Where AI and People Meet

As we look toward 2025, we’re reflecting on how much AI has made deep inroads into our future of work. From creating new opportunities for automation and entrepreneurship to its impact on how we hire new talent, AI will no doubt continue to stay at the forefront of our conversations. But we need to remember that there are actual humans coding the AI, responding to the outputs and implementing change. Where does the intersection of people and AI convene? Are we embarking on a new set of technical and people skills?

To future-cast what this relationship might look like in 2025, we’re excited to welcome back esteemed guests Vaneese Johnson and Chalenge Masekera.

Learn more about UC Berkeley Extension @https://bit.ly/41InVKV

Read this episode's transcript @https://bit.ly/3FB1iii

Episode Transcript

(00:00):
[MUSIC PLAYING]

(00:02):


CHALENGE MASEKERA (00:06):
Last year, I thought of AI as a tool,
as in, it doesn't do anythinguntil you go and prompt it.
It's actually becoming moreof a collaborative partner.
And also, how we use itin 2025 going forward
is more of an agent.

VANEESE JOHNSON (00:18):
It just makes it much more efficient.
And you reallystart to look at AI
more as your partner, asopposed to something that you
need to fight against,or try and work against.


JILL FINLAYSON (00:31):
Welcome to The Future of Work podcast,
with the BerkeleyExtension and EDGE
in Tech at theUniversity of California,
focused on expanding diversityand gender equity in tech.
Edge in Tech is part of theInnovation Hub at CITRIS,
the Center for IT Researchin the Interest of Society
and the Banatao Institute.
UC Berkeley Extension isthe continuing education arm
of UC Berkeley.

(00:53):
As we look forwardto 2025, we're
reflecting on how muchAI has made deep inroads
into our future ofwork, from creating
new opportunities forautomation and entrepreneurship
to its impact on howwe hire new talent.
AI will no doubtcontinue to stay
at the forefront ofour conversations,
but we need to remember thereare actual human beings coding

(01:15):
the AI, respondingto the outputs,
and implementing change.
Where does the intersectionof people and AI convene?
Are we embarking on a new setof technical and people skills?
To futurecast whatthis relationship might
look like in 2025, we'reexcited to welcome back
esteemed guests VaneeseJohnson and Chalenge Masekera.

(01:36):
To reacquaint you, Vaneeseis the boldness coach
and helps leaders bringauthenticity, intrinsic values,
and new levels ofengagement to their work.
She is also the authorand founder of Girl,
Get Your Business Straight andGirl, Get Your Career Straight.
Chalenge is a datascientist currently working
at Pharos AI, acompany dedicated
to enabling enterprisesto get invaluable insights

(01:58):
into their engineeringoperations.
His passion lies in harnessingthe boundless potential of AI
and data driven insights.
Welcome, Challenge.
Welcome, Vaneese.

VANEESE JOHNSON (02:07):
Wow.
I've been waitingfor this topic.
Like Christmas.

This is going to be a veryinteresting conversation.

JILL FINLAYSON (02:17):
Absolutely.
Welcome, Chalenge.
How are you doing?

CHALENGE MASEKERA: I'm doing great. (02:20):
undefined
Excited to be here.
It's been a yearsince we last spoke,
so there's new stuff, and yeah,always excited to be here.

JILL FINLAYSON (02:27):
Absolutely.
And we have discussedAI previously,
but it really seems likegenerative AI or ChatGPT,
Claude, Perplexity,and many other systems
have become mainstream,been adopted in 2024.
So before we lookinto the future,
how would you say your ownjobs have changed or evolved
over the past 12 months?

(02:48):
And did you find yourselfusing AI tools more often?
Vaneese, why don'tyou kick us off?

VANEESE JOHNSON: Yeah, actually, I (02:53):
undefined
started experiencing the changesin my business using AI in 2023,
when I was firstintroduced to AI.
And at first, Ibrushed it to the side.
And I'm, like,oh, it's something
else new I gotta learn.
And then one day, I think Iwas struggling with a task.
I was writingcurriculum and trying

(03:16):
to be creative with ideas.
And so one day I said, letme try this chat thing--
[LAUGHS]
--this ChatGPT thing, andsee what is it all about.
Literally, as soon as I gave ita prompt and it gave back to me
information, ideas, waysto think differently,
I was, like, whoa,wait a minute.

(03:37):
What is this?
And from that point forward,for the last two years,
I really have been incorporatingdifferent elements of AI
into my business to helpme to really be more
competitive in the marketplace.
I've startedincorporating it to help
me to have enhanced customerrelationships, client
relationship management tools.

(03:59):
I've incorporatedit to also help me
with managing my time, as wellas mundane tasks in my business,
and also for some fun.
So that's how it's impacted me.
And I'm really looking forwardto what's to come, what else
I can learn with thistool and teach others.

JILL FINLAYSON (04:16):
And Chalenge, same for you,
you're more in the techand engineering software
side of the house.
You've been using AIperhaps much longer.
But has 2024 changedhow you've seen it used?

CHALENGE MASEKERA (04:27):
Yeah, I think there's been actually
really good changes that hashappened, as Vaneese has said.
I think there's morepeople who've actually
started embracing it more thanthey would have normally done.
And personally, I thinkthe amount of time
I've spent using AI has grown.
Even more so, I think it'sjust come from being something

(04:47):
that I think about asin, like, I'm stuck,
I need to do something,to something that's
as I said, the firstepisode AI native,
like, I think in termsof I want to do this,
how can I use an AI tomake my life better.

JILL FINLAYSON: That's interesting. (04:59):
undefined
So from troubleshooting tonow being a starting point.

CHALENGE MASEKERA: Yes, absolutely. (05:04):
undefined
It's so much moreproductive for me.
And like, I don't knowif I'm losing something,
but my productivityhas gone so much higher
because of how I use AI.

JILL FINLAYSON (05:13):
Well, that reminds me.
You once referred toAI as an accelerant.
What does that mean to you?

CHALENGE MASEKERA (05:19):
What I think of it in terms of an accelerant,
I think what's the onething that you want
to have in life, which is time.
You want your time back.
And using AI has given me thetime back to do other things,
or just be able to accomplishmore in everything that I do.
So in that case, Ithink of an accelerant
as you have an idea, howdo you get to accomplishing
that idea in much faster time.

JILL FINLAYSON (05:41):
And when you think about AI,
do you think ofit just as a tool
or is this something different?

CHALENGE MASEKERA (05:46):
I was waiting for this question.
The answer is yes.
And by that I mean Ithink the big change that
has happened for 2024, lastyear, I thought of AI as a tool,
as in it doesn't do anythinguntil you go and prompt it.
But what we're seeing now isit's actually becoming more
of a collaborativepartner and also

(06:10):
what's going to be probably howwe use it in 2025 going forward
is more of an agent.
I think we weregetting into this world
where there's two typesof, if you have a business,
there's two typesof labor, which
is the human labor and alsothe AI labor, or machine labor.
So we're seeing people,actually, or companies adopt
agents that run independently.

(06:32):
And I think at some point weas humans will be able to,
just like in a normal life,have this AI-powered agent that
will do the work that you wantindependently, autonomously
from you, and you canjust come and review.
And I think that's wherethe world is going.

JILL FINLAYSON (06:47):
That's amazing.
How would you define an AI agentfor people who haven't really
used that term before?

CHALENGE MASEKERA (06:52):
This is one of the fun ones
where it depends on who you'reasking, because there's so
much branding that's going on.
The classic definition of howit's being branded right now,
it's called anAI-powered assistant.
And when you talkof an agent, it's
somebody who's doingsomething and working.
So these AI-powered agents are--

(07:14):
I don't want to say humanreplacements, but tools
that are AI-powered and machinelearning software that's
able to do tasksautonomously without humans.

JILL FINLAYSON (07:25):
Yeah.
So, Vaneese, are you seeing thiswith individuals and companies
that you're working with?
Are they taking onthese AI agents?

