Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_02 (00:04):
You're listening to
the audit presented by IT Audit
Labs.
I'm your co-host and producer,Joshua Schmidt.
We have the usual suspect, NickMellum from IT Audit Labs.
And today our guest is AlexBratton.
Alex is an applied technologistand AI implementation expert.
He's here to share some AIadvancements that are cutting
edge and then stick around tothe end.
We're going to have some tips onhow to implement AI into your
(00:25):
organization in a securefashion.
So without further ado, Alex,thanks for joining us today.
SPEAKER_01 (00:30):
Thanks.
It is awesome to be here and uhlove having any opportunity to
geek out with folks, but make itapplied geekery of how does this
actually drive real businessresults.
So thanks.
SPEAKER_02 (00:40):
I love your chotch
keys in the back.
We're chotch keys here.
Yeah, you can you can tell a lotabout a guy by his by his
collection.
Uh as you can see, I'm a musicguy and producer.
So what do you what do you do inyour spare time when you're not
when you're not geeking out?
SPEAKER_01 (00:54):
Uh well, of course,
as you can tell, equal
opportunity, sci-fi, andsuperhero.
So all things that are lots offun.
Um gaming, and at the same time,uh, sure I run a company, but at
the same time, I'm hands-ontesting the latest AI, diving
in, writing code, playing someinteresting dungeon delve board
games, you know, any of that funstuff.
SPEAKER_02 (01:12):
Uh it's too bad
you're not in the Midwest here.
We have a we have a really sweetuh game night every Wednesday or
first Wednesday of every month.
So very cool.
We'll have to talk more.
SPEAKER_03 (01:21):
I have to touch back
on the Totchkis, though, because
my wife makes fun of me all thetime.
Because if we go anywhere, I'malways looking for Totchkies,
little things, you know what Imean?
Yeah, exactly.
My wife always makes fun of mefor it.
So I'm to have another Tchotchkeenthusiast with Josh and I here
tonight.
SPEAKER_01 (01:40):
And we're kind of
already there.
I'm trying to figure out wherethe Galactus popcorn cup has to
go.
Because I I don't have it yet,but it's gotta go up there.
It's gotta go up.
SPEAKER_02 (01:51):
Um so um we got into
your hobbies there for a minute,
but I'd love to hearprofessionally what you've been
up to.
You have quite an impressivelist of credentials on your
LinkedIn.
I'd love to have uh you explainwhat you've been working on and
tell us what's top of mind.
SPEAKER_01 (02:04):
Uh sure.
My my personal why is all abouttechnology as supposed to help
people thrive at work.
So all of the many companiesthat I founded have all been
about that.
Uh 25 years ago, I launchedLexTech.
It really focused on, hey, howdo we help build the tools
people use to do their jobs todrive great business results?
Um and in the starting days,I'll say the the early days, so
(02:26):
before 2008, that was embeddedtech, web stuff, um, you know,
controlling government systems,all kinds of fun stuff.
But in 2008, the magic of theiPhone hit, and that was
suddenly, you know what?
Wherever people are going totouch technology, that's where
we needed to be.
So we pivoted the whole companyto be mobile.
And that was actually also thelaunch of our partnership with
Apple.
And we've been working withApple's enterprise team for 15
(02:48):
years, um, especially now in theuh the the age of the Vision
Pro, to help these big companiesfigure out what do we do with
this stuff?
How do we empower people?
And that and that is the key forme.
It's all about how do we givepeople superpowers?
And that's something that we thegeeks get to do now, and I love
that.
SPEAKER_03 (03:05):
We know the right
guys.
So if we need some test AppleVision Pros, we're just gonna
call Alex to get a couple sets.
SPEAKER_02 (03:12):
Send them on over.
SPEAKER_03 (03:12):
Alex, I have to just
tell you because when Josh said
you were gonna come on with theApple Vision Pro, I brought up
whenever these came out a yearor two ago, I was telling Eric
that we need to get three setsof these and we need to do a
podcast in Apple Vision Pro.
Obviously, it didn't happen, uh,so I'm very jealous that you're
doing it.
SPEAKER_01 (03:31):
Well, you'll still
need to push that because that
is exactly what I have somethingthat is spatial conversations.
It is a first spatial podcastwhere, you know, yes, you have
virtual me here, but when we canconnect via FaceTime uh Vision
Pro to Vision Pro and bring upto five people into a virtual
space and capture it where, youknow, here we're still in
Flatland, but when we canactually be sitting next to each
(03:54):
other and fist bump from athousand miles away, that
changes everything.
And it changes everything forbusiness, but for podcasts, it's
just incredible.
SPEAKER_03 (04:01):
Yeah.
