Bennett Sung - HR Tech 2024 - AI's Current and Future Role in HR - HR Data Labs podcast

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Announcer (00:01):
The world of business is more complex than ever. The
world of human resources andcompensation is also getting
more complex. Welcome to the HRData Labs podcast, your direct
source for the latest trendsfrom experts inside and outside
the world of human resources.
Listen as we explore the impactthat compensation strategy, data
and people analytics can have onyour organization. This podcast

(00:24):
is sponsored by Salary.com, yoursource for data technology and
consulting for compensation andbeyond. Now here are your hosts,
David Turetsky and Dwight Brown.

David Turetsky (00:38):
Hello and welcome to the HR Data Labs
podcast. I'm your host, DavidTuretsky, live at the HR
Technology Show in Las Vegas,Nevada, and I have with me one
of my best friends for a verylong time, Bennett Sung.
Bennett, how are you?

Bennett Sung (00:52):
I am doing well.
Thank you for having me back onthe podcast!

It is absolutely a pleasure, and especially when
I get to do it in person!

Right?

And actually see your face, your smiling face and
see your your new hair color!

Newest hair color, yes.

And it is a orangish reddish yellow,

yes,

with gray highlights.

It sure is. This is a three, it's three months
old, but it's still showcasingits color.

Yes

I don't mind the gray. Gray is great.

Yeah

Gray is trendy. I love it.

It shows your maturity, right?

right

I'm sorry. I didn't do that. No, I

Oh don't worry.

So what we do for every one of our guests on
the HR Data Labs podcast,Bennett, what's one fun thing
that no one knows about you?

Ooh, or any new fun thing?

We need a new fun thing

I've already told probably the most hilarious one
fainting from from aldehyde inanimal science class. That was
just to bring everybody back tothat, back to three years ago.

Yeah

Well, this time, you know, I had a bit of an
incident getting to HR Tech. Ina context like I had. I tripped
on these amazing new shoes ofmine that are red.

Yeah, they're Pumas. They're really nice!

you know, so, but I tripped over them, and I had
to, like, like, brace myself fora face plant.
No!yeah. And now, now I can barely
shake hands.

So when was this?

From Seattle to Las Vegas.

No!

yeah. So I got off the train in Bay in Seattle, and
boom, oh, I was on. I was on theground. I said, Oh my goodness.
What's going on here? This isnot how I wanted to start my HR
Tech travels. But nonetheless,it is what it is.

But you look okay.

Yeah, I'm fine.

Now show me your hands. Oh yeah.

Did they look bruised?

No, not at all.
So if, for those of you don't

It has been a while, and things have changed.
know Bennett, Bennett's abrilliant guy. I've met him a
long time ago when we wereworking for ADP, his company had
just gotten acquired, and mycompany just got acquired by
I mean priorities. Prioritiesdefinitely have shifted, yet at
ADP, and we kind of bonded overthat newness, but also we did a
lot of conversations aroundtalent management and how, you
know, working in an environmentof payroll and HR, and how do

(03:06):
the same time stayed relativelythe same.
you get more people thinkingabout the world of not just at
that point, applicant tracking,talent management, but also,

Yes

Because I think when I began to kind of look, do
what does that really mean inthe context of business? So,
right? That's been a while ago.
a reflection of, you know, thesince back in 2006 when we were

(03:35):
together.

right, wow

You know, so many things have stayed the same,
right? So many of the topicshave relatively stayed the same!

Right

Candidate matching. We're still struggling
to figure why we can't figurethat out, right?

right

And there's just a lot of small, little, small,
little, mini milestones we allhave to accomplish in order to
get get that piece offunctionality right. And I'm not
sure if we ever will be happywith it anyways. So

Well, it's so difficult, and especially in the
world now where, and we want totalk a little bit about
artificial intelligence, butyeah, within the world of
applicant tracking andrecruiting, there's still so
much complexity with where am Ihiring people, who am I going to
hire, and who do I actually getto talk to, given the fact that
a lot of the AIs are actuallyfiltering people out.

Yeah, for sure. I mean, there's, I mean, AI is
definitely helping with, youknow, getting through the
volumes of candidates,especially in the employer
driven world we're living intoday. I'm sure, you know, next
year will be a candidate drivenworld again, and that's all that
is going to be a differentstrategy. But nonetheless, I
think AI certainly has beenhelping folks out in the context

(04:41):
of, I'm getting work done foryou, I'm getting work done, I'm
screening. I'm looking atpeople. I'm not sure if they're
correct, but I'm looking atpeople, and I'm thinking and
predicting that this personshould be put in your interview
bucket.