VANEESE JOHNSON (07:33):
I think what's different
with this side of the coin isthat the individuals working
in the organizations arestarting to embrace them
as it relates to some of themundane tasks in their work.
However, thecorporations themselves
are not embracing it asfully, and because there still
are a lot of undefined areaswith AI in the workforce.

(07:56):
And I think companiesare just approaching it
cautiously as they aretrying to figure out
what type of guidelinesneed to be in place.
I think companiesdo want to use it,
and they want to beefficient with it,
but I also think they wantto be effective with it
so that it's not doing a hugeshift in business overnight,
and I will say somecompanies that we've seen

(08:18):
and that I've seen, some of myclients in some of these areas,
more on the retail side, when itcomes to inventory management.
So when you see that, becauseyou want your customers to be
able to go and order product--
I mean, we're inthe holiday season.
So you want to be able to getpeople what they need quicker,
sooner, faster.
But you also wantto be able to know,

(08:39):
from an inventory controlperspective, what do
you need to be stocking up on.
Because that's goingto impact the revenue,
and it's going toimpact the bottom line.
So in that instanceyou're seeing
more of the adaptation ofit quicker, sooner, faster.
But on the otherside of the coin,
where you may havemore human interfacing,
you may have moreadministrative type of roles

(09:01):
in organizations, you'reseeing a very slow adaptation,
and especially if a businessis not inside of the IT world.

JILL FINLAYSON (09:09):
So share a little bit of those specific use
cases where you're seeingindividuals starting to use it
and then, maybe,Chalenge, you can
explain how can companiescreate more of those guidance
to enable this to happen.
But what are someuse cases that you're
seeing where people areslowly bringing it in
to be faster and quicker?

VANEESE JOHNSON (09:29):
What I'm seeing is more from a data literacy
perspective.
So a lot of peoplenow are using it
to be able to doresearch a lot quicker,
sooner, faster,and then allow AI
to help to interpretthe data, to help
them to be able to get tothe next steps of a project
that they're working on.

(09:49):
So in that case, I see that.
Also, I see itwith communication.
A lot of individualsare starting to use it.
We've had Grammarly fordecades now, an AI tool.
[LAUGHS] And so now,people are starting
to see more of the value ofAI, because the cross-cultural
communication is reallyimportant, and understanding

(10:09):
what's appropriate and how tocommunicate with colleagues,
with other global partners.
So you're seeing thatembraced, those areas.
And also, theother part I'll say
is innovation and creativity.
You're seeing thatnow, where people
want to really explore a littlemore with some of the work
that they're doing.
I'm seeing that when it comesto presenting presentations.

(10:33):
A lot of my clientsthat I go and do
training with on presentations,they're taking now
into consideration theirtarget audience, what
would be important forthis audience to hear,
and what would be reallykey points to zero in on?
What is it that you say less of?
And so a lot of them now,in the corporate space,
are using these AI toolsto build their slides
and/or enhance their slides sothat they can communicate more

(10:56):
efficiently and effectively.

JILL FINLAYSON (10:57):
I really like that prompt engineering
where they're asking,here's my deck,
but please customizeit for this audience
or for the needs of this group.
And making it more personal,but also more relevant and more
compelling.
And the employee feels better,because what the employees now
are experiencing islike, oh, I don't
have to stay up all night tryingto figure out what to say.

(11:18):
I don't have the pressureand the stress behind, what
am I going to put on the deck.
So now the employee-- when I gettogether with my clients, now
we can build it in ashort amount of time.
Then we can get to practicingthe delivering component of it,
and that's wherethe magic happens,
is that AI isn't deliveringyour presentation for you.
That's where you, the employee,get to shine in your innovation

(11:40):
and creativity.

CHALENGE MASEKERA (11:41):
You hit on the key point
that I actuallylike the most, is
like most of what the gen AIproduct has been able to help
us is the mundane Stuff.
What's important for you, Ithink, is for your clients,
is being able tomake the presentation
and actually showwhat their value is,
instead of creatingthe slide deck.
That's not what you wantto spend your time on.
And that's what sort of genAI, at least in the past years,

(12:03):
has been able tomake strides in.

JILL FINLAYSON (12:05):
It sounds like the folks
that Vaneese isworking with are sort
of where you werea year ago in terms
of using AI to solve problems,but not making it part
of their daily routine.
What do you think, Chalenge,has to happen to make AI
more integrated into workflows?

CHALENGE MASEKERA: It's a journey. (12:22):
undefined
It's like a journey of everytool or anything that you start.
You start somewhere.
It's like, oh, maybeI should, like,
just sprinkle a little bit.
Then over time, as you adaptand you become more comfortable
with it.
I'm like, on theprivilege end maybe,
or on the high end ofthe spectrum of tech.
I'm kind of forced to beusing it because we also

(12:43):
work on building AI products.
For me, it's something thatI have to actually embrace.
But I think what'shappening now is, like,
we've come fromChatGPT came out,
and all these other largelanguage models came out.
It's a prompt box.
It's very, very generic.
It hasn't been, like-- butwe're getting to a place
where they're building moresort of specific models,
or specific usecases, for tools.

(13:05):
I think Microsoft now, theyhave their own copilots,
which are very specific.
Like, you're workingon a Google Doc,
or you're workingon a Sheets doc.
They have an AI that'sspecifically built for that.
So as people start to, Ithink, as companies adopt,
pay into these copilots,we're going to see people just
naturally be able to use them.
I think once you have it,if you have a copilot,

(13:26):
after a while, you're goingto be forced to just say,
hey, what do youthink about this?
This is like where you have togo somewhere else and say, hey.
I have to open ChatGPTand ask this question.
I think as technologybecomes more embedded
in sort of all theapplications that we use,
we're just going tosee the adoption grow.

JILL FINLAYSON (13:43):
And the tools are making it easier
as time goes by.
You still have to have alittle bit of tech expertise
to do some things.
But increasingly, that'sgoing to go away as well.
One of the things thatI've really enjoyed in,
I worked with a student tocreate a RAG, or a Retrieval
Augmented Generation, wherewe got to use the ChatGPT
interface to ask questionsof a small subset of data.

(14:04):
And that, to me, was reallyactionable and valuable
because every companyhas their own rules.
And being able to accessthose rules more efficiently
seems like a great use case.
Are there otheruse cases that you
think, as you're talkingabout kind of onboarding,
we go from problem solving,what are some of the use cases
that you would see companiessort of taking steps

(14:25):
to more adopt?

CHALENGE MASEKERA: What I'm seeing a lot, (14:26):
undefined
depending I think a lot ofpeople I've worked with,
is basically anythingthat you want that needs
speed and performance.
That's where [INAUDIBLE]or what Vaneese
told, like, the novelty,where your style,
where your personalitycomes through.
I see people more usingthe LLMs, more that, just

(14:47):
like the AIs.
So anything that's like,oh, I need something
generated super fast, or canyou synthesize this information
super fast.
So you can see onetypical example
is, you get on a sales call.
I need a summary ofwhat we talked about.
What do you think was sort ofthe main point that my customer,
or this customer,or what's important
to this specific individual?

(15:08):
I'm seeing a lot of adoption.

JILL FINLAYSON (15:10):
That's a good point, because getting a summary
is one thing.
Getting a list of action itemsfrom that call is a step ahead.

VANEESE JOHNSON: There are AI tools (15:17):
undefined
that you can plug intoother platforms that
will summarize a meetingfor you, and then tell you,
here are the next stepsthat you need to do.
It'll summarize themeeting so that you can
send a summation to the client.
You now get your next stepson how to move forward
in the sales process.
And so like Chalenge issaying is that it just

(15:39):
makes it much more efficient.
And you reallystart to look at AI.
You can start to look at itmore as your partner in this as
opposed to something that youneed to fight against, or try
and work against.

JILL FINLAYSON (15:51):
And I think this is all about business
intelligence, gettinginformation, and summaries,
and insights moreactively, more efficiently.
We've talked aboutusing that for business.
But is there somethingthat would be.
I don't know, a careerintelligence equivalent,
Vaneese?
How can we use AI for ourown personal development?

