0:00

All right . Well , thank you everybody for joining us

0:02

on this exciting episode

0:05

of the Boring AI Show , where

0:07

we're going to be dealing with and

0:09

diving into the founder and

0:11

CEO of MeBeBot and

0:27

Bennett Sung , the Bennett

0:30

Marketing , chief Marketing Officer . What's

0:32

the right way to introduce you here ?

0:34

I'm a fractional Chief Marketing Officer

0:36

. Awesome , yep , awesome

0:39

and guru and guru .

0:43

And so , as always , Tally

0:46

is joining us today , Tally thanks thanks

0:48

for joining us .

0:50

Of course , sorry for the mishap there in the beginning .

0:53

Oh good , so let's

0:56

as we normally do . Let's kick off with

0:58

some news , Tally . What's happening out there

1:00

in the world of AI ?

1:01

Perfect . So there's a lot , I feel like . This week

1:03

there's been quite a bit of AI

1:05

news , but one that I think is

1:07

at least really exciting to me came

1:10

out yesterday by the New York Times . It's called

1:12

a Silicon Valley group . A

1:15

super group is coming together to create

1:17

an AI device , and

1:20

again , I'll make sure to link this after the

1:22

show . But essentially it

1:25

shows that the open AIs

1:27

, sam Altman and the former

1:29

Apple designer , johnny Ive , are

1:31

teaming up to develop a

1:33

device that could replace the smartphone

1:35

, backed by SoftBank's investment . So

1:38

again , a

1:40

lot of AI related news this week , but I think this

1:42

really speaks to the exciting possibilities

1:44

that AI is opening for um

1:46

innovators , I know . For

1:49

me this really makes me think of when the first

1:51

iPhone or smartphone with a touchscreen

1:53

was released and just how crazy

1:55

different that was , you know . I think it blew a lot of

1:57

people's minds , and this project

1:59

, I have a feeling , is going to end up doing the same

2:01

, it sounds like . So

2:04

this project is described as preliminary

2:07

, but essentially it would develop a

2:09

device that would succeed the smartphone

2:11

and deliver the benefits of AI in

2:13

a new form , unconstrained by the rectangular

2:16

screen . So really something extremely

2:18

different than what we're used to seeing Further

2:21

in the piece they talk about , you know

2:23

, really think , ambient computing . So

2:26

, rather than typing or taking

2:28

pictures on a device , imagine

2:30

a device that's like a pendant or glasses

2:32

or some sort of simple device that

2:34

could process the world in real time and

2:36

process images , you know , using a

2:39

smart virtual um uh

2:42

agent , um , to kind

2:44

of assist you through that

2:46

method , which is really interesting . So

2:48

again , mr Altman , sam

2:51

Altman , he's been , you know , he's invested

2:53

in a lot of kind of similar companies

2:55

, but one reason he may want to pursue

2:58

his own in this is to avoid dependencies

3:01

on existing companies for distribution . But

3:03

yeah , like I said , it's really

3:05

in the early stages . I just think this is exciting

3:07

to show the potential innovations and how much

3:09

AI really is going to impact

3:12

our world in the near

3:14

future . So I thought that was really exciting and

3:16

I'd love to hear everybody's thoughts

3:18

on this .

3:20

Yeah , beth and Bennett , would you guys wear

3:22

an Alexa necklace that

3:24

can just be listening to everything

3:27

you talk about with everyone and interpret

3:29

it .

3:30

Yeah , is that ? Is that how you're viewing it , tim

3:32

? As a necklace . I mean , I like it if it's

3:34

a bling and a necklace . That

3:36

sounds like a lot of fun . You

3:39

know the way I was essentially reading

3:41

the news on it . I I couldn't help , but

3:43

my brain went straight to the movie Minority

3:45

Report and I thought I've got to go back

3:47

in time and watch that movie again

3:49

, because when it was describing

3:51

glass and not

3:54

a physical rectangular device

3:56

, it really brought up a lot of imagery

3:58

but yet not really getting

4:01

your hands around what it's going to really be like

4:03

for the humane

4:05

solution of the future . So

4:09

I think that's what they're going to call it as a product

4:11

name Humane .

4:13

Right , I mean

4:15

that's like , that's like with uh , with the Patriot

4:18

Act came out and it's like come on , guys

4:20

, y'all have read enough thrillers and sci-fi

4:22

to know anything named like

4:24

for patriotic . You know , patriotism , it

4:27

usually has a sinister kind of overtone

4:29

to it . That's just dystopian fiction 101

4:32

. Exactly , call it humane , it

4:34

does not sound good

4:36

, that was a big little warning

4:39

bell for me .

4:42

So , bennett , what ?

4:43

do you think yeah ?

4:45

I mean , I think it's , think it's , you know , you

4:47

know I'm it'll be very interesting

4:49

to see and observe

4:52

how this , um , this

4:54

device you know , operates

4:56

. Is it going ? You know , obviously , you know , I've

4:58

seen the google commercials if you take a picture

5:01

and all of a sudden , google will tell you how to do this

5:03

and do that and everything about the

5:05

picture , and so if I don't have to

5:07

hit any more buttons you

5:09

know , you know and if it can

5:12

also prevent from , you know , pocket dialing

5:14

, that would be fantastic . That's

5:17

usually my biggest concern about phones but

5:20

nonetheless , you know , considering my phone

5:22

is my pocket most of the time , but

5:30

I think it's fascinating what the potential is going to be in

5:32

terms of its interaction within the community

5:34

and such , and is it going to help

5:36

me win more Pokemon Go

5:38

video games ?