Right

And not the disposition bucket, right?

Right

So that you know, we're going to continue to see
how that continues. See how thatevolves over time in terms of
its accuracy.

Yep

And you know, it's context of not making sure that
decisions are not based onprevious biases and such, and

which has been the case!

which is, which is why, which is why on the
recruiting side, it's very much,for me, one of the biggest
hurdles and challenges going tofacing recruiting technologies
today. All of them are quotingAI. They're making
recommendations.

right

You go talk to EEOC, which I have, they are
very critical about decisionmaking capabilities that are not
made, that don't have a human inthe loop.

Right, right.
So, so, so, I think one of thekeys there is have the
artificial intelligence helpcreate the artifacts, but have a
human review them.

I mean, the artifacts are still honestly
kind of subjective. I mean, youlook at resumes.

Oh, my.

We have no control over the resume or the CV,
however you want to talk aboutit.

Yep.

You look at the the the the content piece
created by recruiters and hiremanagers, the requisition that's
flawed, primarily becausethere's no process of collecting
it.

Right. So actually, a lot of that's being
built by chat GPT, becausemanagers and recruiters are,
sorry recruiters, they're lazyabout this.

Yeah! I mean, for sure, right? You know. And then
you have to think about, is itinclusive, you know? Then you
have to look at the languagesassociated, you know, that are
you being used? But more so it'sthe actual process of intake,
right? You know, is, because thewhole notion is, I as a
recruiter and you as a hiringmanager, really, this is our SOW
to each other.

That's right.

And if we can't agree upon this, then the first
step of the process of going outthere and looking at candidates
and say, yes, no, yes, no, Ithink sometimes humans probably
could do a little bit betterthan AI in terms of getting it
right. We'll see!

But go back to your original point about the
bias.

Yeah

Sometimes we've not gotten it right and and
what, hopefully, what we're nottraining these models on, are
what had been happening on thepast that had gotten us into
trouble in the past.

Oh, yeah, for sure. Yeah,

By the way, we're recording live at the HR
technology show, and it hasn'topened yet. So you're hearing a
lot of the work that's beingdone to get it to open.
Yes, exactly. In an hour or so.

Unknown (07:36):
Let's go to one of the questions that we were going to
ask you, which is, to me, one ofthe fun things about doing these
kind of conversations,especially at the beginning of
the HR technology show, is Ithink your your opinion might be
changed by the time the end ofthe of the show happens. Maybe
not too much. So one of thequestions is, beyond the hype
cycle, where does AI land in theHR stack right now?

So interesting enough. So I've been consulting
with a company called MeBeBot.
Me be bot, one of the hardernames to say, but still a fun
name. But we did a survey of HRfolks to get a pulse check on,
have they made progress since2016 when we first did the
survey, on whether or not theyhave they're ready for AI. Well,
there's a lot of things thathaven't really moved the need,

(08:22):
they haven't moved the needleforward, right?

right

There are obstacles in the way. Their AI
adoption on the HR side, incomparison to recruiting side,
is much slower, much moremethodical, and it's because,
you know, the chat, the one ofthe big ops, one of the
challenges is that now IT hasgotten their, they are driving
in a co shared relationship withHR. They are driving the

(08:47):
initiatives together! Sothey're, they're both kind of

good! like, you know, obviously, you
know, pros and cons and, youknow, having their conversations
and, you know, debates about,you know, where do we want to
take this? What do we want touse? You know, have we set up
the processes and the policiesat front like, right? Do I have
an AI steering committee that'sbrand new?

(09:09):
Yeah,

to to a lot of things, and, you know, that's
just, but it's all, it's all forgood, right? It's all for
ensuring that AI is treated, isdone in a responsible capacity,

yeah

or ethically, or however you want to use it,
right?

right

And it's like so.
So I think for the most part,there are a number of things
that are challenging HR in thecontext of moving this needle
forward faster, right? First andforemost, they're not following
the money. By following notfollowing money. I mean, they
are not partnering as well asthey should be with their CFO.

Why is that?

There just hasn't ever been that that kind of
relationship. But the realityis, the CFO holds the purse

absolutely

purse strings, right? It's like, but, you know,
so the things like, well, youknow? So we want to invest in
this technology, but it's a newline item. How am I going? What
is that? What do you need fromus to to, you know, get you to
sign off on a business case.
What's the business case? Whatare the what are the outcomes?
And by outcomes that we don'tmean how much more, what the
kind of experiences? BecauseCFOs will not buy things based

(10:18):
on experience. In fact, they'recutting technologies that are
exclusively to experience,because there's nothing on the
on the at the tail end thatshows me how much revenue I've
created or how much cost cuttingI've been able to save, right?

right, right.