VANEESE JOHNSON (16:09):
Oh my goodness.
There's so many great uses.
But what I tellclients initially,
when we start the beginningof this conversation,
because I don't thinka lot of people really
are aware of how prevalentAI has been in our lives.
And so that's the firstplace that I start.
When we look at theecho, we look at Alexa,

(16:31):
when you look attalking to your phone,
you look at when you'regetting ready to use your GPS,
we already have been using AI.
It just wasn't called that.
And we saw the efficiency of it.
So when I'm starting topoint that out to people,
they're like, oh.
That's what that platform is?
That's where-- yeah, that'sAI that you're using already.
So if you can tell it toturn things down, turn things

(16:53):
on, turn things off, you'reessentially talking to it,
and it's helpingyou like a helper.
The other component is peopleare concerned about, Jill,
is this going to take my job?
I think the thing that we alsoforget is that we as a people,
and as a culture,we have already
been through severaldifferent stages of evolution

(17:14):
in the economy.
We were in the information agein the 1950s when computers were
first introduced.
Guess what?
We adapted.
Then we went to the digitalrevolution and the automation
age in the '90s.
Guess what?
We adapted.
We also went throughthe digital age.
Remember thedigital divide where
we felt like, oh, thecomputers-- everybody needs

(17:35):
a computer.
People are not goingto have a computer.
What's going to happenwith the people that
are in certain jobsthat don't have access?
Companies found away to make sure
that we try to equalizeit as much as they
could to make sure peoplecould get access to it.
And here we are nowin the AI revolution.
And so there's anopportunity for us
all with upskillingand reskilling.

(17:56):
What we have to help us tobetter position ourselves
as employees of firmsto really leverage
this technology ina way that can help
us to advance professionally,but also to help us
advance personally.
You can get your own coach now.
You don't have to wait, andgo find, and pay for a coach.
AI can help you in ways thatyou can become coachable.

(18:18):
But that has to do withyou engaging with the tools
so that it starts to learn yourpersonality, your skills, what
makes you shine, whereyour deficiencies may lie.
So the more thatyou use AI tools,
that can be a benefit to be ableto help professionals to really
figure out, where do I go now inmy career with the change that's

(18:39):
happening in the market?

JILL FINLAYSON (18:41):
So how are you advising people
to update their resume?
How do they even talkabout these skills?

VANEESE JOHNSON (18:48):
Yeah, so that's another thing we look at.
And I always tell this to myclients is to really look at--
number one, writea career vision.
You need to knowwhere you're going.
Even if your visionis to retire,
you still need to have avision for what's next for you
and your career.
I also tell them to look atwhat's happening, the trends,

(19:09):
for the next five to 10years in their industry,
and also with theirprofession, because when
you can be future focused interms of what's coming down
the pipeline, it allows you tobetter prepare today for what's
going to happen tomorrow.
And sometimes, todayreally does come tomorrow,
tomorrow, like in 24 hours,because sometimes, these tools

(19:30):
get thrusted on you.
Or the type of work that youmay do gets thrusted on you.
And you have to really figureout right away what to do.
I mean, a lot of employeesare experiencing shortages.
So when I tell peoplein advance to get
prepared, and then I askthem to back into it.
So now that we know, orhave a sense of what's
happening in the nextfive to 10 years, so now,
let's dial it back and let'slook how technology is going

(19:52):
to impact the work thatyou do, your profession
and how it's impactingyour industry.
And so we just drillit all the way down.
And what I do is I allowthem to use ChatGPT.
And I say, well, let's playwith Chat and let's just see.
Because guess what?
I'm not going togive you homework
that you got to gotry to figure out
the answer to these questions.
We can get answersto them immediately.

(20:14):
And then that withinitself gives people
a different perspectiveof, OK, so now we
can put together a game plan.
So people are feelingmore confident.
Also with the resumewriting, people
are looking at ways thatthey can enhance their resume
with the right words,the right phrases,
qualitative information,quantitative information.

(20:34):
And so they're no longer justwaiting on a resume writer
to do that for them, becausesometimes, the opportunity comes
from you networking.
And you may have avery short window.
Or you may be limited interms of financial resources
that you can't necessarilypay for that kind of support.
So I think when people start toenvision these tools in a way

(20:54):
that it's not justspecifically for work,
but it also can enhance yourpersonal and professional
development, I think it'llbe a whole different kind
of relationship as theymove forward with it.

JILL FINLAYSON: And Chalenge, sort (21:05):
undefined
of a flip side of thatquestion for you, how
are job descriptions changing?
What are new rolesthat are coming up?

CHALENGE MASEKERA (21:12):
Vaneese mentioned a great point, which
is as any newtechnology, especially
if it's, like,transformative comes,
there's always areshifting, reskilling,
and some peoplewill get upskilled.
And I would be lyingif I say there's not
going to be jobs that are goingto be lost because of gen AI.
I think we've already seen that.
But there's also new rolesthat are also coming up.

(21:34):
I think one of the sort offun ones that I kind of like
is what's calleda prompt engineer.
There are specific focusis like interacting
with the generative AIs andpredicting what people ask them.
So somebody's jobis, like, questions
that somebody is going to askChatGPT or any other chatbot.
And we've already seen sortof a explosion of those roles.

(21:54):
There's also now people who arein charge of the basic stuff
like AI safety.
How do we put awire fence like AI
so that it can besafe and be used?
What are the legal implications?
What are the culturalimplications of AI?
So there's alwayssomething new that
evolves from how any newtechnology that's coming.
And for people who are likeme, you become an AI engineer.

VANEESE JOHNSON (22:17):
What I'll add to that too,
Jill, and Chalenge, onthat is that even though we
see AI kind of opening thedoor for more tactile skills,
it's also going to createan opportunity for people
to really become human centric,because AI can't really interact
with you human to human.
It can only give yousome suggestions.

(22:38):
So this is going to be anopportunity from a leadership
perspective for leaders to finetune their interpersonal skills.
It's going to be an opportunityfor individuals, professionals,
to enhance their emotionalintelligence skills,
and an opportunity to enhancetheir communication skills.
So there's the hardand there's the soft.
And I think it's going tobe important to understand

(22:59):
that it's a marriagebetween the two
so that one isn'tlost on the other.
But again, theyare collaborating
to enhance this personpersonally as well as
professionally.

JILL FINLAYSON (23:11):
I like that balance
of gaining moretechnical skills,
but also gaining more humanskills, and more leadership,
and communication skills.
I think we do need ablending of the both.
Do we have to go back toschool to get these skills?
Or is there some other way?

VANEESE JOHNSON (23:25):
So there's a couple of things
that employers can do.
So one of thethings employers can
do to help in theenvironment, because this
is-- work is where we spendthe most of our time, right?
When you get off work,I mean, let's be honest.
Who's going home to say,ooh, I can't wait to get home
to learn more about ChatGPT?
But what employerscan do is to start
creating AI learningenvironments, where there's
training that's happeningfor employees to become

(23:47):
more familiarized with the tool,and more specifically, how they
can use this tool toenhance their productivity,
and as Chalenge mentionedbefore, to give them time back.
I think approaching itfrom the give yourself time
back is more of agame changer than it
is to approach it and say,you need to learn this tool.

JILL FINLAYSON: Yeah, I think you're (24:03):
undefined
hinting at a brilliant idea.
And I don't know ifanybody's doing this,
but if there waspeer-to-peer coaching,
like if you're inan accounting role,
here's what otheraccountants are doing.
If you're in aadministrative role,
here's what otheradministrators are doing.
I feel like the peoplewho have figured it out
should tell the other people.

VANEESE JOHNSON: Yeah, that's going (24:21):
undefined
to be important as peer-to-peermentoring internally.
So you've got training formore of the technical skills,
the hard skills.
And then there is peer-to-peeropportunity for training.
And then there'sapplication training,
where individuals getto actually go and apply
what they're learning inthat peer-to-peer connection,
and also what they'relearning in training.