I gotta stay on the Apple VisionPro for one more second because
I am a fellow geek and lovingthis stuff.
Have you been on an airplanewith those on?
SPEAKER_01 (04:10):
Oh yeah.
Oh yeah, really?
Oh yeah, because I anddefinitely it definitely starts
conversations.
Now, one of the interestingthings, and I don't know that
you can see it on the real worldcamera here, but uh you if
you're in real person, you canactually see my eyes here so
that it you know it feels likeI'm here.
So if a uh you know flightattendant is walking by, I can
look up and they can see thatI'm looking at them, and it's a
(04:32):
little less freaky.
Uh but lots of greatconversations.
But as a as a business traveler,you know, back to the security
side of things, this is the mostsecure way in existence for me
to work on high confidentialitydocuments on an aircraft.
I can have my laptop up here,and the screen is black, and I'm
looking at a virtual Mac screenin front of me, working on
(04:52):
agreements, presentations,whatever it is.
It's awesome.
SPEAKER_03 (04:55):
Yeah, that is very
cool.
I when I had a pair when theyfirst came out, I wanted to get
on an airplane.
It just was too weird for me atthat time.
But now I think I'd be ready todo it, so I think I need to give
it another shot.
SPEAKER_01 (05:08):
It it's much more
common now.
And uh and I've never hadanybody be weird about it.
Lots of curiosity, and that'scool.
You get to have greatconversations with folks.
And I've been on multipleflights where I wasn't the only
person.
SPEAKER_03 (05:21):
I think too, the
curiosity is awesome because I
think, especially in our space,most of us are very passionate
about cybersecurity.
And if somebody asks you aboutit, generally you we want to
have the conversation with youabout it.
That's just part of the trade,right?
We want to educate people.
So this is kind of the samething, right?
If you think it's cool, Ialready think it's cool.
It's you know, we obviouslyalign, let's let's have a chat
(05:42):
about it.
SPEAKER_01 (05:43):
Yeah, and there's
there's so much we can talk
about.
And again, this is this is theiPhone 1 moment for the world.
That came out, and you know, 5%of us were all over ourselves
and waited in line for 12 hoursto get them.
And the rest of the world waslike, who are these crazy
people?
Fast forward to the iPhone 5,and it really changed everything
of how we work.
Sure.
But the that's the part that Ireally give Apple a lot of
(06:05):
credit on with this, is thatthis is the first time ever that
they have telegraphed this iswhat the future is going to be
in five to ten years.
And they started in theenterprise because of some just
great enterprise and work usecases for it, but it's the kind
of thing where if folk, if youhaven't done it, get to an Apple
store and get the demo.
This is so different than the VRheadsets out there.
Spatial computing is verydifferent.
(06:27):
It's it's bringing everythingthat might normally be on my
computer, and I've I've gotwindows floating around me here
where I can see what I need tosee.
It changes everything.
So in five years, you know, whatit's not gonna look like this.
Is it glasses?
Is it something else?
But this is the future the sameway the mouse was when Apple
brought it out.
SPEAKER_02 (06:44):
You mentioned that
uh you're you're kind of testing
some of these things out.
Have you seen any new workflowscome about or have you found any
specific tools that will blowpeople's mind that that you've
been using in their spatial uhwork life?
SPEAKER_01 (06:56):
Uh definitely.
Uh and it I'll say for us, itreally comes down to just the
collaboration and being able tobring a small team together and
uh put content together.
So pair programming in this isinsane.
You know, doing that in person,you're kind of stumbling who's
got the keyboard, who's on topof each other.
Virtually two of us sitting nextto each other looking at that
10-foot screen of one persontyping while I've still got my
(07:18):
computer doing it.
So collaborative engineering,collaborative design, awesome.
Uh, from a business perspective,um, we had actually put together
some originally it started asdemoware of hey, what if we
could bring some people in andthere's a lot of companies out
there where for a high valueproposition sale, it's a
million-dollar sale or a tenmillion dollar sale, I'm selling
(07:39):
you an aircraft, I'm selling youa piece of construction
equipment.
You have to travel to go see it.
And what if instead if we'reapart, we can actually be
sitting next to each other and Ican put the bulldozer right
there and we can get a feel forit.
Or we can actually be out onyour manufacturing floor and
(07:59):
take the piece of equipmentyou're thinking about buying and
put it where it would go andtalk about that.
So uh essentially I see todaythe Vision Pro is the only thing
possible for million-dollarconversations.
If something matters that muchto a sales process or we're
collaboratively designing a newcruise ship room, it matters.
(08:21):
That then I don't have to putpeople on planes.
It's amazing.
SPEAKER_02 (08:24):
That's amazing.
There's a video online that mydaughter has gotten really into
the Titanic lately and all ofthe history about the Titanic.