So following the money is important for HR to
really master and build thatrelationship. And then you'll be
able to realize when you have,when you find the tools that
you're looking for, you workwith, you work with CFOs to then
go and really build a strictcase that's going to be for the
long term. Not like this is,this is not one that is going to
be cut for the next year.

Unknown (10:55):
Sure, I would like to think about this though, there's
also another side that the CFOwill be interested in which is
the risk side.

yeah.

And what are the risks to not only the adoption,
but what are the risks to nonadoption,

right!

Because so many of their competitors might be
adopting AI tools, an AI stack

yeah.

But, but also, we've seen a lot of AI in the
consumer world.

yeah

ChatGPT 4.0 is available for people to kind of

for sure sign up!

What happens if people do it and they expose
data, I'll be talking about thisa lot, and it's not been done
with IT's knowledge orinvolvement, and so it's being
done rogue?

Yeah, I think the reality is we all we already
know that It's already beingdone. So you have to just
realize that it's a good it'sactually a really good thing
that individuals are using AIfor their for purposes of just
getting themselves acclimated towhat, what is my what's the day
in the life of what I'm doingtoday? Going to how is that

(11:57):
going to change? Until theystart experimenting, they will
really, not really feel theimpact. And that's one of the
major benefits and kind ofmissions of AI. At this stage,
we're about changing thebehaviors of work, right?

Right

And if we like, so, so I we're encouraging folks
to experiment, but we have toexperiment with with guardrails.
Well, they have to develop theseskills, because they have to!

if not, it's going to overtake what they're
doing, and everybody else isgoing to be doing it, and
they're going to be like, Well,why didn't we invest in AI?

I mean, we're seeing that. And I think there's
also the reality is, like, youhave to understand the problems
you're trying to solve. This isnot like you mean, if you keep
on layering technology ontechnology, and they're not
really solving any real, realissues. Then there's, then
again, we're not going to whenit comes to the renewal and they
ask for the outcome that's nevergoing to be very clear.

Unknown (12:52):
Let's look at a consumer technology in the world
of AI that everybody adopts. AndI'm not talking about Siri or
Alexa. I'm talking about like, aGrammarly like or or think
about, you're just a spellchecker that, or the grammar
checker that you use in Word orwhatever.

For sure,

There are so many people that think that a
Grammarly is a, what a lazyperson uses it, or someone who's
uneducated, no, no. A lot ofpeople use it so that they,
they, you know, what's the rightthen versus than, yeah, or then
or right? What's the right? Youknow, spelling of right or
higher, or

Sure, the context

Yeah, I mean, and getting it right in a
business context is soimportant!

Exactly

But, but that's a consumer version of an AI
tool,

yes,

that people have just kind of built into a lot of
tools, whether it's makingemails work or,

yeah, I mean, it's all anything that is content
generation is probably going tohave a Grammarly, a Grammarly,
like functionality built intoit, without you maybe even
knowing!

Exactly!

And so a lot of times, there's a lot of tools
that we're probably using thatwe don't ever realize.

right

Like, you know, we'll take a step back into the
days of Virtual Edge.

That was one of those technologies that got
bought by ADP, that broughtBennett to ADP.

Yeah, but what folks didn't know is that it had
a built in candidate matchingtool that was using a machine
learning, natural logicprocessing, you know, algorithm
called ingenuine back in theday. It was like a desk, it was
like a premise based, likemachine learning tool. It wasn't
even put into the cloud yetuntil, until Virtual Edge got a

(14:29):
hold of it. But the reality is,not many folks knew about that
in the days and and so, youknow, it's we, it's been around,
let's just be real. I mean, AIhas been embedded in so many
different things. I mean, youtalk about, you know, Grammarly
being lazy isn't a calculatorthe same way?
Like, Excel, likemy mathematical capability of

Oh, yeah! comprehending mathematical
equations have definitely sloweddown.
Oh, sure, right.

But nonetheless, it's the we know in the back of
our heads. It's made us moreproductive.

Yep, right.

We can get to answers faster.

absolutely, right.

But when it comes down to investments in
technology, and we're asking forlots of new money, like,
sometimes efficiencies andexperiences will not cut it,
like we have to, again, not touse the term phrase follow the
money. The reality is, CFOs arelooking, how much are you gonna

(15:23):
rate, how much more revenueyou're gonna give me, or how
much money are you gonna reduce?