(24:41):
But I think it needs to becomenormalized language when
it comes to workperformance, and use it
from a perspective ofenhancing the output of work
as well as saving time.
Also, it's a way toinvite individuals
who have been less proneto be creative, less prone
to be innovative in their role.
I think it willinvite and create

(25:01):
an opportunity forthose individuals
to play a little bit.
And I think it'll makeit more acceptable to be
able to come to projects andmeetings, and saying, hey.
Here are a fewdifferent scenarios
that we could possibly look at.
Or, here's a differentway to interpret the data
with these different modelsthat we're doing sales or client

(25:23):
relationship management.
So I think there's stilla really great opportunity
to create the environment that'simportant to allow us to thrive.

JILL FINLAYSON (25:31):
Almost a bottom-up model of people
solving the problemsthat they need to solve,
and then tellingthe administration,
this is how we'regoing to do it.

VANEESE JOHNSON (25:39):
Yeah, yeah.
And I think the administrationwill be open to embracing,
because the administrationstill is leading.
The administration is reallylooking at data analytics.
The administration is reallylooking at technological tools
to enhance the waythat it does business.
So it cannot do everything.
And so if you're moreon the ground floor,
if you have more high-touchcustomer service type

(26:01):
of positions, or you havemore high-touch type of work
across the organization,or if you're
working with globalpartners, again, this
can just enhance the outputand the productivity.
And the person that'sdoing the forward reaching,
they can start to feel thatmuch more better about the work
that they're doing.

JILL FINLAYSON (26:18):
Chalenge, I feel like you also
said you've had to do thisbecause you're in an engineering
environment.
So you've needed toadopt these technologies.
Is there somethingthat we can learn
from the software side of thehouse about continuous learning?
And how have youbeen allowing people
to test and try new things?

CHALENGE MASEKERA (26:35):
It is my bias because I work in tech,
and we are always tryingto-- at least the way
I wear my head is we're alwaystrying to be the cool ones,
trying to be ahead of the curve.
I think, but oneof the things, I
don't know if many peopleabout which Vaneese mentioned,
is at least in all the placesthat I've worked so far,
we usually do whatare called, demos.
Sometimes, they'relike, oh, this
is what I worked on this week.

(26:56):
But sometimes,it's just like, oh,
I found this really new thingthat I've been testing out.
And at least in ourcompany, we're super slow.
So we kind of have that funand ability to try out things.
And our demos nowadays, theyset it up all being engineering.
But now, it's pretty mucheveryone across the company.
And especially when thatfirst wave of AI came,

(27:19):
we're seeing a lot ofpeople, sales people,
marketing people was like, hey.
This is what I found this week.
And I've been using this tool.
This is what it did.
Like, oh yeah.
We were talking about nextsteps from a sales call.
It's like, I was tryingto create this marketing
image that we want toput on our website.
And I put it in one of theother AI generative tools.
It created all these images.

(27:40):
Something that I would normallyhave taken half the day,
it took me a couple of minutes.
And I think that sort of wayof thinking about things,
bringing back the fun to worksometimes, saying, hey, this
is what we shouldbe focusing on,
or this is what weshould be doing.
And this is how we canget better at the things
that we're doing willhelp organizations a lot.

(28:01):
And even as an individual,I remember when I was still
in school, well, itdidn't happen to me,
but I'm sure some people werelike, don't use calculators.
It was frowned upon that youshould not use a calculator.
But now, nobody, ifyou go to school and--
yeah.
So I think that sortof mindset of, these
are the things that actuallyprovide value to the company,

(28:22):
and being able to focuson those, than sticking
to, oh, this is the waywe've always overrated.
And also being able to havethose open conversations
with the management, like, hey.
I think we've been doing this.
These are the thingsthat we can do better.
And these are the AItools that are coming up,
or just like anyother tool that we've
found that will helpus be more productive.

JILL FINLAYSON (28:43):
I want to hashtag that, bring back
the fun.
How can we make peoplereally feel ownership and be
like, hey, I think I founda better way to do this,
or I found a new tool todo this better or faster?
I think that is part ofmaking organizations,
academia, businesses, moreinnovative is to empower people
to try and solve thingsin new and better ways.

(29:05):
I would love to see thatadopted by other divisions
outside of engineering.

CHALENGE MASEKERA (29:08):
Yeah, so it's actually funny.
When I was, like,reading through for this,
I was going through,I think it's BCG.
And one of the points thatthey were talking about,
what are the inhibitorsof AI being adopted?
It's like, management support.
I was like, this is interesting.
And I'm pretty sure almostevery sort of consulting group,
any recommendation foranything they talk about,

(29:30):
is always management support.
There's always that slidefor management support.
And I think that'ssort of usually
where sort of the focusand the problem is.
And as long as seniorleadership doesn't
provide a way for loweror middle-level people
to embrace new technologies,or just adopt new ideas,
it's going to be pretty hard.

JILL FINLAYSON (29:50):
I always have to temper things
with a little bit of skepticism.
With all thisincreased productivity,
is AI actually making us better?
Are we actuallygetting more free time?
Vaneese, what are your thoughts?

VANEESE JOHNSON (30:01):
I will speak for myself.
And the answer is yes.
I remember pre-AI,pre-chat, that when I--
I am a corporate trainer.
That's one of thehats that I wear.
And I remember literallyhaving to set aside
between six and nine hours towrite curriculum for a course.
And I literallydid that for years.

(30:25):
And so my brainwas trained to it,
I made sure that I hadnothing else to do, because I
was literally going to beall day writing curriculum,
and sometimes all night tryingto find the right images that
conveyed the message, tryingto find the right design
layout for the slide,researching all
of the different articles andpapers, all of these things
that I was researchingbecause I wanted to make sure

(30:47):
that the curriculumis valuable content,
and it's rooted in theinformation that I'm providing,
it can support the data.
And so when I gotintroduced to ChatGPT,
it cut down my time75% of the time.
So now, instead of mespending hours of researching,
I can ask Chatand other research

(31:07):
tools to bring that informationto me quicker, sooner, faster.
And then that allowed me to beable to sift through information
quicker so that I couldzero in and extract
exactly what I wanted to share.
Also, likePowerPoint, PowerPoint
has enhanced that softwareso much with AI tools.
So now, instead of mehaving to figure out

(31:28):
how to design a background.
So now, PowerPoint can giveme different iterations.
Even in their owntoolkit of imagery,
of videos, thingsthat are in action
that people liketo see movement.
So it really has givenme some time back.
So I literally can go to bedat a normal time of night
when I have to deliver--

(31:49):
preparing for curriculum.
Now, I can just put moretime on the front end
of talking to the clients.
I can spend more timeinterviewing the participants
in the training programso that I can better
understand their individualneeds and expectations.
Now, I can take that and plugit into part of the research
process.
I couldn't do that before,because I didn't have the time.

JILL FINLAYSON (32:09):
And it gives you a more polished look as well.
I remember the dayswhen people would
be putting orange andgray together or-- no 18
different fonts.
And now, you get the templateand it just looks a little bit
cleaner as well.

VANEESE JOHNSON (32:21):
Yeah, and then I can look and say,
I can put what myoutline is into Chat
and say, how does this flow?
Is it a progressiveflow in doing this?
So I think it'sreally important is
to look at perhaps some of themundane areas of a person's job
to really see where you'respending most of your time on.
And is where you're spendingmost of your time on things

(32:43):
that you really enjoyin the work that you do?
How productive doesit allow you to really
be when you're doingthat specific work?
So I think when we startto treat our careers,
you may have heardme say this before.
And this is new to Chalenge.
But we start treating ourcareers like a business,
and we step back, and we startto take more of an overview
of, how are we doingthe work we do?

(33:05):
And then when we alsolook at the vision
that we have for ourselves inthe next step in our career
journey, I think it allowsus to be able to see,
OK, what type of tools can I usefrom AI to help me to get there
quicker, sooner, faster,but yet being efficient,
and allowing meto really showcase
my own professionalbrand that is of value?

(33:27):
I'm not showcasingthe brand of AI,
but I'm gettingto really showcase
the brand that I have as avalued contribution to a team.

JILL FINLAYSON (33:37):
And are you realizing time savings
as well, Chalenge?