And it's a video where youvirtually walk through the
entire recreation of theTitanic.
I I feel like that would besomething that would be really
cool to experience, you know, oror being alive in Jurassic Park
or um doing a Dungeons andDragons type game where you're
(08:45):
actually dungeon fellow.
SPEAKER_01 (08:46):
And there you're
speaking to me.
Yeah, absolutely.
Yeah.
SPEAKER_02 (08:49):
Game night just got
a lot more interesting.
SPEAKER_03 (08:51):
Yeah, I think it was
re or when I had the headset,
they'd they were doing the musicon Apple Music.
You could like sit in therecording room when you know one
of the artists, or I think itwas like Metallica has one where
you can like be on stage withthem.
Oh, cool, or something of thatnature.
And that to me, again, I'm goingback to the solo experience, but
I think that is probably thefirst thing that most people are
(09:12):
gonna have be mind blown by thekind of the tips, like like the
party, the party trick, right?
Oh, yeah, like you get in aTesla, right?
And it's just so fast, right?
That's the party trick.
Like the party trick for this tome, for the average user, is the
they're on stage with theirfavorite artist or they're
watching a basketball gamecourtside, right?
Exactly.
(09:32):
Right.
So that's that's very cool tosee that coming up.
SPEAKER_02 (09:35):
I was wondering
where's the crossroads between
VR, augmented reality, and andAI.
Uh you've worked a lot with AI.
I think you even have a book umthat you just wrote uh about
internalizing AI into yourcompany culture.
I'd love to hear about that.
And then where the crossroad isbetween AR, VR, and and AI, AI.
SPEAKER_01 (09:55):
Uh so the the book
is uh practical AI for leaders.
And it it really is um it'sabout our journey of how do you
bake it into your culture?
How do you start gettingeverybody?
And I'm not gonna say AI firstbecause that's not what it is.
It's just AI leaning.
Every time we're touchingsomething, we should be asking,
hey, could AI help me do thisbetter?
Um and it's it's telling ourstory of how we did that, which
(10:16):
was um last year uh we reallyput together some simple
programs around expectationseverybody in the company needed
to experiment and then sharewith everybody else in the
company how that went, and thenthe the fast forward on that of
how do you start building upteams to figure out which tools
we can use and how do we put theright processes in place and
those things.
So the the book was all aboutthat.
(10:36):
Um and the uh AI side of thingsgets really interesting when we
link it to spatial.
And I won't say VR, I'll sayspatial.
Um, because I mean again, thethe avatar of me that you're
seeing here, my persona, there'sa lot of processing power going
into that.
That's you know, scanned onceand used forever.
(10:57):
But being able to allow me tothen interact with something
else at that level of fidelitythat isn't real, that's powered
by AI, becomes an incredibletraining tool as an example.
Um, we've talked to a number offolks that have uh sales forces,
and maybe they're apharmaceutical sales rep that
has to get good at talking todoctors.
Well, I can practice this withanother salesperson, but they're
(11:20):
not an oncologist.
If I can actually practicetalking to a virtual oncologist
in their office that feels real,the skills retention there is so
much deeper.
So that that kind of thing justgoes is huge.
Uh, we were actually talkingwith someone else at one of the
airlines.
They're using the Vision Pros tofirst prototype out um what new
(11:43):
gates at an airport might looklike and then to go through that
because they do that todayphysically.
It takes them 30 days to buildphysically this gate and then
bring a bunch of people in toreview it.
Their thought is, well, they canbuild it digitally, and then
they could talk to an AI andsay, hey, you know, give me a
chair that looks like this.
Put 10 of these over there andessentially use that to help
(12:06):
them do the design work in themoment and just accelerate the
heck out of it.
SPEAKER_03 (12:11):
Yeah, it's all
happening real time.
You can test, tweak right thenand there.
SPEAKER_01 (12:15):
Exactly.
Exactly.
And that that conversationalpartner piece, I think that's
one of the huge powerfulelements of AI.
SPEAKER_03 (12:20):
Yeah.
There was when the, you know,we're seeing the training
aspect, kind of what exactlywhat you're saying, but they
were showing like a doctorpracticing surgery.
They're showing, I think it wasFormula One or NASCAR, they were
practicing changing the tires,right?
Like the getting some musclememory down by, you know, going
quick and like through arepetitive nature.
Not you don't have to actuallydo the physical tasks.
(12:43):
You can you can practice on yourown, you could practice with
with the whole group, but youcan do it in a controlled
environment.
Love that.
Um speeding it up, cutting a lotof cost out for organizations
that could be for medicalindustry, it could be for racing
industry, it could be forcybersecurity industry.
There's a lot of use cases forthese testing, this testing
range that we're looking for.