5:41

Yeah , that's the value proposition . That's

5:44

my value proposition . This will be my value proposition

5:46

. That's awesome . So

5:55

, you know , I kind of I come back to this with years ago , like 2016, . I was

5:57

at a conference and it was . It was a BPM conference of business process management , and

6:00

they were talking about communities of systems and

6:02

I thought that was a really interesting topic

6:05

and kind of far-fetched at the time . But

6:07

as we get back , having

6:10

these AI assistants , one

6:12

of the things I think is really interesting is how are they

6:14

going to work together ? How would this tool

6:17

work with Siri ? How would this tool

6:20

work with Alexa ? Because , you

6:22

know , one of the things I love about uh

6:25

, you know , uh , apple's home kit is

6:27

like it is a centralized place to

6:29

work with all kinds of intelligent devices I

6:31

have through my house . So , you

6:34

know , I think making sure that it's not

6:36

built as a silo and

6:39

like built as part of an ecosystem , I

6:41

think that's going to be a real , like critical

6:44

part to all AI in the future

6:46

. And so you know , Tally

6:48

, you know what are your thoughts on

6:51

the Humane device ?

6:54

Yeah , I mean a lot of you know a lot

6:56

of products have been released and failed , so

6:59

there's no guarantee anything's going to be the

7:02

revolutionary product . But

7:04

to me it's just exciting because you know , because

7:06

I think it's been a while since we've seen , especially

7:08

with personal devices , a huge shift

7:10

in that area . I

7:13

know most of at least my group

7:15

we still have iPhones or Androids and smartphones

7:19

. Think of these new possibilities

7:22

and ways in which our personal

7:24

and professional world is likely

7:26

going to pretty drastically shift

7:29

in the next coming years , which is very

7:31

exciting .

7:35

Yeah , I agree . Very related to this have

7:37

you guys read Johnny Ives' book

7:39

Build that

7:42

?

7:42

is a phenomenally good book . No

7:45

, I have not , but I will add it to the major list

7:47

of books that I still have not yet read this

7:49

year that are coming out .

7:51

So the tbr , the

7:53

tbr maybe the cliff notes version

7:55

here , uh tim so

7:58

, uh , somebody I was talking to recently

8:00

is like um , you know , their , their

8:03

list has gotten so long . They're like well , well , you just

8:05

get it on Audible and put it at like 1.2

8:08

speed , oh there you go . Go

8:12

through it , but in this book the

8:14

narrator speaks pretty fast so I

8:16

would not encourage that . But really

8:18

good book that dives into

8:20

you know his mentality

8:22

and methodology around building products and

8:24

how he approaches it . Eddie Jenkins

8:26

at Mind Over Machines was the one who suggested it to

8:28

me . Really wonderful book

8:31

. He's a good one to get book recommendations

8:34

from . So if you all are following him on

8:36

LinkedIn you'll see he talks about all

8:38

kinds of stuff he reads because he has a long commute

8:40

and so he's audio booking it as

8:43

he's driving around Awesome

8:48

. Well , you guys want to talk about employee experience . We want to move

8:50

on , absolutely cool

8:52

. Well , before we go there , why don't

8:54

you , beth , tell us a little bit about who is me be bought

8:57

?

8:58

well , thanks to tim . I know you've been on

9:00

our , you know ask me be bought anything

9:02

. The Past , which was exciting to have you on

9:05

to help us in the world navigate

9:07

the world of AI and how

9:09

companies are utilizing these technologies

9:12

, and where we stand

9:14

at Me Be Bot is . We think a lot about

9:16

the employee experience and there's

9:18

a number of different types of ways that AI

9:21

can be applied , of course , to the world of work

9:23

, but if you think about what we've had happening

9:26

in our lives just in the

9:28

past few years , we've

9:30

all moved to working various locations

9:33

. Some of us are back in office , some are still

9:35

remote or going to continue hybrid

9:38

, and you need technology that

9:40

really connects employees

9:42

to their employer

9:44

to really make the company successful

9:47

. And so we at

9:49

Mibibot really think that what's been missing

9:51

is and talking about all the future

9:53

of digital glass

9:55

. It's really just a front door inside

9:57

the ecosystem of a workplace which

10:00

becomes Mibibot . We are

10:02

that front door that greets employees

10:05

, we help support their employee

10:07

needs , we push communications , take

10:09

feedback from pulse surveys and

10:11

we show data back to the business

10:13

. And the reason we do all this at MeBeBot

10:16

is we really believe in this concept

10:18

of if we're connecting

10:20

ourselves in the world as consumers

10:23

. Why aren't we doing more with inside

10:25

the workplace to connect employees

10:27

to all the different disparate

10:29

data that's out there with inside

10:31

the walls of an organization

10:34

? But , frankly , it has a lot

10:36

of its own challenges . Just like you

10:39

know the world of consumer , you know

10:41

products and AI technologies

10:43

where there's a huge

10:46

lens and I think we're going to hit on that a little bit

10:48

here in the conversation on how

10:50

do you do this in a secure way , you

10:52

know , within the walls of a business , to protect

10:54

the intellectual property of the company , to

10:57

protect the privacy of the employee , to

10:59

make sure information is delivered in a compliant

11:01

, consistent , accurate way , and so we

11:04

think a lot about that at MaybeBot

11:06

and how we're delivering our solution

11:08

to our customers .

11:10

That's awesome , and you

11:12

hit the core

11:15

topic when it comes to AI

11:17

. You were saying security in governance , and

11:20

so that was one thing . In

11:22

our pre-writer's

11:25

room meeting . We talked a

11:27

lot about governance and security

11:30

, and so I'd love to hear from you guys . Bennett

11:32

, if you want to kick us off here , why

11:35

can't the chatbot just

11:37

say whatever it wants in the world of HR ? Just

11:40

make up stuff on the fly . I'm sure it's probably accurate

11:43

. So tell us about governance

11:45

.