CHALENGE MASEKERA: Yeah, of course, (33:41):
undefined
even if I have to say that.
But I actually, a goodchunk of the work that I do,
I think that I'm actuallyseeing a lot of time savings.
Things that used to takeme like maybe a day,
I can do some of these thingsnow in a couple minutes,
because now, what I'vebeen sort of playing with
is, like, if you use any ofthese sort of generative AI

(34:05):
models, they havethese context windows.
So I actually have some whichhave been running for, like,
six months.
So and I've been kind oftraining it every time
I work on somethingthat's very close to it.
So now, there are somethings that are like, hey.
I want to start workingon this and like,
can you copy what we did lasttime, make these few changes?
Then I have probably 200lines that are, like, written.

(34:26):
Some of it is veryboilerplate, but it's still,
I would have normallyhad to do it myself.
But then all I now dois what Vaneese said.
I just want tostart reviewing it.
Then I can iteratewith it and say, hey.
No, make this few changes.
And you start havingthat collaboration.
So I actually do see alot of that happening.
And I think as thetools become better,

(34:47):
it's just going tohelp me do more things,
giving me time back.
It's going to be hard.

JILL FINLAYSON (34:53):
So if you're both individually saving
a lot of time, why mightcompanies not be seeing it,
not rely on AI atthis point in time?
Is there something thatcompanies are not doing
right in terms of using AI?
What are you thinking, Vaneese?
I see you nodding.

VANEESE JOHNSON: What I'm thinking is, (35:08):
undefined
the components thatscare corporations
are policies, ethics.
Because right now, withoutthose guardrails in place,
a company is like, hands off.
We don't know wherethis is going to go.
We don't know how peopleare going to respond to it.
We don't have a way torespond to issues, challenges.

(35:29):
We're afraid of lawsuits.
So I think the bestthing that a company
can do is to haveits leaders first,
to really startto become educated
on all of the differentimportant facets of how
AI would impact a business.
I mentioned in ourjust pre-workshop
that I recently took a courseat Haas School of Business.

(35:49):
And it was AI for Executives.
And it was an intensive coursefor three days where there
were 60 of us in the room.
And we all got a chanceto really collaborate
as thought leaders aroundwhat we were learning.
We saw the valueof the direction
of the future of how AI isgoing to impact the workforce.

(36:10):
But we also saw thevalue as leaders.
How can we start toembrace this tool?
And what areimportant components
that will support us ingoing back and pitching it
to our boards, pitchingit back to our teams.
And I think one ofthe biggest things
that I can say that welearned was that, you
can do this in small stages.
You don't have to do acomplete whole rollout

(36:33):
in a short amount oftime because the learning
curve can be so steep.
Also, what we learned is thatwhen some companies didn't
employ really a rolloutstrategy, there were penalties.
And I think companies arekind of more hampering
towards the penaltyside versus looking
at the value side of that.
And then the otherpart that I'll add

(36:54):
is that there arecase studies out there
already where companiesare doing this efficiently.
There are municipalorganizations
that are rolling outAI that we-- of course,
we've got the tech side, thenwe've got the consumer goods
side.
So there are someindustries that
actually are rolling this outand doing this successfully

(37:14):
in different increments.
So corporations andleaders need to ask
which pieces ofthese components can
we adopt that are in alignmentwith how we run our business,
and alignment of howwe serve our customers,
and ultimately inalignment with how we
want to enhance our workforce.

JILL FINLAYSON (37:30):
Yeah, Chalenge, obviously, we're
also talking about alot of data and privacy.
And there's a differencebetween, obviously,
a paid version of ChatGPT and anenterprise version of ChatGPT.
What are you seeingthat companies
need to do to setthings up correctly?

CHALENGE MASEKERA (37:46):
Like any sort of technology,
or partner, or anyperson you collaborate
with, it's generally,it's a question
of how well youuse the technology.
This is, like, whatthe technology is.
Especially if it'scoming from the top,
we're now having theseAI evangelists who
are coming to organizations.
They don't have tobe tech companies,

(38:07):
but trying to findthose niches, like for
our specific organization, whatare those things that actually
we can adopt these technologiesto make our organization
to be much more effective.
And I think that'swhere it starts
is like, before you sort of,like, let's say, all right,
let's all startusing Copilot, you
need to take a stepback and say, OK,

(38:28):
what exactly do wewant to get from it?
What are the sort oflow-barrier entries
that we can actually adopt thistechnology and be successful?
Then the secondpart is it always
matters the quality of theinformation that you have.
So whether you're aperson, your experiences,
what books you've read,what you've learned.
And it's more truefor sort of AI

(38:49):
with these largelanguage models.
I think inasmuch asI'm a super optimist,
I think there's still lotsto be done for them in terms
of general quality.
So to say we're going toreplace all our knowledge
workers in the shortterm, yeah, that's
completely out of the question.
So just sort of to tie thetwo points is the quality
of the model mostly matters.

(39:10):
So you have to figure out whichones are very good at adopting,
or which ones work well forthe specific use case you
want to have for your company.
ChatGPT may be good forthis, but now there's
no Gemini from Google.
It's very good at other things.
At least for me, forprogramming, we have Anthropic.
It's better thanChatGPT for that.

(39:31):
So I think having thatlayered way of thinking of,
like, this is what wewant, what are the skills
that our people have?
And what are the problemswe're trying to have?
And then having sortof this exploratory
is like, I think wecan start with this.
Then building ontothat, and just
exploring the toolsthat are there.

JILL FINLAYSON (39:47):
I really like that layered way of thinking,
and that there aredifferent tools that
are better for different needs.
So really understandingwhere this will be adding
value as opposed tobeing a distraction.

VANEESE JOHNSON (39:59):
Jill, I want to add to that that companies also
can build their own AI.
That's really a valueadd that I don't
think a lot of corporationsare really aware, or having
the conversationsaround, because there's
a concern about privacy,intellectual capital, trademark
information.

(40:20):
So that's a hugearea that I think
it's important to addressin this conversation
that corporations,they have the power
to design AI program that reallycaters to that specific company,
and that businessline of that company,
and be able to putthose guardrails.
So creating jobs where someoneactually is monitoring AI tool,

(40:43):
I think that's anotherimportant component.
And when I say monitoring, notjust somebody that's sitting
there from 8:00 to5:00 looking at it,
but this needs tobe a 24/7 operation,
where the personis monitoring that.
Also, they have to--
looking at ethics and biases,depending on what is being used,
or how this informationis being used,
and what you're plugginginto that information.

(41:05):
Health care isshifting for all of us.
And there's a lot ofAI in health care.
The good news isthat it can make
a patient visit morepersonalized because it
learns you.
The other side is, how do youprotect the data of that patient
that you are now providing thatcustomized personal care to?
So I do think that companiesget to have the conversation

(41:27):
with people like Chalenge andothers in his field in terms of,
how do we build amodel that really works
for our businessand our employees
so that when we are giving themsome level of freedom and level
of autonomy, that they canfeel comfortable innovating
and being creative?
But we as the company, canalso have a level of comfort

(41:49):
that we have someguardrails in place.

CHALENGE MASEKERA: Yeah, I think it's (41:52):
undefined
one of the sort oftough conversations,
especially in tech.
Because as a tech person, Ithink there's always like,
let's move fast all the time.
And then, deal withthe consequences
later, which is funbeing part of it.
But I think there'salways, we've
just had too manyinstances of where

(42:12):
this technology has actuallyhad really, really bad outcomes.
And at least from what I see,I think for the most part,
I think anybodybuilding an AI tool,
I think that hat is now there.
We've gone from the wilddays of the internet
where it was just like, let'sbuild it out and throw it
to people.
Now, there is some focus.

(42:32):
I wouldn't say it's therewhere it needs to be.
But I think havingthat mental model of,
this is what we wantfor society and, like,
what are the ethical and safetyimplications of our models
is becoming forefront.
I think the beautyof technology,
especially with AI and all thesenew things, they can reach--
they move fast, but alsoif they're uncontrolled,

(42:55):
they just also spread harm,and at a much faster rate
than anything we've seen.
So I think it's very important.
If you are usingAI to figure out
what are the ethicalimplications,
and also if you're a businessthat's adopting it, actually
making sure this is what--these are our priorities.
And like, will this AI serve usto the fullest of our abilities,

(43:15):
and not put us in trouble?