Right

Those are the two things, that's all they care
about. So so you have to haveyour you have to have that
business case tight and reallyfocus

absolutely

on the on those two line line items. So

absolutely

Announcer (15:41):
Like what you hear so far? Make sure you never miss a
show by clicking subscribe. Thispodcast is made possible by
Salary.com, now back to theshow.

I want to take us on a little bit different
directions. This is, this is alittle bit selfish for me, but,
How does itchange the game in recruiting
but you've been in therecruiting space for a long

Yeah.
when now they have to disclosethe pay range when they're in
time, and now we're seeing paytransparency. I have to take you
this direction. Pay transparencyis really huge in a lot of
different states, and it's andit's going to be huge across
many organizations who are usingthe, let's call it the highest
common denominator, whatever thestate is that has the most

(16:16):
rigorous regulations.
the midst of the requisition andin that candidate cycle?
Yeah. I mean, Ithink in the long term, it's
leveling the playing field,right? It's giving it's going to

(16:36):
make candidates, you know, maybeit's going to be a vehicle for
screening candidates out or in,

absolutely

right? Because they, you know, they have their
own personal expectations ofkind of money that they want to
make, right?

right

So I do feel, from a recruiting perspective, it's,
it can be used as a absoluteattraction, right, you know,
and, and the reality is, I thinkit also reflects culture.
Transparency is one of those,one of those cultural elements
that employees value, you know,and so I think it's, you know,

(17:12):
it's also then aligning, like,if it is something you really
put forth and prioritize, andit's gonna be reflected in your
company's entire, entire way ofcommunicating. So, yeah, so that
those are the things that when Ilook at pay transparency, or
anything that's transparent, AI,you know, AI is a it's all about
transparency! It's nothing. It'snot about hiding anything,

(17:32):
right, right? It's only going toget it's only going to get more
granular and more visible andmore accessible.

Absolutely. When I think about transparency in
the world of recruiting, I alsothink about, what are the
programs and how do I educatethe candidate and the employee
as to, what is it the valueproposition is that they're
working here for what are therewards there they have the
opportunity to get?

right

And also making a more mature relationship
between the employer and theemployee, because now you're
trusting your managers, all thestakeholders in this. You're
trusting them to make the bestbusiness decisions for the
company

totally Yeah, I, you know,
I think one of the things that

and the person.
candidates and companies don'tprioritize, or it's very rare to
see they focus exclusively onsalary, and never really, I
never really bring in all of theother influencers of
compensation. Benefits, how muchthey put towards your benefit

(18:28):
programs or contributions toyour 401k, all of these things
that don't what they call totalcompensation statements.
Yeah

Those need to be.
Those are amazing recruitingtools, if you you know when you
put it into, when you put itinto, put it into play. I just
don't see a lot of folks, theysay, Here's your here's your
offer and salary and and here'show many days you get off,
right? I just don't feel like Ihave a full understanding of,
like, what I'm what value I I'mgetting from the company, right?
I know what I could bring tothem, but I really don't feel

(18:58):
like I have a good sense of theend to end, understanding of how
much they're investing in me.

And I think once we get beyond the regulations in
pay transparency, what you'regoing to start to see is that
more companies are going to bemuch more open about their other
benefits and other pay payelements and things like that,
because the regulations thatexist only talk about base pay.
Why? Because it's so complex inthe world of everything else.

Oh gosh, everything

You know, what is an incentive? What's a sales
incentive versus a commissionversus just a short or long term
incentive?

Right

Because those things are so variable by
company, by culture, as youmentioned before, that the a lot
of the regulators, and one ofthe regulators I was speaking to
last week said, Look, we thoughtabout doing beyond base pay, but
we really couldn't get what agood definition of those other
things are. So to me, what'sgoing to happen is, once we get
beyond the regulations,companies are going to use their

(19:52):
culture,

yeah,

and they're going to and they're going to
say, well, our culture is atotal compensation culture, and
then people are going to learnmore from the beginning.

Yeah, you know, I also just feel like it's an
education for candidates andemployees. It's like, let me
tell you what, let me help youunderstand the total investment.
And that's as good. Once I hearthat, I think them, they're
going to be very like, they'regoing to embrace a lot of the
things that they're talkingabout, and they're going to
like, really revalue theorganization that they joined,

(20:22):
and it's going to give them moremotivation and and, you know,
and they're going to feel more,you know, belonged in the
organization and valued and suchso

and now, and that will be a better tool for
retention than trying to changetheir pay or giving more of
something.