SPEAKER_01 (13:02):
Definitely.
And AI becoming critical becauseit's observing how you're doing
and providing feedback.
Hey, you missed step three.
You left the screwdriver overthere.
Um, or fast forward, that's someof the things that again, um
we're seeing Apple really leaninto is the uh I'll call it more
the the frontline worker usecase of all right, I'm wearing
(13:23):
this and I'm in a high-techmanufacturing job and I'm
cleaning and manipulating a bigpiece of equipment, and it's
watching me do the process.
And the AI is actuallyrecognizing, okay, did I put the
parts in the right places in theright order?
That gets really interesting.
We're still super early in that,but it's all technically doable
(13:44):
now.
SPEAKER_03 (13:45):
Do you think, you
know, when we're having the
conversation about it, we'reearly on.
When do you think we're gonna befar enough down the pipeline
where this is gonna be more of areality every day?
Is this like 15 years away?
Are we talking three to fiveyears?
What what's your thought onthat?
Is it soon?
SPEAKER_01 (14:00):
It's this is coming
so fast.
Uh and again, iPhone 1, what I'mwearing here.
Yeah.
Um, but the iPhone 1 took, I'llsay, five or six years to start
changing everything.
I think this is gonna take maybetwo or three before Spatial is
having the impact.
Um, AI coupled with that, Imean, next year.
(14:21):
It's moving so fast.
The technology is um the rate ofadoption of AI is probably ten
times faster than it was for uhmobile.
SPEAKER_03 (14:31):
As Josh mentioned,
we're doing an AI class, or
actually rather just uhcompleted a couple weeks ago.
And I don't know if we're justso much more immersed in it now.
I mean, we were before, but wewanted to really strengthen our
ability.
So we sought out this or lookfor this class.
And I don't know if we're somuch more immersed in the world
that we're seeing it, or if Ithink AI is even picking up more
(14:52):
speed over the past, let's sayone year, it's everybody's using
it now.
My mom is asking about Chat GPT,right?
Or you know, everybody's askingabout everybody's using it, it's
writing our emails now, right?
It's helping everything,unfortunately.
Unfortunately, but we're seeingit more and more and more.
So that's what prompted me thequestion of kind of where you're
(15:13):
getting at what's what's next,what the next two or three
years, like you're saying, it'sgonna be more mainstream.
People are gonna start adoptingit.
Because I think that is a partof the problem we're seeing in
not only in our industry, butacross almost all industries is
the adoption rate and howthey're adopting AI.
Are they need to be doing itprobably a little bit more
(15:35):
quickly, but doing it correctly,right?
Policies, procedures, setting upthat governance within the
organization.
And you know, and I brought thisup before on the show, that one
of the biggest problems ispeople needed to get on board
quicker because it's a you know,the discussion of Netflix versus
Blockbuster, or way back beforethe the iPhone came out, Nokia,
(15:56):
right?
They had the whole market space.
They're virtually gone.
Kodak with cameras, they're notmaking film, right?
They're they're out of businessbecause they didn't follow
trends in the in the industries.
SPEAKER_01 (16:07):
Agreed.
And that's that's part of whichwhat's exciting me again.
I I first touched AI touched AI,I'll date myself a little bit,
uh, almost 40 years ago.
SPEAKER_03 (16:16):
Okay.
SPEAKER_01 (16:17):
It was not ready for
prime time.
It was very different.
Fast forward, um, what I lovedabout mobile technology was it
empowered people.
It let folks who weren't geeksreally lean in, and AI is that
exponentially.
Anybody can use it.
If you can talk, if you cancommunicate effectively, you're
suddenly in charge of atechnology assistant.
(16:38):
And that that's just amazing.
Uh and to the Netflix point, I'mhaving conversation after
conversation with I'll just saypeers, you know, folks who are
leading mid-market companiesthat are looking at this of, you
know what?
If we adopt AI, we can reinventhow we're doing things.
And these aren't tech companies.
Absolutely.
(16:58):
Um, I've uh someone inparticular um in the um the
kitchen remodeling world, he hastaken what was historically not
a lot of tech and turned itinto, well, we can map out what
we're talking about while we'resitting here, show you a
picture, what it could looklike, hit a button, and the team
knows what to go build.
SPEAKER_03 (17:18):
We're accustomed to
thinking or judging somebody on
what they know, but not whatthey can learn.
That's it.
Right?
So if we can reverse thatthinking, so so let's say Josh
isn't a cybersecurityprofessional, but we give him
AI, he he can fill in, he can doa lot of things.
He can be a lot more than justscary, he can be very dangerous
(17:38):
in the space and doing a lot ofthings, great work that maybe
somebody with a four-year degreethat's right, that's just out of
the gates.