JILL FINLAYSON: Yeah, making sure (43:16):
undefined
the purpose itself isethical, and responsible,
and then having accountabilitythat I don't think
any company sets outto do the wrong thing.
But if you aren't monitoring,to Vaneese's point,
you could be causing harm.
And I think there is agreater push on accountability
for the companies thatare producing these.

VANEESE JOHNSON (43:37):
Yeah.
Yeah, you cannot implementa program, an AI program,
institute that inan organization,
without havingdiscussions of ethics.
You can't have that security.
They have to be at thetable of the conversation.

JILL FINLAYSON (43:50):
Well, can I turn to a more mundane question?
With AI autocorrecting,finishing our sentences,
predicting whatwe're going to say,
I worry about thesea of sameness,
everything starting to be verygeneric and one size fits all.
And I don't think itone size fits all.
So what are some of thedownsides to using AI?
And how do we avoid them?

VANEESE JOHNSON (44:12):
I'll say my experience on this.
And I know Chalenge has gotthe engineering side of this.
My experience on this is themore that I have used AI tools,
the better it learns mypersonality, and who I am.
So whenever I'mdoing prompts for--
I was working on my bio.
I was fine tuning mybio for a speaking event

(44:32):
that I was speaking at.
And I put in what I had paidsomeone to do for me last year,
really great writer.
And I asked Chat based onwhat it has experienced
in me over the lastyear, update my bio,
but make sure you'readding my tone, my brand
tone, and my brand personality.
And so it's givingme information back
that really reflectswords I do say,

(44:56):
the energy of the context inwhich that information is there.
And sometimes it's like,wow, that was really good.
You really know me.
So for me, the more thatI'm using these tools,
the more that I feel a level ofokayness with the output of it.
The other thingthat I'll say is,
AI is showing up in other areasof our lives, on our phone.

(45:17):
AI will tell you, hey, Vaneese.
You normally call Jillon this day at this time.
Would you like to call Jill?
Sometimes, that mightbe a bit overwhelming.
Like, wait a minute.
How did you know I call Jill?
Well, I'm using the computer.
And so other times, it mightbe like, oh my goodness.
Thank you for remindingme to call Jill.
I get in my car.
My phone tells me how long it'sgoing to take me to go home,

(45:39):
to get to my house.
And so there are aspects whereAI really does bring some value,
but I think it's importantif we're using these tools,
and we're in thatcontinual learning mode,
that we are training these toolsas much as possible about who
we are as the individual.
And when you put inthose prompts in there

(46:01):
to really stay away fromtrying to ask questions,
to be like somebody else, toreally approach it and use it
from a place of personalidentity or personal information
versus, again, tryingto be like somebody else
and trying to come up with,being Beyonce's formula.
Give me Beyonce's formula.
I don't want Beyonce'sformula because I
can't deliver at thatlevel that Beyonce does.

CHALENGE MASEKERA (46:21):
I totally agree with Vaneese.
Right now, where we are,like, if you look at AI tools
that we have spokenabout, like ChatGPT,
Gemini, whateveryou can think of,
there are these one superlarge language model that's
OpenAI hosting their system.
But even with that,we're starting
to have these contextswhere you talk to it, like,

(46:43):
oh, you're asking it.
It starts to get you.
It's still this model, butwe're going to get to a point
where these are all goingto become commodities.
If you look at your phonewith an iPhone or whatever,
if you look at how you type,I type, I'm from Zimbabwe.
So we have another language.
So when I talk to myfriends back home,

(47:04):
I use like whatwe call Shonglish,
which is a combinationof Shona and English.
So some words are halfEnglish, half Shona.
The first time I typeit, it'll autocorrect me.
It's like, what areyou trying to say?
I keep doing it, andafter a while, actually,
if I misspell thoseShonglish words,
it will actually correct them.

(47:25):
We're going to get to apoint where everybody's
going to have their own personalAI, which actually understands
them pretty well,either on your iPhone
or whatever the next generationof the technology that we're
going to be interacting within your pocket, or whatever,
that understands you andknows as much as you,
or even sometimes[INAUDIBLE] better than you
that we're going to use.

(47:45):
But however, I thinkwhen you interact
with it for thefirst time, yeah,
there is that sortof lack of novelty
that you find when youinteract with somebody.
If I talk to Vaneesefor the first time,
I'm going to getVaneese's personality.
But if you talkto all these AIs,
if I just ask for something,like especially, sometimes I

(48:08):
play this game where Ijust, somebody text me,
then I go to ChatGPT andsay, can you respond to this?
And I'm pretty sure I'm goingto know what it's going to say.
We're going to--initially, there's
going to be some loss of it.
But I think interactwith it a lot more.
It's going to get morecustomized, and then more
personal.

JILL FINLAYSON (48:24):
I like that.
And I like the fact that itwill get better at helping you,
because it'll be doing itmore like you as a person.
But I also put my shoe onthe other foot, as it were,
and say, OK, peopleare applying for jobs.
They're applying forawards, or grants,
and they're nolonger writing those.
They're writing the first draft,they're instructing the AI,

(48:46):
but the AI ischurning out an essay.
How does this affectthe people who
are trying to choose thepeople to let into a program,
or into a college,or into an award?
How did they assessthis program that
has really notnecessarily been written
by the individual themselves?

VANEESE JOHNSON (49:05):
Yeah, I think this is an important question,
because I know that the redflags were raised, especially
in the academia environmentabout people submitting papers
that are not reallytheir papers.
And we also heard that infamouscase where the lawyer was there
doing an argument and hepresented an argument,
but it wasn't really real.
It was AI generated.

(49:27):
But I think what's stillimportant in that component is
the human-centric connection.
I still think there's anopportunity for whomever
the person is submittingany type of documents
to, especially when itcomes to something that's
kind of personalized in a way,I'll speak on both fronts,
career and business, isto interview that person.
Because you're going to beable to tell right away.

(49:48):
Whenever someonegives you a resume,
and the resume looksamazing, and you
start interviewing thatperson, you have two options.
Either you realize theperson wrote this content,
or you realize the persondid not write this content.
So I think thosecomponents still
will be an importantfactor that whomever
is on the receivingend, that they
get to take it a step further.
As a teacher andinstructor myself,

(50:09):
I know the personality ofthe students that I teach.
So I can tell.
And I recently did this.
I introduced ChatGPT to a groupof women entrepreneurs that I
was training in a 10-weekbusiness planning class.
And when they wouldsubmit their homework,
Jill, I would givethem feedback.
I was like, thisis a good start,
but this is Chatpersonality all right here.
What I need you to do isI need you to rewrite it

(50:29):
in your personality.
Because I use the tool, I'm ableto recognize when something just
looks a little toopolished, given
the fact of where it came from.
On the other side,when a business
is submitting forgrants, because there
are AI tools out there that willreview proposals and help you
to write a proposal, you doneed to read that proposal
and make sure thatyou're putting

(50:50):
the human side,the relatable side,
of the people that willbe delivering the project,
of the organizationand the work that it
has done, in the field thatthey are responding to the grant
with.
You still have to bringthat human-centric component
into that.
And if those people readingit on the other end,
they know their field.
They've seen hundreds ofthousands of proposals.

(51:12):
They should be able to,and may be able to detect,
when something's a little off.
But at the end of the day,the human-centric component
and connectivity still needsto be a part of that process.

CHALENGE MASEKERA (51:24):
When you said it feels a little too polished,
you're being generousto all these AI tools.
They're mostly too robotic,at least in my experience.
And I think if you have usedany of these, especially
these generative tools, theyall kind of sound the same.
And you can just tell that thisis not written by a person.