11:46

Governance . I mean , I think HR

11:48

in general has obviously a

11:51

reputation of one

11:54

of their major kind of DNAs

11:56

is compliance right

11:59

To protect the employees , to

12:01

protect the company , the employees

12:03

to protect the company , and

12:05

so really each individual within

12:07

HR and across the business really

12:09

are focused on risk aversion

12:12

. And

12:17

there's a lot of legalese around a lot of

12:19

information . So making sure that when you talk about specific

12:22

policies , whether it's an AI policy

12:24

, you

12:37

know , when you talk about specific policies , whether it's an AI policy , whether

12:39

it is a , you know , pto policy or you know anything else around the employee

12:41

handbook so much of it is governed by the legal team and so we have to continue

12:43

to protect it because it's obviously you know a big part of the communication

12:45

requirements out to employees , because the one thing employees

12:48

don't want at least I don't want is

12:50

I don't want to get misinformation . So

12:53

a lot of what I'm asking

12:56

of HR or

12:58

the line manager , my supervisor , is give

13:01

me the accurate information , give me the accurate

13:04

answers to my questions , and

13:06

I think that is the big

13:08

part of it . And so just having

13:11

AI come up with an answer

13:13

based upon a document you

13:15

know they may misinterpret certain

13:17

types of legalese that have

13:20

been rigorously reviewed and

13:23

run through a various approval process

13:25

. So I think , within

13:28

the workflow of how

13:30

HR and the business operates

13:33

, a big part of the

13:35

requirements at least currently

13:37

, until we all continue

13:39

to build trust in what

13:42

content is being created

13:44

from existing documents or

13:47

other system data is

13:49

this process of going

13:51

through an approval , going through some

13:53

level of verification that these answers

13:55

are correct , before they get published to employees

13:57

?

13:58

Yeah , Tally Beth , what do

14:00

you guys think Go ?

14:02

ahead Tally .

14:05

No , I think that's huge . I think Bennett hit the nail

14:07

on the head in terms of making sure misinformation isn't

14:09

spread , making sure we're confident with not

14:12

only the information we're using but that our teammates

14:14

or employees are using and we're

14:17

feeling comfortable with where data is going

14:19

and being stored . I think all of this is huge

14:21

and relates to some of the major

14:24

fears that I think a lot of employers

14:26

and individuals have with . You

14:28

know , learning how to use AI

14:30

in a really ethical and appropriate way

14:32

that doesn't put people or

14:34

organizations at risk . So

14:37

I think , yeah , that's obviously a huge piece

14:39

of it and I don't think a solution is , you know

14:41

, don't use it . I think it's figuring out how to use

14:43

it appropriately because , unfortunately or

14:45

fortunately , it's here

14:48

and , based on the data that

14:50

we have , whether you like it or

14:52

not , odds are your employees are using

14:54

it . So how can we coach people to use it in a

14:56

way that's responsible ? I think is where

14:58

we need to put our focus yeah

15:07

, absolutely , and that's what we hear from our customers .

15:08

We've been delivering a native ai solution from day one , leveraging , you know , natural language

15:10

, processing , machine learning frankly

15:12

, before the era of generative

15:14

ai , just a year ago , and now

15:17

we've had the ability to embed

15:19

generative ai within our products so

15:21

that our customers can take a

15:24

sensitive document like a

15:26

. This is a great example . You've

15:28

got an FMLA policy for

15:30

, you know , the US and then you have one

15:32

for California , right , and there are

15:34

other states that have different laws

15:36

around leave of absences related to

15:39

medical , you know , and types

15:41

of leaves that are very

15:43

critical to companies and their

15:45

employees . And so if

15:47

you leverage AI , to use

15:50

a generative AI to come up with an answer

15:52

on , say , a three or four-page

15:54

document that has actually

15:57

been reviewed by your legal team

15:59

, that applies to

16:01

the legal letter of the law of a particular state

16:04

and you're trying to deduce it down to

16:06

a paragraph so it summarizes

16:08

something for an employee , you

16:10

have to be very careful with that content

16:13

, because even one letter out of

16:15

place or two several

16:17

letters out of place could really redefine

16:19

how a policy is described to an

16:21

employee and , frankly , what companies

16:24

are worried about , because we sell

16:26

into a lot of the HR and those are

16:28

our customers as well as IT

16:30

teams , and it

16:33

could result in a lawsuit Right

16:35

right against the business , and so

16:37

they're looking at how do we leverage

16:40

the benefits of having the AI

16:42

, especially a generative AI , help us get

16:44

to more of a succinct answer , but , like

16:47

we need to have human eyes review that

16:49

and we need to have them verify

16:51

it to make sure that's the answer that people need

16:54

to receive and , if it needs changes , they

16:56

can make edits and lock it down so

16:58

that they know that it is in sync

17:00

with their governance

17:03

and the compliance processes that

17:05

the business is following , so that they

17:07

can be comfortable embracing this technology

17:09

without fear of repercussions

17:11

, frankly , in the future .

17:13

Yeah , and that human in the loop

17:15

. We get so enamored

17:18

with AI as this magic box

17:20

that can just do stuff and a

17:22

lot of times and we see it working

17:24

with processor engineering . What

17:26

is the value of the human in this process ? And

17:29

don't lose that . That's a great

17:31

example that human is providing

17:33

so much context that the

17:35

AI system just doesn't have . So

17:37

the human in the loop for

17:40

whatever the AI solution that you're

17:42

working on , people in

17:44

the listening crowd , keep your humans

17:46

in mind and don't forget . If you have a

17:48

comment or question for us , throw it into

17:50

the comments on the event and we'll pick

17:52

it up . But , bennett , I hear you're chomping

17:54

at the bit . What do you want to say ?

17:56

No , so it's like

17:58

it happens that A good friend

18:00

of mine , David Teretsky , who works at salarycom

18:03

, was commenting . Because he and I have worked in HR for

18:06

20 plus years , you

18:08

know he was referencing . The whole notion is

18:10

that HR operates in

18:12

compliance mode , always right

18:14

, and they are continuously . You

18:16

know it's not just their own company

18:19

policies that they have to keep

18:21

in mind , but the reality is the complexity

18:23

at the federal , state and municipality

18:25

level . So it is mission

18:28

critical that humans are always in the loop

18:30

, because those policies , those

18:32

that are being governed

18:35

outside of the company , are

18:37

, you know , those are continuously being modified

18:40

and changed , and so you have to keep , you

18:42

have to have someone who's very , very

18:45

on top of things in terms of making

18:47

sure that those are the right types

18:49

of answers .