Right. I mean, you know, I think, I think there's
some folks that no matter whatit's always going to be the base
pay and the salary, and that'sall they can look at, right? I
mean,

right.

And I think there's a good part of

pay the bills

they have to pay the bills, right? I gotta keep
the lights on!

right

But, but, you know, I think over time,
hopefully as more folks do likestart to kind of think about
their large, the larger picturein life, then they're going to
realize the importance of all ofthose additional pay elements.

Absolutely.
Hey, are you listening to thisand thinking to yourself, Man, I
wish I could talk to David aboutthis. Well, you're in luck. We
have a special offer forlisteners of the HR Data Labs
podcast, a free half hour callwith me about any of the topics
we cover on the podcast orwhatever is on your mind. Go to
salary.com/hrdlconsulting toschedule your free 30 minute

(21:29):
call today.
So let's talk beyondtransparency now. Let's go back
to our list of questions and getone last one, which I think a
lot of people are thinkingabout, is 2025 the year that we
see AI break into HR in a majorway? I know you have an opinion
on this.

Um, we're gonna wait another year. It's gonna be
2026 I believe.

So you think 2026 will be the year that HR
really adopts AI in a major way?

In a, in a, in a more in a scalable way.

Okay,

We're still asked there's, there's way too many
organizations who haven't evendone basic things around using
AI! Like again, I'll kind ofreflect back on MeBeBot, which
does AI employee support. Thereare 24 digital HR generalists.
Who doesn't like, whatorganization HR people ops teams

(22:23):
want to answer repetitivequestions every day?

None of them.
They hate it.

They hate it! but

slows everything else down

But yet no investment. So there's kind of
this like double edged sword, oflike, do more with less, yet
they never get the investment toactually augment the team to
help do more with less, in termsof, like, the overall head
count, right? So it feels so forme, I feel like there's still a
big stride away. There's a lotof cool things happening here at

(22:51):
HR tech!

Oh yeah!

Right? I mean, there's a lot of innovation, but
so much, like many, like a lotof things, it's all about
timing. It comes down to time.

Last year we saw a lot of hype cycle here at HR
tech around AI, this year is noexception.

Yeah, we're still in the delusional stage.

Oh my gosh, yes.

Delon, delusional.
We just don't because what, whatwe don't see is there's not
enough, like there's not enoughability to play with the
technology. So the AI, thedifference between AI and
applicant tracking system is, inthe case of AI, you have to vet
like, is it doing the job asit's designed? So, which means

(23:32):
you have to get access to thatAI algorithm, and you have to
play with it.

That's right.

between traditional software and what
we're seeing today in this AIsoftware. And yeah, it is
definitely kind of flipping orchanging the dynamic of how the

(24:00):
solution providers are actuallyselling.

Unknown (24:02):
well, you have to make sure that decisions it's making
would be the same decisions youwould have wanted it to make!

No, exactly. No.

At scale!

And that's what, and that's the hope. That is
why, also the solution providersare going to be pressured to
provide transparency on how thealgorithm works, right? Which,
which? I think a lot of folksfeel like, oh, that's, that's
like, secret sauce, yeah, Ican't give that away! Well,
you're not gonna have muchchoice, because you're gonna
have legislation

Exactly.

Legislation is gonna tell you. You gotta expose
this all

absolutely.

You cannot hide.
This cannot be hidden.

Well, if you don't have it through
legislation, you're gonna haveit through lawsuits. So,

right? Do you want reputation hurdle? You want to
like we already see in theconsumer side?

Yeah, right.

I mean, the consumer side just saw Air
Canada get through, go through alawsuit because their chat bot
was delusional about bereavementtravel policies.

Yes, I heard

right? And then you hear the same thing
happening in, I think, New York,New York City there are, they're
they're giving information,wrong information about
everything, because their chatbot is ingesting wrong
information, outdatedinformation. So,

yeah,

the reality is, we ought to hold everybody
accountable. You got, but youalso, you realize that you have
to understand, again, what arethe problems you're solving? How
would you go about solving that?
And it's no more different fromthe days of assessments, right?

Yeah, of course.

Assessments were one of those tools, early AI
tools that were done by, onpaper,

yeah,

and but nobody they had to provide the evidence
and the receipts to to defendwhat, what potentially could be
a lawsuit or a reputation issue.

Well, we also saw WorkDay.

