

David Turetsky: 32:13

Sure.

 

Nicholas Rhodes: 32:14

But I think what I would really suggest that someone can do is since these tools are free, you can take any one of your tests for the math example that I used earlier, or helped me figure out the best schedule for this or I have these ingredients at home, what are some good recipes, and you can start to see how the how it responds to you. And you can literally say if this is unclear ask me more questions to get clarification. You can also Google around and find pre existing prompts that help you build future prompts. So before I got very comfortable in Chat GPT myself, there was a prompt, that was a prompt builder. So you put it in as your first step. And then it would ask you a series of questions that help narrow down the question you were actually asking chatGPT.

 

David Turetsky: 32:58

Sure. There are a million things racing through my mind right now of things I could ask Chat GPT to help me with. But in the world of HR, that has to be more limited because we need to be able to focus on the human element of it, not just the data element of it. And because I can't bring in to your point before I can't bring in proprietary data, nor data about the people that I work with, then I still need to come up with this myself. Or at least that's my bias right now.

 

Nicholas Rhodes: 33:31

Yes. And no. I mean, I think I think that the data element, like we said, is really interesting, but I keep coming back to the activity concept of it. So there's so many HR activities that happen when someone's already a member of the team, whether it's team building, those sales kickoffs, we do a ton of sales kickoffs. There's a lot of ways to utilize these tools to help people imagine a different future. Or imagine the future that they're already aiming for. We did this really, really cool event with Axios out in San Francisco, where people were asked a series of questions to imagine what they wanted the future of San Francisco to be. And it created a completely new San Francisco and you saw the Golden Gate Bridge, in its modifications in the background, right. And so, myself, I'm not a painter, for example. But I can, as a creative director, I can describe a style of painting an artists that I really wanted to be inspired by, were three artists that I want it to be the combination of, and you can create these things. So it's not just about aesthetic, but it could also I could take a photo of myself and say, what would I look like as a CEO of 500 person company? And based on the data that's in these tools already, they'll find these CEOs that have 500 person companies and change me to look more like them. Right? And those are, once again going back to what's the data, you know exactly, but you can imagine these things in a way that or I should say you can see the things that you have in your imagination. And a lot of the times when we're dealing with our clients, they really know what they want until they see it. And they're like, Oh, that's not what I want at all. Right. So using it as a visualization tool. Even with drafting an email, you get that you're like, Oh, I thought that's what I wanted to say but the sentiments that I really need to be front and center aren't there. So let me go back and re work it.

 

Dwight Brown: 35:24

And there, there are ways to anonymize data too that you're feeding in to chat GPT and still be able to achieve what you're looking for with it. Yeah, kind of like the the Imagine San Francisco there's, there's inputs that go with that. And, and I mean, probably not a whole lot of privacy issues with that particular piece of things, but at the same time, to the, to the point that you're making with that, that there really is this creativity that goes with it. And you can apply that in many different arenas, especially in the HR arena, as long as you're careful about the privacy piece of things anonymizing your data that you feed in there, and, and being able to arrive at, you know, arrive at good points for what you're looking for.

 

David Turetsky: 36:16

But Dwight, the more data you put in, and the more specific data you put in, the better the model should be. But that then goes against the privacy and the simplicity or what you're talking about the masking,

 

Dwight Brown: 36:31

Depending depending what you're looking for. If you're looking for something on a specific candidate, for example, that is an issue. If you're looking for something, let's say you want to chat GPT to help you with an HR process, that's something totally different. And you, you know, you that's one where you you could run into issues with internal processes with the company you don't necessarily want to send out into cyberspace, but you can do some anonymization of the of the data to still be able to arrive at the same point. So your point is, is spot on. Yes, privacy, and there are elements where you can't you just have to be very, very careful. But there are also elements that you know, things you can build with it.

 

David Turetsky: 37:22

But let me ask I guess let me ask the question around that, though. And Nicholas, I'm asking this to you, unless Dwight, you can answer this too. The question I'd ask is, is there any way to have a private chat GPT server where you can do these things in a box and not worry about it getting outside of that box? And having that data leaked to the world?

 

Dwight Brown: 37:43

Air gap your data?

 

David Turetsky: 37:44

Yeah, exactly,

 

Nicholas Rhodes: 37:46

I think, well, so currently in the world that I work in, we can for generative AI for visuals, there's a tool called stable diffusion, which if you use the public version, then it goes into the same big pool, if you install it, it's open source, you can install it on your own server, and then bring in existing libraries. So you could in theory, erase that server and everything is wiped. My assumption is that as we see ChatGPT and Gemini, go into Enterprise Mode, there definitely are going to be siloed out scenarios. I mean, Gemini right now they sent me an email via Google saying like, you know, for a limited time only I can get feeds for $25 ahead to use Gemini. So I don't know, you know, it's going to be a very expensive tool, but they will silo it out. So you will not have to worry about your data being mixed with data outside of your scenario. Well, I think, but then you kind of lose some of the magic. So I think

 

David Turetsky: 38:43

Exactly.

 

Nicholas Rhodes: 38:44

It's about finding the happy spot where you have access to anonymized data that you could insert your data into, and then reflect on how it fits in those scenarios.

 

David Turetsky: 38:55

And I think as we grow as a people, people dealing with this new technology, whether it's the practices that we use, or the regulations that we're getting from the government will enforce on us, whether it's from the EU, which have already kind of put some regulations on privacy and data sharing about employees, or, you know, our world in the US. That's kind of going to be taken out of our hands a little bit. But we have to have good hygiene and good practices until that time.

 

Nicholas Rhodes: 39:27

Sure.

 

David Turetsky: 39:36

I think we could talk about this for another year, Nicholas?

 

Dwight Brown: 39:38

Yes, it's fun stuff.

 

Nicholas Rhodes: 39:39

We've barely scratched the surface.

 

Dwight Brown: 39:41

Really fun stuff.

 

David Turetsky: 39:42

Yeah, this is. I mean, what I want to do now and I'm kind of energized to do this is to kind of open up a sandbox in chat GPT and start playing and asking it some real world questions about whether work or personal stuff.

 

Nicholas Rhodes: 39:55

Yeah.

 

David Turetsky: 39:56

And seeing what kind of answers I got. I tried to early on and I wasn't good enough with prompts to be able to get anywhere? So I think I might actually have to learn a little bit.

 

Nicholas Rhodes: 40:04

Yeah, there. I will say if you're if you want to play, there's a pro version, I think it's $20 a month, it is worth the extra 20 bucks to play with. Because ChatGPT 4, I think 4.2 is what it's up to now, is significantly better than the free one. I know because I've run out of credits on the the pro one and have to use the I have to wait an hour or use the old one. But it's really it is wild. And I do encourage you also to check out those GPTs. You could search on them for HR

 

David Turetsky: 40:33

Yeah. All right, cool. Well, now our next tools, I'm sure someone built them already and dig into those as well. conversation with you will be much smarter about chatGPT and our use of it.

 

Nicholas Rhodes: 40:47

Sounds good.

 

David Turetsky: 40:48

Thank you very much for being here! Dwight, thank you.

 

Dwight Brown: 40:50

Thank you. Thanks for being with us.

 

Nicholas Rhodes: 40:52

Thank you for having me.

 

Dwight Brown: 40:53

I could talk about this for hours.

 

David Turetsky: 40:55

Thank you all for listening, take care and stay safe.

 

Announcer: 40:59

That was the HR Data Labs podcast. If you liked the episode, please subscribe. And if you know anyone that might like to hear it, please send it their way. Thank you for joining us this week, and stay tuned for our next episode. Stay safe