18:51

Exactly , and just to elaborate that and

18:53

Bennett's spot on , you know , when

18:56

we meet with our customers and other

18:58

people that we're talking to , there's

19:00

always this curiosity

19:03

about how AI can help them in their

19:05

roles . But there is obviously , like

19:07

every role across the business , a concern

19:09

of what's this going to do to my job

19:11

? Well , the idea and I think what

19:13

everybody's been talking about is AI is going to help

19:15

you . It augments you to do your

19:17

best work . So if AI

19:19

can help summarize something that you

19:21

need to just use that critical

19:24

thinking that you have as a human being

19:27

to help make it better

19:29

. That's how this world's going to play

19:31

out , with people and the technology

19:33

working really symbiotically together

19:36

. And the more we keep continuing

19:38

to talk about this , the better , because

19:40

I think people have to understand and I think

19:42

, frankly , having the I call it the

19:44

sandbox of the chat GBT

19:47

out there in the world for everyone

19:49

to write a blog with or

19:51

to come up with an article or

19:53

recipe for something that they wanted

19:55

to make . You know , people are realizing

19:58

, you know , yeah , this is pretty good , but you still

20:00

need to get involved , and I think that was such

20:02

a great light bulb moment that happened

20:05

. You know , in a culture of you know

20:07

the AI community , when you

20:09

know it spread to more of a mainstream

20:12

effect right Of knowing that

20:14

this is powerful and helpful . However

20:16

, we still , as people , have a

20:18

role and responsibility

20:21

to continue to improve on

20:23

content . We continue

20:25

to improve on the training of

20:27

the AI , etc . And

20:29

last night I had a pleasure of

20:31

being at an event that was combining

20:34

internal comms people with

20:36

internal HR people and how

20:38

delivering messaging to employees

20:40

has changed with the usage of AI

20:43

. And , frankly , there was

20:45

a time when , a few months ago , I was part

20:47

of an internal comms webinar where

20:50

there was a lot of fear about AI for

20:52

people's jobs in that community because

20:54

they thought , wow , if this chat GPT can

20:57

write up content , then what's left for me ? But

20:59

I think there's become an aha moment

21:01

where people realize that there's definitely

21:04

this human in the loop aspect and that

21:06

embracing it is going to make them

21:08

you know , frankly better

21:10

, more productive , be able to prove

21:12

out the ROI of the results of what they're

21:15

doing and the effect . So

21:17

there's a tremendous benefits that

21:19

everyone can achieve .

21:20

Absolutely , and I think that's a great segue to

21:22

you know . I know the topic today

21:24

is really employee experience and I would love

21:27

to hear you know , bennett and Beth

21:29

and Tim , you know ways in which you

21:31

guys have used AI

21:34

either in your personal or professional life

21:36

and how that's maybe impacted your employee experience

21:38

and what you're hearing out in the marketplace

21:40

you know . I'd love to hear any use

21:42

cases in terms of how ai's impacted

21:45

um some of the employees and organizations

21:47

that you guys have worked with go

21:51

ahead , bennett , I'll let you kick this off sure

21:55

.

21:55

So you know , for for my personally

21:57

as a marketer , I think think

22:01

I'm a one-person machine and

22:05

while there is , within

22:08

the marketing profession , quite a bit of fear

22:10

, especially if you're a content developer , but

22:13

the way I've looked at it is , it's actually enabled

22:15

me to offset all

22:17

of the real tactical

22:20

work when you're putting together

22:22

content . So , when you think about

22:24

the whole process of building a

22:26

blog , you have to research

22:28

keywords , you then have to go find subject

22:31

matter experts , then you have to

22:33

write an outline , then you have to write the content

22:35

, then you have to proof it and everything else . The

22:38

reality is . The most important part of , for

22:40

example , writing a blog is going out

22:42

there and interviewing and getting proprietary

22:45

information from individuals

22:47

. That , for me , is where I want to spend

22:49

my energy , because that's where the

22:51

content is becoming really special

22:54

for the audience that I'm writing for , that

22:56

I'm writing for . So in my world I

22:58

use a hybrid of ChatGPT

23:02

plus . I also use an enterprise

23:04

writing tool called Writercom and

23:08

it's just allowed me to spin

23:10

up emails , spin up meta

23:12

descriptions there's so many little pieces

23:15

of content that need to be written and

23:17

such . So it's now allowed me

23:19

to do a lot more without having

23:21

to out , you know , get somebody

23:23

to help me out with it . So that's personally

23:26

how , as a marketer , I'm

23:28

beginning to use , beginning

23:30

to be able to do a lot more

23:32

without you know a

23:35

lot with not to use a cliche doing

23:37

more with less , but that's pretty much what it ends up happening

23:39

to be .

23:40

Right , yeah , and

23:48

I'll tie into that as well from a marketing perspective , not work-related , so

23:50

kind of stepping out of that . A lot of people know I like to write fiction

23:52

. It is a very

23:54

different skill set to tell a story in fiction

23:57

than it is to write an ad to

23:59

advertise that fiction . And

24:01

so , leveraging tools like Jasper

24:04

AI that's a huge tool

24:06

for me to say , okay , what are some

24:08

ad copies for this content that

24:11

I could use and give me 10 options

24:13

. And then further leveraging

24:15

AI when building out

24:17

those ads and meta , using

24:20

the creative , their ads

24:22

algorithms to select , well , what's the right

24:24

content to apply to this viewer

24:26

. And again , this isn't rocket

24:29

surgery , this isn't super experimental

24:32

. There's zero lines of code to do everything

24:34

I just described . So I

24:36

really think , from a content generation

24:38

, there's huge value there and

24:40

also huge

24:42

value where we use it

24:44

to help us understand content and

24:46

what might not be written in that

24:48

content . So we have our own large

24:50

language model . We use internally that we

24:52

use to analyze documents like RFPs

24:55

, and so it's a walled

24:57

garden , it's private to us . This data

24:59

doesn't go anywhere outside of the

25:01

data ecosystem , but we're using

25:03

these large language models to say what's

25:05

not being said in this document , that you've

25:10

done

25:15

it

25:18

before , which might sound like

25:20

well , of course they want to know that , but nowhere

25:23

in the document was it ever spelled

25:25

out like , hey , show us that

25:27

you've done this before , you

25:30

really , really know what you're doing and we're in good hands . But that's where the large language