We can get speed things up andand really supercharge
somebody's abilities uh ifthey're used it the correct way
and you teach them how to dothat.
It's all about exposure to AIright now.
SPEAKER_01 (17:55):
It it is, and you
nailed it, it's the it's the
speed and inclination tolearning.
That that's the key, being opento those things and using it as
a tool, learning those newthings.
And and again, today it's AI.
In ten years it will besomething else.
The same way that you knowmobile was the key word back in
2008 through 12 of I need amobile app.
(18:18):
The word mobile mobile is gone.
Fast forward five or ten years,we're not gonna say AI anymore.
It's just gonna be a part ofeverything that we do.
SPEAKER_02 (18:26):
And I still have
family members that can't even
sign into their iPad or downloadan app.
So I wonder what kind of yeah,what kind of tech support's
gonna be needed there?
Most people are getting used toAI through writing prompts,
right?
Or or just maybe replacing theirtheir search engine with ChatGPT
or Claude.
But what would you recommend forpeople that have gotten to the
(18:47):
next level?
Like we're we went from writingprompts and asking questions to
creating Chat GPT projects orcustom projects or cloud
projects.
What would be the next stepafter that to further immerse
yourself or your organization?
SPEAKER_01 (19:00):
Good good questions
in terms of what that continuum
looks like.
Um so looking at AI, there itreally breaks into kind of three
categories.
So the first I would call reallya reactive assistant, and that's
exactly what ChatGPT is, orcustom GPTs.
It's I ask it a question or Iask it to do something, and it
does it.
Um and there are thousands andthousands of tools that are
(19:21):
doing that.
The next category is a biggerstep just because it requires a
little bit of learning, butthat's where we start getting
into task-based agents.
So we use that magic word agent,and we have to be careful
because it means 50 differentthings to everybody.
But it's tasks step-by-stepautomation, where if I do this,
so for example, I post anon-disclosure agreement into
(19:44):
this Slack channel or thisteam's channel, that the agent
would grab it, review it, andput comments back in the
channel.
That that's a task agent.
The next step is what I'll callum more actually digital
employees, or we call themdigital staff, where they're
goal-based, where that digitalstaff is focused on, you know
what, you're a businessdevelopment rep.
(20:05):
Go find prospects for ourbusiness, research them, reach
out to them via email, andschedule the salesperson
appointments.
It has a goal and a bunch oftools, and it figures out what
to do when and how.
So let's put that totally aside.
The automation piece in themiddle, that's where tools like
uh a Zapier or an N8N get to beinteresting, of just to start,
(20:27):
okay, if I could plug thesethree things together and add a
brain in between them, whatcould I do?
SPEAKER_02 (20:31):
How much time and
stock should we put into
learning Zap and N8N when AI isjust around the corner?
Because they're pretty intenseprograms, right?
Um they do require a little bitof attention, a little bit of a
learning curve.
But then I'm also thinking like,well, if this is agentic AI is
just around the corner, should Ireally be spending, you know, a
thousand hours or even a hundredhours learning N8N?
SPEAKER_01 (20:53):
Uh interesting
question.
And uh I don't think Agentic AIis going to replace any of that.
Um the so for example, the MCPprotocol that's coming out that
allows you to essentially it's ageneric API for an AI to talk to
a system.
That's really interestingbecause it would allow an agent
to talk to the CRM or the emailsystem or whatever it is without
(21:15):
having to care what that API is.
Many of the tools, like N8N,there's an MCP server that you
can put in front of that.
So you could describe theautomation that you're looking
to have built, and it couldactually craft that and put it
together for you.
But to the question, um I thinkthis is a good example where
understanding, hey, what does itlook like to integrate things
(21:38):
together?
What is a tool?
So that when that agent is doinga thing that we know, oh, okay,
it's a tool, it's talking to asystem.
Um, I think that's actuallyreally important for most folks
to uh if you're technical, youhave to.
If you're inclined, you should.
SPEAKER_03 (21:53):
Yeah, how you are
actually interacting with the
tool.
Yeah.
It's interesting.
Random question, then, Alex.
Do you have a favorite rightnow?
If you had to pop open one ofthe tools, what are you reaching
for?
SPEAKER_01 (22:05):
Um, so my my
defaults, um, because I've done
a ton with it and it does agreat job.
Uh so chat GPT is my that's mygeneral go-to.
Um been experimenting with thethe 5-0 model or the five models
that just came out to see, okay,what's that really add to the
mix?
Um, love doing deep researchwith it.
That is absolutely my go-to.
Anytime I have a brain sparkabout a topic of I wonder, a
(22:30):
research request goes off.
So map it out.
And whether it's a go-to-marketquestion of, hey, who's doing
things in this industry, or howwould I do technical topic X?