We're still, we're, we're waiting for that
judgment to happen, right? Wekind of, it's been exposed. It's
nothing secretive, but now it'sgoing to be in, you know, who's
accountable

right

for the AI, the algorithm itself. I think this
again. This is history repeatingitself. It comes full circle
like we're we're here again, 20years later, 10 years later,
still talking about the samethings, maybe in a slightly
different context, but but theramifications and the thought
process is all the same.

Unknown (26:12):
but I think you mentioned before, transparency
helps provide that layer oftrust,

yeah,

whether it's talking about a chat bot,
Whether it's talking about an AIassessment of a candidate. And
why did you choose this one overthat one? And why did you choose
to, you know, let this one gofrom the process or pay
transparency or whatever, yeah,treating people with respect and
providing them the insight tounderstand why a decision was
made or how it

Everybody wants an answer

It has to and have to. I mean, we'll get
legislated, or we'll get alawsuit.

you know, and or we're just going to be where we
are today, with frustratedemployees and candidates, where,
you know, it's like, Why didn't,why didn't you choose me? Well,
you can't really say, I already,like, there are these, there is
these communication like, kindof restrictions about what you
can say, What you cannot say.
It's like, you know, let's, oh,we have to open this up! Like
there is no reason why youcouldn't tell somebody they were

(27:04):
not chosen for this particularreason. The problem is that they
actually, and probably the AIdoesn't actually know the
reason.

Right. They just made it.

They probably just made it up

Unknown (27:15):
Well, and for those of us who have gone through the
process of trying to apply for ajob,

oh gosh, yeah

and getting an email back five seconds after
you hit submit, that said, thankyou very much. You've got really
great experience, but we've goneon with other candidates who are
better suited than you,bullshit!

right, right? What they probably forgot to do is
close the requisition. That isgoing to be my guess, that the
requisition has forgot to beclosed. What they're saying is,
ooh, we've got a dispositionfrom these people, same way, and
this is how we're gonna do it.
We're gonna give them as themost generic email that you can
possibly exist.

But doesn't that? I mean, talk about
reputational risk there though,Bennett, I mean, isn't that,
like, really embarrassing?

It's embarrassing.
But you know what, who's talkingabout it? Like, nope. Like,
you'll get a few of thesenaysayers on TikTok.

Yeah,

we've heard them all,

yes,

right? I mean, they're, some of them are very,
very like, self promotional andwill put themselves out there
saying, Oh, my goodness, can youbelieve what just happened to
me?

Right?

Versus, you know, but most people just, you know,
they're accustomed to it.
They're just like, okay,

yeah, right.

There's not, I mean, there's nothing else I can
do about it, right?

Right

I mean, I can call them. Nobody answers the phone.
Nobody even responds to emails,right? So,

Oh there's, there's no phone number. You
can't, you can't call arecruiter and say,

I'm gonna find a way to say

why?

I'm gonna figure out a way to reach the recruiter

Why didn't you choose me? I'm the best
candidate!

exactly, you know.
So, I mean, does? I mean it'sjust, you know, it's a vicious
cycle of things that there's,like legacy practices that are
still in play that just have tobe kind of like, what, let's,
let's find a way to tell peoplewhy they didn't get the job, why
they were, why, what, why theircompensation is where it's at

right, right,

or why the answers to these questions are the way
they are at right? So it's, it'sall comes down to again,
changing the changing thechanging the behavior, to change
the culture, to reflect andchange and get different
outcomes.

Absolutely. And let's just say this, because I
did crap on recruiters before,and I apologize, I wasn't
calling them lazy. I was joking.
But recruiters have a tough job,especially these days, trying to
find the best candidates in avery, very big sea, and these
tools are trying to help therecruiter get the best person,
because it's about theirreputation. It's about the

(29:30):
recruiter's reputation too!

equally as that, yeah, for sure.

And they're looking for these technologies
to be able to make their lives alittle bit more livable,

yeah,

to be able to do that,

I figured, I think some of the technology, you
know, at the end of the day, youhave to look at the people
process and and people processand technology tools, right? So,
yeah, so the people are thepeople is not the issue.
Usually, the process I could beimproved a little bit. The
technologies are the ones thatyou know, have much further to
go, because expectations arejust great.

Oh, yeah,

right? It's like, I just don't want you to tell me
exact the exact match will nowalso expand that exact match.

Yeah,

do who potentially could do this? Like, there
should be tiers of candidates inyour pool that you until you
talk to them and talk to them,you'll never be able to get a
full sense of whether or notthey'll be a good fit.

That's right

fit into the organization, the role the team,
right? So, so at the end of theday, some of the technologies
have to understand, like, how doI build a pool of candidates
that are based on potential?
oh. I know!