25:33

model is like a recurring theme is you know , trust

25:35

and history

25:37

of performance , and so we really

25:39

wrote our RFP response to highlight

25:42

those things , and we heard it directly

25:44

from the client . We were the only ones to do that

25:46

because we , you know , instead

25:48

of just diving in , we had to help

25:50

her analyze it . So

25:53

those are fun use cases .

25:54

Yeah , that's a fun one , tim , and you

25:56

know I'm thinking of use cases , of employee

25:59

experience centric , for

26:01

example . The world of

26:03

people is really becoming

26:05

more of how do you personalize

26:07

the experience for employees with inside the company

26:10

, and so there's not

26:12

the one to many type of messaging

26:15

and outreach and even

26:17

learning and development programs

26:19

and recruiting efforts and career

26:22

pathing . It's becoming hyper-focused

26:25

on individualizing that

26:27

experience and some of the AI

26:29

solutions in the market today can really help

26:31

companies do that . There's

26:34

companies that are really great at writing

26:36

job descriptions , leveraging AI

26:39

so that you can do it in a very inclusive

26:42

manner and have

26:44

that really great lens to making

26:46

sure you're not creating a

26:48

job description with biases and you

26:51

make sure it has the lens toward the

26:53

diversity , equity , inclusion

26:55

and belonging efforts that you have with inside

26:57

your company . There's learning

26:59

and development tools that are

27:01

perpetuating the concepts

27:04

of not just micro

27:07

trainings and learnings , but very

27:09

individualized right Individualized

27:11

to what people are requesting

27:13

, as far as mapping it to

27:15

a career pathing tool as well . So

27:18

the more you can kind of combine technologies

27:20

from learning to getting people

27:23

down a path to becoming you know

27:25

in and moving and growing

27:27

into roles with inside their organization , I

27:29

think that's incredibly exciting because

27:31

at the heart of it all is you

27:33

know employers , you know care about their people

27:36

they really do , and they know that their

27:38

employees are a reflection of their brand

27:40

and their customer experience . And

27:42

most employees , the

27:44

average tenure at companies is two to three

27:46

years , and why ? It's because

27:48

a lot of times they're not getting matched opportunities

27:51

with inside the business , they don't know how to grow

27:53

within the company , and so there's a lot

27:55

of ways that companies are going and

27:57

personalizing messaging to

27:59

target that . We do some

28:01

of that at MaybeBot as well . With our

28:03

push messaging functionality . We can deliver

28:06

a lot of targeted messages to employees

28:08

within channels that can be hyper-driven

28:11

around . You know either

28:13

reminders or suggested

28:16

you know , you know direction or

28:18

a learning that may be applicable to

28:20

an employee . So that

28:22

is really the wave of what I see happening

28:25

, and the more we can do that

28:27

approach to really targeting individuals

28:30

with inside a company , I think it's

28:32

going to be a lot more impactful for the

28:34

organization as a whole . More

28:41

impactful for the organization as a whole . And then , personally , I really

28:44

like I'll put a plug real quick out to a tool that I've been using for years

28:46

called Beautiful AI , and it started when there

28:48

weren't a lot of tools like this . But

28:51

if you're not a great PowerPoint

28:53

designer and you need to come up with lots

28:55

of slide decks , you can leverage this

28:57

tool , and now they've embedded some

28:59

generative AI within Beautiful

29:01

AI so you can write a prompt description

29:03

in text and it will actually

29:06

produce some slides for you as a starting

29:08

point . Or they have

29:10

great templates that make you look like you're

29:12

a much more professional presenter

29:14

slash designer than you are

29:17

.

29:18

Very awesome . I also big fan

29:20

of beautiful , beautiful AI . I think it

29:22

is a such a powerful tool . Yeah , very

29:27

cool , tally . How about you ? What tools I mean

29:29

? You and I talk about this a bunch , but you know , for

29:33

the for the population here .

29:33

A lot of it is just been experimentation . Obviously

29:36

, I know at we use Jasper

29:38

, which has been great for me personally

29:40

, just because I am not the strongest

29:43

writer , so that takes a lot of time

29:45

for me to write any new content

29:47

. So to be able to tell

29:49

Jasper what I want and then edit from

29:51

there , just as a starting place to get me going

29:53

, has saved an insane amount of time

29:55

for us to be able to generate a lot more content

29:57

, um , which has been really really helpful

29:59

both internally and externally . Um

30:02

, I know we had fun recently um

30:05

developing . You know we've been doing some internal

30:07

trainings and , uh , I worked recently

30:10

with um . Shout out to Molly

30:13

at Mind Over Machines who was playing around and

30:15

experimented with . I believe

30:18

it's Adobe Express's new

30:20

. They have a feature that allows

30:22

you to input an audio file and

30:25

it'll take that audio file and

30:27

essentially

30:29

have a visual character

30:32

that will voice over to make it sound like it's coming

30:34

from that visual character . To make

30:36

it sound like it's coming from that visual character . So

30:41

it's a really easy way to just give some visual interest to any sort

30:43

of audio-based training modules . So that's been really

30:45

fun to play around with as well , but again , it's just

30:47

really been huge for any

30:50

sort of content generation . So I highly recommend

30:52

folks take a peek with what's

30:54

out there because it's super helpful and

30:57

really fun and interesting to use .