(51:44):
And I think if you're goingto submit that and say,
this is my work, anybody whohas actually played with it
will, a couple sentencesin will be like,
this is not written by a human.
At this point,it's easy to tell.
They'll get betterat some point.
But I think, again,like, I think
as somebody who's been reviewingstuff, like for a long time,

(52:07):
you can tell what'shuman and what's not.
Your client has apresentation, use the tools,
the generative AI, tocreate the content.
But then what actuallyputs you over the hump
is your personality,your experiences,
which none of thesegenerative models have.
What makes us special orbetter than other people, sort
of different, is our individualexperiences, the way we talk,

(52:29):
the way we sort of craft--
Well, it may be at some point,but that human element is
what you're going tolose when you just
submit AI-written stuff.

JILL FINLAYSON: Yeah, so you have (52:39):
undefined
to give it this specificstory, the personal story,
and say, incorporate thispersonal story into my essay,
into my answer.

VANEESE JOHNSON (52:48):
I'm applying for TEDx stages.
And so you have tosubmit the application,
you have to submitshort form bio,
you've got to submit asynopsis of what your talk is
without it being the talk.
Then you have to fillout the application,
and you have to make surethat things aren't repetitive.
So there were timesmy coach and I,
my coach is acommunications professor.

(53:09):
So there were times when shewent in, and I sent her this,
and then she tweeted with AI,and then gave it back to me.
And I'm like, thatdoesn't sound like me.
This line here, let mechange this line here.
But 80% of it sounds like me.
And so being able to do that.
But at the end, I also hadto submit a 60-second video.
So not only am Isubmitting you paperwork,
but I'm also showing youmy personality within that

(53:33):
60-second window.
And what I asked Chat todo was to narrow down,
give me so many words.
And I had to narrow downthe time to 60 seconds.
Chat couldn't narrow it downto 60 seconds on the dot,
because I needed tobring the human part
to be able to do that.
And then as I trial and error,I was able to cut out words,
and then give it prompts tohelp me to really negotiate

(53:56):
what I was saying, andwhat needed to come out.
So those things arereally important.
When you look atusing them in tandem,
I felt like astronger candidate.
I was very proud of myself thatI was able to create now a model
that I can go and applywith other stages.
But by no means do I want Chatto be 100% all of my data,

(54:18):
because I'm going to show up.
And they're going to be like,you're the boldness coach?
I don't see any ofthat on the paper here.
You are not bold, lady.
Next.

JILL FINLAYSON (54:25):
Well, I like the personality side.
It kind of ties back towhat you said earlier
about the soft skills,or essential skills,
that you need to develop.
And I was thinkingabout the fact
that just workingwith ChatGPT means you
have to ask better questions.
You have to docritical thinking.
You have to analyzewhat you get back.
What are some ofthe other skills
that you think you need todevelop to use AI effectively?

(54:49):
Chalenge, what haveyou seen that has
been helpful in adopting thesetechnologies, but also skills
you had to develop?

CHALENGE MASEKERA (54:58):
I think it sort of, like,
ties all the skills thatI've learned before.
I wouldn't say, like, I'verediscovered how I think.
One of the things thatI sort of had to do back
is, always take beginner'smindset as in, like, start from,
OK, this is somethingthat I haven't adopted
and I've never used before.
How do I now try and incorporateit more into how I work?

(55:20):
So I started fromwhere I was like, oh, I
need help for, like,troubleshooting,
to I actually have noidea what I want to do.
And maybe, just let'sbrainstorm together.
I just have thisrandom thoughts.
I'm trying to do this.
Then I start from there.
I wouldn't say there's more ofnew skills that I had to adopt.
I think it's just trying to putall the skills that you had,

(55:42):
just like adoptingit and treating it
like a beginner's mindset.

JILL FINLAYSON (55:44):
You know that famous garbage in, garbage out?
If you put bad information intoAI, you get bad information out.
There's a new websitecalled gozigzag,
which allows you to create abusiness plan, Lean Canvas,
over two seconds.
And the only thingit asks you for
is a prompt to describethe problem and solution
that you're working on.
But depending on how goodthat little prompt is,

(56:07):
you're going to get a muchbetter or worse Lean Canvas
model from it.
Vaneese, what haveyou seen that has
changed how you think about AI?

VANEESE JOHNSON: You said it earlier, (56:16):
undefined
Jill, at the top of thisquestion, is critical thinking.
It's really challenged me tothink and ask better questions.
And you can tell ifyour question is not
as deep as it could bebased on the output of it.
And you're like, no, no, no.
This is what I mean.
So that critical thinkingcomponent is really important.

(56:36):
It's also helped meto enhance my data
literacy in termsof interpreting
the data into the story.
What is this datareally telling me?
Do I like the story?
Do I need more data tocreate a different narrative?
Do I need to havemultiple narratives?
And so it's reallychallenging me to learn that.
The other component isthat AI is challenging me,

(56:57):
we mentioned earlierabout lifelong learning,
is to really lookat micro learning.
A lot of my clients now, they'reasking for micro training.
So I've done a whole librarynow on micro learning
is because people reallynow are looking at,
how do I learn somethingquicker, sooner, faster,
and then go applyit, and then make
the adjustments along the way?

JILL FINLAYSON (57:16):
I think one of the challenges
is that quicker, better,faster that some people, AI can
be perceived as the lazy, orthe shortcut, or the quick way
to get to an answer.
But in fact, you have to bevery critical of what comes out.
You have to change it.
You have to reask your question.
And so my concern is people whotake the first answer that they

(57:37):
get and don't do the evolution.

VANEESE JOHNSON: I don't recommend (57:40):
undefined
doing the first answer.
I always, and I practicedthis with my students.
And I always tell them,that's a good start.
But let's really look andsee, what did it give you?
Because a lot of times,what I was experiencing
was, they would say, it's notgiving me a better answer.
And it's like, OK,so what that means
is we need to askbetter questions?
But let's reallythink about, what

(58:02):
is it that you reallyneed to know in order
to move forward to a next step?
And sometimes, I don'tknow what that is.
OK, well, let's startat the impact, then.
What is the impact thatyou're expecting this to do?
And then let'swork back that way.
What I found was that developingthose critical thinking skills
and approaching it fromdifferent perspectives

(58:24):
can help you to form betterquestions so that now, you
can be able to weigh theoutput that you are receiving
as to what's better.
And then what Idiscovered, too, is
that when people askbetter questions,
and we're more applyingmore critical thinking,
they felt better about using AI.
The next time, they'relike, OK, so now
I know I need to aska better question.

(58:46):
So now, they starttraining the model.

CHALENGE MASEKERA: You're absolutely (58:47):
undefined
in terms of critical thinking.
Previously, sort ofthe evolution of this
is like, we used tohave Google, then
you have, like, seven links.
Then you go througheach one of them
and try and find the answer.
But now, you have thisvery specific answer,
which if you can'tprocess the information,
analyze, and thinkcritically, you never

(59:09):
know if it's the right answer.
If you're just goingto take it as is,
it's not going to besuper useful for you.
You're going to actuallybe led on the wrong path.
The fun exercise todo sometimes is, like,
ask any of these AIagents, like, hey.
What do you think about this?
Or like, can you giveme an answer for this?
And then you tellit, it's wrong.
It'll confirm to youthat it was wrong.

JILL FINLAYSON (59:27):
I had a really interesting experience
with AI trying to askit to create an image.
I just wanted animage of a person
being interviewed by a robot.
And of course, the firstimage was a white male.
And I'm like, no, no.
I would like to havethis be a woman of color.
And it gives me thisglamorous person.
I'm like, no, a normal.
And the AI is responding backto me, oh, I see what you mean.

(59:49):
Oh, you want this person tobe professionally dressed.
Oh, I see.
And the odd thingwas that the robot
became feminine when I askedfor a woman interviewee.
And I'm like, why didthe robot become a woman?
And the AI waslike, oh, I'm sorry.
Here's the one witha generic robot.
So there's all ofthese things that
happen that if you'renot asking questions,
you could be getting verystereotypical, or biased,

(01:00:13):
or incorrect information.