Right? Well, and God knows, we also need
succession plans, right? Oh, andso somebody moves up. Well,
yeah, people leave, People go,

yes,

and then, and actually, that's a really
important thing. People aregonna be going a lot more now,
because the bubble of babyboomers, the people in my
generation, you know, Gen X, GenX, we're gonna be retiring!

Yeah,

I mean, I'm 57 so I got, I got at least 20 more
years

right? We'll be doing 20 more years of
podcasting

yeah, I'll be talking to Bennett in like 2037

Is it still gonna still be in the same facility?

The HR Tech show will still be here in Vegas.

It will, it will never go away.

Never! But, but, I mean, seriously though there's
gonna be this demographic bubblethat is leaving, and there are
gonna be a lot of holes left inorganizations.

Yeah.

I mean, look at recruitment, retirement
statistics are going to throughthe roof soon and and the
recruiters job is going to benot just about filling today,
but also filling tomorrow.

Yeah. I mean, I think this is why organizations
need to really get a handle onretention, right? Because the
reality is, I am it's likeretention that you can control,
right? And so, because the thethe it's just recruiting and
retention are the same coin, buton the opposite sides.

absolutely.

So we need to continue, you know when, when
retention is high, recruitersare just strapped to refill
seats

absolutely

versus refilling for the future,

right

So, so we got to help recruiters. We all have to
help each other by reallyaddressing being rigorous on
retention, so that recruiterscan actually recruit for the
future, recruit to to fill, findthe folks to be able to come in
to support or replay or so, youknow, replace the the the
retirements that are are goingto happen in droves!

absolutely.

And they need skills. They need experiences.
And so organizations have tofigure out what, how do I get
these individuals theexperiences that they need?

Yep,

right? And that's not so much saying they need
technical skills. Sometimes it'sjust, it's just, they just need
the experiences of being inleadership, the experiences and
doing specific types ofprojects.

or mentorship too!

Mentorship!

Being able to do it the right way, not just being
able to do it.

Yeah, exactly. So,

so we're gonna be talking about this for
another 20 years. So next, thenext 20 years, we're gonna talk
about how AI took over theworld.

Yeah, took over the world. But you know, what's
also interesting is this newgenre of chat bot called AI
agent.

Yes,

We've heard about it from WorkDay, Salesforce and
such, right? There's all theseagents coming about. What's
really exciting about that isthat it's really going to
change. It's really going to prothis version of AI that's
currently being pushed out isgoing to be able to be proactive
in seeing the areas of gaps inorganizations, organizational

(33:27):
systems, right, to then fill it,figure out how to get them back
in order, right? So, which isthat one of the hardest things,
which is, you know, why a lot ofcompanies have struggled with
their you know, the HR techstacks, is that they're so kind
of non integrated data is all,kind of all over the place, and
so it's very hard for them torealize, oh my goodness, this I

(33:49):
have all these gaps incompliance because people moved,
or we went from remote work toyou better get your get back
into the office, folks, whichmeans it's going to have a
trickle down effect on taxes, soand all sorts of other things.
But no, no.

So what you're talking about agents are, these
are little AI bots that thatserve a specific purpose. They
do a specific job, and they'retrained on one thing,

yeah,

and they fill that gap. Or it may be, it may
be what someone used to do, orit may be something completely
new.

It's probably somewhat somewhere in between.
It's because, I mean, anybody,if we had the time and the time,
enough, more time, and, youknow, to be able to go into our
systems and figure out the data,and then look at it and then
realize, oh, here are the 20compliance gaps that we have
right now, and here's a list ofthings that we need to do to fix
them all that's very routine.
Those are all routine thingsthat can be nicely performed by

(34:43):
an AI agent.

sure

Right? And so, so, so a lot of times the non
compliance things, because oursystems are just out of order,
right?
right Becausethey're missing something.
They're missing things. They'remissing they're not talking to
each other, they're missing datapoints, all sorts of things. And
so, so we have, it's, I thinkit's the AI agent is a great way
to maybe scale the fixing of alot of systems, because the AI

(35:08):
agent can't be really, you know,do its like true, automatic
automation, until they get thedata straight.

Unknown (35:16):
So, so what you're saying is, is that we need
another layer to help fix thedata, and then things will be
okay?