30:59

Yeah , and all of this is dependent

31:01

on having the skills to know how to use

31:04

this stuff and making sure people are familiar

31:07

with what are these tools and what are approved

31:09

and what's the policies . Given

31:11

that governance again , and one

31:14

of the things that we talk about often with

31:17

, we give somebody like okay , you

31:19

now have access to this tool . Now

31:21

, the very first thing we need to make sure you understand

31:23

is you are accountable for anything that comes

31:25

out of this system and

31:29

that talk . I feel like it's the

31:31

talk you have with your kids about drugs . It's the

31:33

talk you have with coworkers about AI . You are accountable for this . This is on you and your

31:35

actions . Kids about drugs . You know it's the talk you have with coworkers about AI

31:38

. You know you are accountable for this . This is on you

31:40

and your actions . So

31:42

you know , bennett and Beth , what are

31:44

you guys seeing from an upskilling , technical

31:46

confidence growing . You know

31:49

how are people handling that in the

31:51

market ?

31:54

Well , you know , I think it is challenging

31:56

right now . Frankly , if I'm being

31:58

direct , I think people are behind

32:00

on their learnings , I think

32:02

if you haven't gotten out there and

32:04

tried a few AI tools in your own

32:07

time . And this is a moment in

32:09

our lives where we're going to have to spend a little bit

32:11

of those free time hours and dabble in

32:13

a few solutions . But I

32:15

think it's moving so fast that

32:18

people can't keep up and

32:20

they're not quite sure even where to begin

32:22

, and so it's a

32:25

matter of you know taking some use

32:27

cases . Like Tim you mentioned , you like

32:29

to write fiction . Bennett has some

32:31

specific needs . So do you Tali and

32:33

just trying out a few technologies

32:35

so that you can gain a bit of understanding

32:38

of how this works . And then it takes

32:40

you down a path , because then , as soon as you play with

32:43

chat , tbt , then you want to know more about prompt

32:45

writing , and then you want to know more about

32:47

how it's really kind of coming up and what

32:49

it's really doing , and so you're starting to dig

32:51

into the machine , and

32:53

that's what we see . Frankly , some of our customers

32:56

doing is , once they see how maybe bot works

32:58

, they get to go well , wait a minute , how

33:00

did it come up with that answer ? There

33:03

you know , and , and being able

33:05

to use the technology and have your hands

33:07

on it , is much easier

33:09

to get that learning curve moving faster

33:12

than just to , frankly

33:14

, just live in the world of reading or yet

33:16

ignoring , which there is a bit of

33:18

that happening as well .

33:20

Yeah , yeah

33:23

, I agree 100% . And I look at

33:25

my own journey

33:28

from using generative AI

33:30

tools . You know it was like

33:32

, oh , for me it was a

33:34

mid-journey Well , actually technically night

33:36

cafe . So we're going way back . You know , so long ago I think it was a mid journey , uh , well , actually technically

33:38

night cafe , so we're going way back . You know , so long ago , I think it was like two years

33:40

ago , but

33:43

the images were , they were crazy

33:45

. Like what was coming out was like , you

33:48

know , twisted , nightmare , dream stuff . And

33:50

you know , yes , that's my art aesthetic

33:52

, I love that kind of stuff . So , and

33:54

then mid journey shows up with stable diffusion and it's like

33:57

whoa , the , this is different , this is so . And then mid journey shows up with stable diffusion and it's like whoa , this is different , this is

33:59

getting interesting . And then you start seeing

34:01

tools like runway and people making

34:04

these really engaging videos

34:06

with generative AI . And

34:08

then it's 11 labs and Murph

34:11

AI for voiceovers that

34:13

you know . So it really does

34:15

spiral up . And , with

34:17

all that said , you have tools like

34:19

ChatGPT , jasper , bard

34:21

, claude , pick your large language model

34:24

of choice that really do help

34:26

you in work as long as you are on

34:28

the guardrails , and you've set up

34:30

the guardrails so that you're not just

34:32

doing whatever , yeah

34:34

, yeah , whatever . So , yeah

34:36

, yeah , and that guardrail man . That's that

34:39

I think . From an employee perspective , playing

34:42

with the tools , experimenting with the tools , is really

34:44

important . Getting

34:51

you know , growing your knowledge and AI literacy is really important

34:53

. From a management perspective , you also need to be growing the AI literacy , but

34:55

from a different perspective of how do we build the guardrails

34:57

, you know what's the protections . We need to be growing the ai literacy , but from a different perspective of how do we build the guardrails , you know what's

34:59

the protections we need to have in place .

35:01

Um , and frankly , tim , you won't

35:04

know those until you try . You know

35:06

you've got to get in there and just and just

35:08

see where it's headed , because

35:10

that you know . It's kind of like how

35:13

probably people started driving in cars

35:15

. You know you had to get out there and try it before

35:17

you could understand like where's the road

35:19

begin and end , right . So

35:22

I think that's just the challenge is getting people

35:24

to have ways to overcome the

35:26

barrier . And you

35:28

know , I know you're out there with a lot of people

35:30

that are probably a little bit more , you know

35:33

, advanced in the world of AI and

35:35

we talk to people all day long that may

35:37

have a little understanding , zero understanding

35:40

or just want to learn , which is great

35:42

and it is pretty . It's

35:44

pretty disparate right now . There's people

35:47

all over in their levels

35:49

and , frankly , as much as companies

35:51

want employees to learn about

35:53

AI , I think that they haven't

35:55

offered enough opportunities to get

35:57

exposure to it within the business and

36:00

many times you think about where

36:02

the role of the employer is for a

36:04

lot of people . They get their health benefits from

36:06

them , they , you know they , they get the paycheck

36:09

from them . They also , at times , expect to

36:11

learn these types of things from them as

36:13

well .