I love that.
SPEAKER_02 (22:40):
I'm gonna defer to
Nick on some of this stuff, but
I wanted to generally ask thequestion you know, what kind of
blind spots are you seeing withAI security?
We're we're basically rushing togiving AI a lot of power and a
lot of insight into our personallife, into our schedule, into
our emails, um, so that it makesthings easier.
What do you see that balance of,you know, making tasks easier,
(23:03):
taking the workload off, andthen kind of preserving some
personal security orinformational security?
And how do you weigh thecost-benefit uh uh ratio on
that?
SPEAKER_01 (23:15):
So that was about 50
questions.
I gotta get them in while I can,Alex.
But no, that it that and that isthe challenge that we've got
right now because people areblazing straight ahead and and
rewinding a little bit, and I'lluse the iPad as the example uh
when it came into the businessworld.
It was brought into the businessworld by business leaders who
are like, I need this, I wantthis, and whether that was a
(23:37):
doctor or a CEO, it didn'tmatter.
And the technology teams were sofar behind that it was it made
it difficult for everybody.
They couldn't figure out, okay,how does this connect to the
network?
How do I secure it?
How do I do anything with it?
And many IT teams today, anycompany size, are doing the
exact same thing.
They're struggling.
There are so many ways that wecan empower teams, but is this
(23:59):
tool secure?
Did somebody just sign up for atool that's free that's stealing
all of our information?
Which so that step number one isthis the stepping back and
communicating with the teamaround, hey, what are our
expectations?
What is responsible use of AI?
It doesn't mean here's the fivetools you're allowed to use, or
something we see very frequentlyin big companies is the answer
to AI use is no.
(24:21):
Well, that's not real.
Come on.
What are we gonna do with it?
What does responsible mean?
Number one, you have tounderstand what are the
licensing terms of this thingwe're using?
What does it do with ourinformation?
And for non-geeks looking atlegal agreements, that does
that's not awesome.
Um, internally, we wrote acustom GPT to analyze that.
So we could give it a tool name,it would go out and grab all the
(24:42):
documents, bring it down andsay, yes, they're gonna train on
your data.
And that that's the magicquestion.
Um but the simple statementbeing if it's free, you're
giving them your data.
Forget it.
Don't ever do that.
If it's paid, okay, well thenget somebody involved if you're
gonna use it officially.
But while you're tinkering,don't give it state secrets.
(25:03):
That doesn't work.
Um but taking those to the pointwhere we know what's going on.
Um now for me personally, again,I'm coming in on the Vision Pro.
Um, one of the things I loveabout Apple is their privacy
stance.
I think one of the hugechallenges, and I don't know
that folks see it coming, isthat when we couple AI with the
(25:25):
company that hosts all of ouremails and our documents and
that makes money by sellingadvertising, that's a bad combo.
And we have two megacorps thatsit there at the center of that,
that's dangerous.
SPEAKER_03 (25:41):
You you lob it up
perfectly, Alex.
I'm was thinking about thesethings along the same lines.
We're in a we need to, iforganizations haven't gotten or
stood up a governing body forAI, get it get a board of
people, four or five people,whatever you decide on, and
start deliberating on whatdirection you're taking with AI.
(26:01):
Because it's coming, you gottado it.
You have to do it now because werun the risk of if we don't do
that, your employees are gonnado it on their own anyway.
SPEAKER_01 (26:09):
Shadow IT is real,
it's gonna happen.
Of course, yeah.
SPEAKER_03 (26:11):
We've seen it all
the time through plugging proof
point.
You know, we see it at one ofour clients that they're
emailing each other back thetheir work and their personal,
they're emailing back and forth.
So, what do you think they'redoing, right?
They're taking their informationthat they're working on at work
and they're sending it to theirLLM and their personal machine,
and they're sending it back withtheir output, right?
So they're using it already.
(26:32):
Uh sometimes they might be usingit uh you know on their machine.
But so first off, it's you know,figure out the direction you
want to go.
Let's peel the hood back andfigure out where you want to go,
what we want to use it for, ifyou're gonna run it, uh, if
you're gonna let if you're gonnause cloud, if you're gonna run
on cloud or if you're gonna runon-prem, right?
How are you gonna do this?
Um, and and then you have to getpolicies and procedures out.
(26:55):
You have to uh educate yourstaff, and then you have to
train your staff on how toactually use it, right?
It's one thing for us to say,oh, don't use state secrets.
Right, that's just a given.
That's things that the three ofus we just know that is a no-no,
we're not gonna do that.
SPEAKER_01 (27:10):
But what does it
really mean?
SPEAKER_03 (27:11):
Yeah.
What does it really mean?
Exactly, right?