VANEESE JOHNSON (01:00:15):
Yeah.
Yeah.
And I like thefact, Jill, that you
kept giving it moreprompts in terms of,
and you kept fine tuningit, and narrowing it down.
I think that's really important,especially from a new user,
is to understand that youneed to train the model
to help you to get towhere you're ultimately
trying to go, and not tosit back and say, see?

(01:00:37):
This is biased.
I don't like using that.
That's why I don'tuse AI, because it'll
give you these kind of answers.
Well, let's reallylook at, how are
we applying criticalthinking to the prompts
that we're putting into there?
So I'm just glad that you sharedthat example so that people
know that, don't stop.
Don't give up on it.
Continually prime it.
Prompt it correctlywith critical thinking

(01:00:58):
so it can give you backmore closely-related to what
your expectations are.

JILL FINLAYSON: And Chalenge, did (01:01:02):
undefined
you see we could tell it it waswrong and give it corrections?

CHALENGE MASEKERA (01:01:06):
Yes, you can give it a correct wrong answer
and it will agree with you.

VANEESE JOHNSON (01:01:11):
I love that.

CHALENGE MASEKERA (01:01:12):
That's how sophisticated the technology is.

JILL FINLAYSON (01:01:15):
Amazing.
Well, as we thinkabout now the future,
so we're coming up to2025, what do you think
is going to happen in 2025?
We saw a lot moreadoption in 2024.
How is 2025 goingto be different?
Chalenge?

CHALENGE MASEKERA (01:01:30):
I'm not so sure, actually.
I've thought about this.
I think we're at a stage where--there are some people who
argue that we've sortof reached the limit,
or we're close to thelimit, of this phase of AI,
like these new languagelearning models.
So I'm not so sure how muchwe're going to progress.
But what I think I'm going tosee, I think over the past year,

(01:01:51):
I've actually beenseeing people adopt AI,
and in very specific use cases.
And those use cases havebecome really well polished.
We're in the 50, 60% correct.
We're going to getto maybe the 70s.
So I think we're going toget more tailored use cases,
and of how people usetechnologies in their roles.
I think it's going to getbetter at the things that it's

(01:02:14):
currently good at, thingsthat are generating text,
things like generatingvideos, helping you polish,
come up with plans.
For me, what I seeis the biggest use
case is having these automatedAI-powered human workers,
specifically for industrieslike service, where you're like,
you just want to say, Ididn't get my Amazon package.

(01:02:35):
Instead of havinghumans in the loop,
I think we're going tosee a lot of more software
companies adoptingAI-powered agents for that.

VANEESE JOHNSON (01:02:42):
Yeah, I agree in that too.
We're seeing that now, Ithink what I'm seeing just
from the consumer endis that the chat bots
and some of these companiesare more sophisticated.
Before, the chatbotwould give you
some real kind of simple stuff.
Now the chatbot asks you reallyspecific information like,
what's your account number?
What's your address?
So it collects the data of whoyou are, and the interaction

(01:03:05):
with that company,with that brand.
And it's able to tailormore towards responding
to you around a specific thingversus something general.
So I'm seeing that now.
I think also, what we'regoing to see in 2025
are more technologicaltools that
have AI embedded in its use.
I mean, Apple has the new phoneout that has AI in the phone.

(01:03:27):
So I think consumersnow are going
to find themselves,professionals and that consumer
spot, are going to findthemselves interacting and using
these tools more from athoughtless perspective
so that they can start todevelop levels of comfortability
in using the tool.
I also think thatthose individuals
in the professionalsectors who want

(01:03:48):
to advance, I think they'regoing to find themselves really
taking it upon themselvesto play with these tools
when it comes to careercoaching, when it comes
to resume writing, when itcomes to interview prep,
when it comes to researchon target clients
that they want to work with.
I see people nowstarting to say, hey.
It's been around for a while.

(01:04:09):
They were right.
It's not going anywhere.
So I'm going toreally look at how
I can start kind of taking bitesizes out of this technology,
and start to makeit work for me.

JILL FINLAYSON: Yeah, I think you (01:04:19):
undefined
both hit on a couple of things,these really tailored use cases
where things will be optimizedfor particular uses, as well
as the integration.
It'll just be part ofthe tools, like we've
seen with Zoom transcriptionor these other functions
where it's just partof how you do business.
And we will seeit less and less,

(01:04:40):
but it'll be theremore and more.
So as we thinkabout this new world
where it's going to be much moreembedded in how we do our work,
what are your finalwords of advice to people
as we enter this new year?
And what can they do to prepare,and take advantage of it,
and really grow?

VANEESE JOHNSON (01:04:56):
Start really recognizing
how you're using itin your everyday life.
That's number one.
You're already using itin your everyday life.
Become comfortablewith it, because when
you're using itfrom your home, it
doesn't have the same pressure.
When you are programming yournew TV, when you are programming
your new phone thatyou're going to buy,
when you're buyingthat new vehicle
with all thetechnology in it, start

(01:05:17):
using it on a personal level.
So that way, you're notfeeling the pressure of it.
The other part that I'll sayfrom a professional development
side, I think it's reallyimportant for the individual
to take ownershipof your career,
really start looking atways to reskill or upskill
what you have.
Reskill is you'relearning a new skill that
may be in alignmentwith the next direction

(01:05:38):
that your career is going.
Upskilling may be, howdo you do the current job
more efficiently?
So start takingownership of that,
and don't waitfor a major change
to happen in theorganization where
you are forced to adoptto using this tool,
but you can better kindof manage and assimilate
with the change, because itis coming and it's happening.

(01:05:59):
And it's not theemployer's responsibility
to make sure that yourskills are up to par.
You have to own that andtreat it like a business.

JILL FINLAYSON (01:06:07):
So start using ChatGPT for your gift guide
recommendations for yourtrip, travel, whatever you
need to do for the holidays?

VANEESE JOHNSON (01:06:14):
Exactly.
Because I think whenyou play with it,
Jill, you'll start tosay, wait a minute.
It's giving me giftguide and vacation?
Well, let me see whatit can do for my career.
What's the besttime to take off?
What's the best way tosave money for the trip?
So I think it'llbreadcrumb you to where
you will create more curiosity.
And you'll find yourselfusing it for information
that you didn't evenplan on using it for.

(01:06:37):
And I think you can discoverthe value in that way.

JILL FINLAYSON (01:06:39):
And what breadcrumbs
are you following, Chalenge?
What advice do youhave for folks?

CHALENGE MASEKERA (01:06:44):
Yeah, I think I would just, like, add
a few more breadcrumbs there.
But I think I reallylove what you said.
Your careers should betreated as a business.
And I think every business,what do businesses do?
They always find waysto be more efficient,
how to further improvewhat they're doing.
And I think we havegone away from,
AI is going to kill us, AIis going to replace our jobs,

(01:07:05):
with oh, AI is here.
We just, it's going toembed in everything we do.
So just finding thoselittle tidbits of,
how can I be better at somethingthat I either don't like
or even somethingthat I'm good at,
like, how can I be incrementallybetter by using AI,
is the way to go?

# JILL FINLAYSON (01:07:21):
So hashtag #bringbackthefun,
and use these thingsto innovate, and make
your job better.
Make your job better in 2025.

VANEESE JOHNSON (01:07:29):
Right.
# And hashtag#treatyourcareerlikeabusiness.

CHALENGE MASEKERA (01:07:32):
I like that.
I love that.

JILL FINLAYSON (01:07:34):
And with that, I want to thank you
both for joining me today.
It's been a wonderfulconversation.
And I hope thateveryone listening
has enjoyed this latest ina long series of podcasts
that we'll be sendingyour way every month.
Please share withfriends and colleagues
who may be interested in takingthis Future of Work journey
with us.
And make sure to check outextension.berkeley.edu to find

(01:07:56):
a variety of coursesto help you thrive
in this new working landscape.
And to see what's comingup at EDGE in Tech,
go ahead and visitedge.berkeley.edu.
Thanks so much for listening.
And we'll be back next monthto talk about neurodiversity
in the workplace.
The Future of Work podcastis hosted by Jill Finlayson,
produced by Sarah Benzuly,and edited by Matt DiPietro.
[MUSIC PLAYING]

(01:08:19):