36:14

Right , I

36:16

love the analogy of a car . I think that's a great

36:18

point because , you know , obviously cars

36:20

open up such a large world

36:22

and you're able to get from point A

36:24

to point B and have so many more opportunities

36:26

. But in order to get to the place where you can even drive

36:28

, you need to make sure you're taking your driver's test

36:31

, you're doing the research on how to drive

36:33

a car , what are the rules and regulations you're

36:35

sitting with , you know , hopefully your manager

36:37

or somebody you know who's a

36:40

little bit more versed , to kind of go through

36:42

your first couple runs of driving together . So I think that's

36:44

a great analogy here of how to kind of ease

36:46

into the world of AI and make sure that you have that

36:48

support and you're doing that background research to

36:50

get to the point where you can then start driving .

36:53

Absolutely Awesome .

37:00

Well , is there anything else ?

37:02

we want to talk about on this topic or we want to move on to wins . I

37:05

think I'm good . Like I said , I think my last little

37:07

bit would just encourage people to go

37:09

out and try different types of

37:11

things , even if it's just something for fun

37:13

. If you really like to cook

37:15

, know cook . You know use it for recipes

37:18

. If you really like to , you

37:20

know um build . You know create

37:22

beautiful . You know powerpoint presentations

37:25

and everyone seems to have to do that at some point

37:27

try beautiful ai .

37:29

So just get started yeah

37:32

, and and and you'll

37:34

see , um , some fascinating

37:36

things . So if you're a runner , you

37:38

you know , I know , like Michael and Nicole here

37:40

, he's a runner , and one

37:43

of the things I always joke about is there

37:45

was this example that was you

37:47

know , chatgpt was used to create

37:49

a marathon training regimen

37:51

and , like the day before

37:53

the marathon , chatgpt

37:55

had the person running like twice the

37:58

amount of miles as a marathon , with

38:00

the logic being well , you're

38:02

taking it easy on that next day , you're all set

38:05

to go . But the reality is a human body

38:07

is not going to handle that very

38:09

well . And so when

38:12

you are playing with the tools

38:14

, like this , you know , when you're asking it for a recipe

38:16

and it comes up with something , you're like , ew

38:18

, like who would eat that ? Now

38:21

you're seeing , these are the guardrails , like this

38:23

is what can happen in your business where

38:26

, like , if it starts talking about a policy

38:28

you don't have , you know it's

38:30

hallucinating , it's making it up , you know , and so

38:32

, um , you know in . Uh

38:34

, I always like to say you know , we , you know tech

38:36

, we call it hallucinating . Where I

38:39

come from , we call it bullshitting . That's

38:41

really what it is , and just recognizing

38:43

. That's what we're talking about here . So

38:47

, okay , cool Wins

38:50

, tally what you got

38:52

. What's the win this week ?

38:53

So I think our win my win at least

38:55

it actually really aligns with what we were just

38:57

talking about . You know , I think that there's

38:59

a lot of examples out

39:01

there of ways in which AI

39:04

can hallucinate or BS or

39:06

lean into and exacerbate

39:08

existing biases based on the data that it's

39:10

fed . The example I have

39:12

today it kind of flips that on its head . So

39:14

it was a Stanford article . It

39:17

was titled AI shows

39:19

dermatology . Educational materials

39:21

often lack darker skin tones

39:23

. So this has been an issue in

39:26

the recent years wherein

39:28

many medical textbooks or learning materials

39:30

lack images of various skin

39:32

conditions as they appear on darker skin

39:34

tones , which obviously could lead to really

39:37

, you know , dangerous

39:39

, you know misdiagnosed or delayed diagnosed

39:42

outcomes which could impact , obviously

39:44

, somebody's survival rate

39:46

. So it has really dangerous implications

39:49

for patients . So

39:51

in this use case , roxana

39:54

de Jong co-authored

39:57

a study introducing this

39:59

skin tone analysis

40:01

for representation in educational

40:03

materials . She called it STAR-ED for short

40:05

framework that uses machine

40:08

learning to assess bias in

40:10

skin tones in the frequently used

40:12

medical training materials . So this

40:14

is huge . She mentioned you

40:16

know she's not the first to do this A lot of folks

40:18

are manually going through these materials

40:21

to try and integrate a wider

40:23

range of skin tones to show what

40:25

different medical

40:27

conditions look like on a wider

40:30

range of skin tones . But obviously this takes a lot of time

40:32

. There's a lot of medical materials

40:35

out there , so she's

40:38

really using this tool , using machine

40:40

learning that

40:43

they've trained to detect

40:45

this human bias throughout all of these

40:48

medical journals and textbooks and things of that nature

40:50

. So I just think this is a really cool example of ways

40:53

to , instead

40:55

of having AI enforce some

40:57

of the bias , actually call

40:59

out the bias , so that way we can then make

41:02

different materials

41:04

less biased in the long run . So I just thought

41:06

this was a really interesting win and use

41:08

case of showing ways that AI could be beneficial

41:11

if it's used correctly .

41:12

Yeah , that's awesome .

41:14

I love that example .

41:15

That's amazing .

41:18

Awesome . So from

41:21

a win perspective , you know , Beth and Bennett

41:23

, we kind of sprung the

41:25

win thing on you . Sorry guys , that's

41:28

all right . Any

41:31

wins you'd like to share ?

41:33

Well , you know , on a daily basis , there's just so

41:35

many exciting things going on that

41:38

we're winning every day . We show

41:40

up and we're building

41:42

new technology , bringing it to and having

41:44

just people that are passionate , dedicated

41:46

. That's such a huge honor and

41:48

win to have on a team , for

41:50

sure . But what also is a win is when we have

41:52

customers who give

41:54

us some amazing compliments and

41:57

are super psyched to launch

41:59

our solution within their companies , and so we

42:01

have one customer in particular

42:03

that's going live to a specific group next

42:05

week and just to get that

42:07

general excitement of the team to

42:10

see something they've all really collaborated

42:12

on across the business to bring to

42:14

light , I mean it's fun , it's

42:17

energizing , it's something new and different

42:19

. Um , it frankly , often

42:21

, as times where you know HR

42:23

gets to be cool , they get to bring an HR

42:26

, an AI solution to employees

42:29

and HR was behind it

42:31

and especially if they're at a company

42:33

with a lot of tech people , they're liker

42:35

, all right , we see

42:37

you , you know so , um , and and

42:41

when they get the support of the it team

42:43

and other folks , I mean it's just they

42:45

. I think they get excited because it's it's

42:47

different . I mean , everyone has been in

42:50

utilizing a lot of database

42:52

systems for forever and you

42:54

know things and tools that are hard to implement

42:57

. And

43:04

when they see that , wow , this is all accessible , I understand it and I can actually do something like

43:06

this , it really makes them feel good as people right

43:08

, like they're achieving something with inside

43:11

their business , and

43:13

that's been very exciting because

43:16

it's happening all week long . We've been having

43:18

conversations leading up to it . That's

43:20

awesome .