SPEAKER_01 (27:14):
Yeah, totally agree.
And it's helping them becomfortable with it.
And again, for me, that's abouthow are we baking this into our
culture?
This is the new norm.
This isn't as simple as, hey,we're just gonna hand everybody
Excel who doesn't have it.
It's it we're not giving you atool.
We have to change how we'rethinking about work.
Every time I'm about to do atask, how could AI maybe help me
(27:35):
with this?
And just giving simple guidance,not IT scary stuff of no, no, no
tools.
Like, no, we have we have toembrace the whole company and
surfacing this stuff.
IT can't own all of it.
SPEAKER_03 (27:46):
Totally agree.
Totally, totally agree.
I I I actually thinkorganizations could run a risk
of overprotecting this.
You can tighten the bolts toohard.
Um, and that and we run thatbalance in cybersecurity all the
time, right?
We people need to be function,they need to be able to
function, right?
I mean, if we're in a perfectworld, we would just disconnect
from the internet, right?
(28:06):
And then we're gonna write inthe cybersecurity world.
That's not an option for us,right?
So we gotta ride that razor thinline of tightening the bolts too
hard.
And we're having the sameproblem, I think, with AI.
I think a lot of organizationsare gonna come in and they're
gonna put too many policies,procedures around it.
We should be deleting enough orwe realize we need to add things
back.
SPEAKER_01 (28:26):
It's interesting
when you hit on the potentially
clamping down too hard.
Um, that's something that we seevery, very frequently in the big
businesses in the Fortune 500because they have very
sophisticated IT security anddata teams.
So, especially over the pastyear or two, many of those teams
have raised their hands saying,we own everything.
Nobody touch it, nobody doanything.
(28:50):
And unfortunately, many of themhave been successful in wrapping
their arms around it and kind ofshutting everybody out, and
which again has totally missedwhy does this technology exist?
It doesn't exist for that groupof people, it actually exists
for the people who aren't theexperts, aren't the geeks?
Wait, I can just talk tosomething, I can chat with
something.
(29:10):
And so where we're seeing peoplebe successful is actually
focusing on the employees.
Maybe it's an airline pilot or aflight attendant, maybe it's a
salesperson.
What would help this person dotheir job?
What would empower them?
What would take friction out ofthe way?
And the part that, again, thegroups that are locking it down
are missing is if we canidentify the couple things we
want to help people with, we canjust trace a thin line through
(29:32):
the back end systems, the data,the policies.
We can figure out that partfirst.
And then we can do the next lineand figure out that part.
Instead, we're boiling twooceans.
We're boiling the ocean ofwhat's all of our data?
Make it accessible.
Well, you know, to powersomething for this person over
here, you might not even havethe right data yet.
But most folks, hey, let's wehave this, so let's take our
(29:53):
ocean and figure that out.
And then let's look at thesystems.
How do we AI enable all of thesesystems and make them all
accessible?
I appreciate the thinking, butspending three years on that
means you are going to be so farbehind.
When I look at mid-market andsmaller businesses, I love the
passion of business leaderssaying, we have to do this.
Hey team, let's figure this out.
And they are moving so muchfaster, which is actually
(30:16):
hinting to me that I think we'regoing to see a lot of leadership
positions in a bunch ofdifferent industries change.
The internet drove that, mobiledrove that.
This is driving it even faster.
And I think we're going to seesome small to mid-sized
companies that become really bigbecause of what they're doing
with AI.
SPEAKER_02 (30:33):
Interesting.
Is that a sea change in thinkingof employees not just as another
cog in the wheel to accomplish atask, but treating each
individual like their ownentrepreneur within the company?
Is that kind of where you'recoming from with that AI
enablement?
SPEAKER_01 (30:48):
And that's an
interesting way to put it.
And that that is my core beliefis that things like AI should be
giving people superpowers.
We should be helping that oneperson be 10 or 100 times more
effective, not how do Iimplement AI so I can fire my
whole team?
I think anybody doing that,number one, you're focused on
cost cutting and that's not goodfor growth.
(31:10):
That's not where you grow.
Number two, the great ideas ofwhere AI is going to transform
the business come from thosepeople.
And they're the ones that havethe idea.
So once they get comfortable andthey can lean in and have that
entrepreneurial always learning,hey, what if?
When they start to ask the whatif at times 10 or 100 or 1,000
(31:31):
employees, that changeseverything.
SPEAKER_02 (31:33):
Seems like it
started with an ultra kind of an
altruistic stance.
I mean, baked into the name,right?
Open AI, it started out opensource, I believe.
But now we have this kind of AIarms race going on with nation
states.
Um, how do you see that squaringoff with kind of it seems like
things should be moving more toopen source?