I think it's a good, it's a good, yeah, it's
not a it's not a patch, but it'sa good thing that it's a good
starting point, because I don'tthink people know where to do
instead of, like, starting fromground zero and implementing new
systems and importing all thisdata, it's like, Okay, I think
we can, we can fix existing datathrough these AI agents, and you

(35:45):
really get the actual integrityof the data in a in the in the
way that we need it to be sothat we can now just move on
from that, and then look atother AI tools to be able to
layer on top.

And those, those agents don't stop. They continue
doing their job, refining,refining, refining,

Always refining, and that is, that is a job in
itself!

Right

That's worth, it's priceless in my books!

Unknown (36:10):
Well, and since it's an AI, it's probably relatively
inexpensive compared to peoplethat you would have been paying
for doing that.

Yeah, exactly,

And they'd be bored out of their freaking
brains!

Yeah. Talk about things, probably some jobs that
will cause them to want toleave,

yes,

right? Unless they really, I don't really think the
most obsessed data person wouldwould want to stay for that kind
of job.

Nope!

you know?

nope. I think you're right.

But that's kind of, that's kind of exciting part
of like, where AI is, where AIis today, that compared to last
year, like, I mean, these arejust this, all of this AI agent
conversation, and we're seeingit in real life, is, is really
kind of another layer of vision,and

it's a refinement.

It's a refinement,

yeah,

and I think it's also going to help refine the
employee, the employeeengagement. So when you think
about, remember the days ofEmployee Self Service Portal?

Oh, sure!

ESS?

yeah. Well, well, they're still here, but,

Yeah, they're going to get replaced. They're
going to be replaced by these,chat, these AI agents for
employees.

So the AI agent is going to contact the
employees and say, Hey, have youupdated your w4 in a long time?

It's going to it's going to enable them to go in
there and say, Tell me how. Tellme I need to fill out my expense
report. Well, this is how you doit, and I'm gonna lead you
through. And don't worry, youdon't have to log into concur.
You don't have to log intoexpendify. We're gonna work

Here's an expense that we found on your
Amex. Give me the did you take apicture of the receipt?

Yep.

If you did, right? Send it to me.

Yeah, yeah.

Send it to me at this address. Yeah.

So it's really gonna, you know when, because
when you pull back the ITapplications that is being
supporting organizations. It's,it's astounding,

yeah

right? There are some companies that have 600
employees and have 600applications. It's like,
application, it's, it'ssuffocation,

yes

of applications and like, these are
applications, like, maybe once ayear I use it?

yeah, right.
It's like, who's the expert in

AI agent, hey, help me figure out how to use
this?
this one tool that I do once ayear called benefits enrollment!

Oh yeah.

So, you know, I mean, there's some exciting
things coming down when itlooks, when you look at like
trying to stream, simplify thesimplify the employee engagement
in these technologies, and thenalso, most importantly, really
helping leadership andoperations teams really get the
things that they need to getdone so that they can focus so
not to so they can focus onthings that you know, maybe move

(38:39):
their their career for it, orhelp bring in new, new solutions
to address other

scalability. It provides them a scalability that
they wouldn't have had, becausethey have to call every employee
to make sure that they're doingtheir benefit enrollment. Well,
I said, train the bot to do it,and then

Bot keep reminding these folks, go into the system.
Oh, who hasn't done it? Get sendthem another reminder. Oh, you
need help? We are, yeah,

let me help you!

Recommend. I can recommend you something
your dependents change? No. Nextstep now, yeah.
So exciting things. I mean,that's why they're calling, at
least in the healthcare world,they've called this digital
front door. The digital frontdoor to employee engagement,
which just enables organizationsto not to still have all these
disparate systems in the backback office, but really could

(39:27):
provide a seamless, one, oneuser interface to the employee,
which is really kind of the whatemployees crave, because it's a
consumer, it's a consumerapproach.

That's right.

right?

That's right.
Well Bennett, I think we couldtalk about this forever.

Oh yeah, give me a couple more hours.

But I know, but I know the show is going to
start soon.

Is it really?

Yes!

How long have we been here?

Well, we've been doing this now for 38 minutes.

There we go!

Isn't that before, actually 39 now!

right?

Bennett, thank you so much for your insights.
It's such a pleasure to talk toyou.

Always!

Thank you and take care and stay safe.

Announcer (40:10):
That was the HR Data Labs podcast. If you liked the
episode, please subscribe. Andif you know anyone that might
like to hear it, please send ittheir way. Thank you for joining
us this week, and stay tuned forour next episode. Stay safe.

All Episodes

Bennett Sung - HR Tech 2024 - AI's Current and Future Role in HR

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Bennett Sung - HR Tech 2024 - AI's Current and Future Role in HR