43:22

And all of us who've been in huge

43:24

projects and

43:26

by that I mean just lots of team

43:28

members trying to all row the same direction

43:31

. When you all get there , it

43:33

is awesome . It is so awesome

43:36

, so cool , excellent win . Thank you , beth

43:38

Bennett . Anything on your side ?

44:00

with mentors and I'm just like my mind is blown away at

44:02

all of the different use cases that are popping up that I probably

44:04

wouldn't have never thought about in uh , in solutions that are leveraging

44:07

generative ai for doing

44:09

everything from helping me helping

44:11

out with , like , advertising workflows to

44:14

having a chef come in and cook

44:16

for you every day , all the way through to

44:18

, you know , caregiver marketplace

44:20

. So so what I'm ? You know

44:22

this big win and shout out to all the start

44:24

, you know founders who are building these new companies

44:27

with all of these very interesting applications

44:30

. It's , it's very inspiring

44:32

to to see the

44:34

you know this this the-stage

44:36

startup world just be so

44:38

kind of really massively

44:42

experimenting with anything and using the latest

44:44

kind of waves of these macro

44:46

technology kind

44:48

of tailwinds . So I'm excited

44:51

for the health of the

44:53

early-stage startup companies who

44:55

are using AI to build

44:57

a business Awesome

44:59

.

45:00

Very , very cool . These

45:03

are great wins . So

45:05

for my win this week , it's

45:07

definitely around the Writers

45:10

Guild and

45:12

in the Writers Guild description around

45:16

, that AI can't be a

45:18

I believe the wording they use is like an originator

45:21

of content , and so

45:23

writers can use AI

45:25

to create content , but the writer is who

45:27

is accountable for it . That

45:30

AI cannot just be like we

45:32

use this system to create this content , use

45:40

this system to create this content . That is a very big , big thing , because what

45:42

it leads to is there always needs to be a human accountable for the ai's work . Um

45:45

, that the ai cannot be like

45:47

the source . Um , and so , yes

45:50

, while while we look at that and say

45:52

, oh , the writer's strike

45:54

, that's important , but when you start thinking about

45:56

the ripple effect in society and in business

45:59

of , like , well , why'd you make that decision

46:01

? Well , the AI said to do it . Oh

46:03

, yeah , that's not an answer . You

46:07

know you can't do that Because

46:09

the AI couldn't originate . So very

46:12

fascinating to me and from

46:14

my perspective , the win about this is we

46:17

are taking the first steps of clarifying

46:19

AI's role in society along

46:21

with humans . That's huge

46:23

, so very excited about it

46:26

. Yeah , that's cool

46:28

. All right , we

46:32

are almost at time Beth

46:34

. Thank you guys for joining us today .

46:36

Thank you Always a pleasure , Bennett Beth , thank you

46:38

guys for joining us today .

46:40

Thank you , always a pleasure .

46:48

Any parting things to share or anything that you guys just want to kind of

46:50

say on employee experience as we go out the door

46:52

? I don't know , Bennett , if you want to give

46:54

a preview to some announcements next week or how

46:56

you feeling about that . I'll let you do that

46:59

. Yeah , as

47:01

we earlier described , we have a press release coming out next week . Um , so you all are

47:03

getting the sneak preview to that uh

47:05

, where we're announcing our digital employee experience

47:08

ai platform . That is using

47:10

embedded , you know , generative

47:12

ai within our solution to again

47:15

provide another opportunity for

47:18

our customers to leverage

47:20

existing content . They have to deliver

47:23

accurate answers to employees with

47:25

a verified approach . So just

47:27

getting to market , like the concept

47:30

that there's a lot of buzz and noise

47:32

about generative AI and

47:34

what does it mean , but there are

47:36

ways that you can embrace it today

47:38

within the enterprise and feel

47:40

good about it , and getting

47:43

education and messaging out

47:45

. There has been one of our kind of key

47:47

drivers from day one , so that's kind

47:49

of an exciting announcement

47:51

that you all are privy

47:53

to in advance awesome , very

47:55

exciting .

47:55

So very excited , super looking forward to this guys so are privy

47:58

to in advance . Awesome , very exciting , very exciting , super looking

48:00

forward to this guys . So thank you

48:02

again for joining us today . Thank

48:05

you to Beth Bennett

48:07

you guys have been really great . Thanks

48:09

to MeBeBot and all

48:11

that you guys are doing in the community . Mind

48:13

Over Machines . Thanks for sponsoring .

48:15

Yes , thank you , and

48:21

Tally as always thank you for keeping us

48:23

on time . Yes , tally tim , thanks for inviting us

48:25

. We always have fun chatting with you guys .

48:28

all right , everybody , have a wonderful day and

48:30

we will talk to you . Oh , next week . Um

48:33

, actually we'll talk to you next week , so I almost almost

48:35

forgot Due to scheduling

48:38

things . We will have another Boring AI Show

48:40

next Friday , so keep an

48:42

eye on LinkedIn . We'll be sharing

48:44

it here shortly and we'll

48:46

see you all there talking about healthcare

48:48

. So healthcare and AI , that's great

48:51

.

48:51

Good topic , awesome

48:53

, that is very important All

48:56

right , thanks everybody , thanks

48:58

guys , thank you all , bye .