(31:53):
Um, if everybody's enabled tokind of be their own
entrepreneur and um everyone hasLeonardo da Vinci and Albert
Einstein in their pocket.
Um, how do we get from where weare in this kind of like keeping
things close, but allowingpeople certain liberties?
And how are we going to balanceall that moving forward?
SPEAKER_01 (32:12):
Uh that's an
interesting one.
Um, I don't know that itnecessarily needs to go open
source.
And I think again, back to whatwe the humans control, we
control being greatcommunicators.
We control coming up with theideas and framing what it is
that we want the AI to do.
Um, and then even, hey, here'sthe process.
When we're building things, forexample, even in the agentic
(32:32):
world, I mentioned ChatGPTearlier.
You know, it's a simple go-to.
But you know what?
If I'm writing code on my Macand I want to be having it
generate different types ofcode, I might be pointing at
different systems.
So the the simple statement ofnot ever being locked into a
single vendor is more importantnow than ever in technology.
We've got to be able to takethat prompt that I created and
(32:55):
say, you know what?
Since this vendor is not betterat this thing, I'll just move it
over here.
Maybe there's a little bit ofrewiring, but we have to be
crafting things in such a waythat there's flexibility there.
SPEAKER_02 (33:06):
I'm gonna hand it
over to Alex, but I just want to
say, you know, you're using thepaid version, which it's saying
we're not training off yourmodel, but uh is there any
promise there that they're notgonna be selling data or you
know, packaging it up and formarketing um usage or
essentially.
SPEAKER_01 (33:20):
And that's where
those license agreements come in
and are so critical of yeah,what can they do?
And I'll I'll use Midjourney asthe example.
There's the free version, youknow, you publish stuff and it
goes out in the world.
There's the paid version tierone where you can do it and you
can commercially license it, butthey retain the rights to use
anything you created in theirmarketing.
(33:41):
That's not cool.
So then you have to buy that thenext tier up to be have shadow
so that anything that you createis yours only.
So you have to pay attention tothat if it's it's commercially
important.
Um Nick, back to the writingemail since it's come up a
couple of times.
Uh to me, that's actuallyanother example where I think
Apple is um as they continue toevolve the on-device models,
(34:03):
that's our future.
Being able to run the model onthe iPhone, I'm running the iOS
26 beta.
I built my own version of an appthat is a mini chat GPT that
runs on the phone so I can chatwith it, chat with the model,
but that's 100% private.
Anything that's going throughthe Apple ecosystem, let's
forget all the comments of who'sthe leader on what right now.
(34:25):
They're gonna surprise the heckout of the world because there's
so much AI baked into thosedevices already.
And with those models running ondevice, they could become the
largest AI company overnight.
SPEAKER_03 (34:38):
I do appreciate on
the on-device because as
security engineers, that's thebiggest thing, right?
We want to hold our data andhold the keys to that.
SPEAKER_01 (34:45):
Exactly.
And I think that's where, again,over the next 12 to 24 months,
um, and I don't know that mostfolks have really acknowledged,
and having worked very closelywith Apple for 15 years, I've
seen their privacy stance, I'veseen their security stance, and
there's no other company that'stheir peers, and I will conclude
the AI companies now, especiallyin that, that is paying any
(35:05):
attention to the kind of stuffthat Apple is.
That we the user own our data.
Even for a corporate enterprisedevice, the okay, I'm what does
the employee own and what doesthe company own?
They they have put so mucharound that to allow you to
separate those things and dointeresting things with that.
I'm really expecting that as wehave 26 going live and 27, um,
(35:27):
the secure AI, that's gonna be abig part of the marketing and
what we all have to payattention to going forward.
And I think Apple's gonna driveit.
SPEAKER_02 (35:35):
We gotta get our
Apple Visions.
It was probably, you know, wegot to get Eric on.
It was probably good that hewasn't, because then we could
have got all of our questionsout and we weren't stepping on
him because I'm sure he could goall day with you too, Alex.
So um I want to thank you foryour time and uh we really
appreciate you coming on theshow.
You've been listening to theaudit presented by IT Audit
Labs.
My name is Joshua Schmidt, yourco host and producer.
(35:55):
Nick Mellum riding Shotguntoday.
And our guest is Alec Bratton.
And you can check him out onLinkedIn and see some of the
things he's got going on andthen check out his book as well.
Um, you can find the auditanywhere where you source your
podcasts.
We now have a video on Spotifyand YouTube.
Please like, share, andsubscribe.
And we'll see you in the nextone.
SPEAKER_00 (36:15):
You have been
listening to the audit presented
by IT Audit Labs.
We are experts at assessing riskand compliance while providing
administrative and technicalcontrols to improve our clients'
data security.
Our threat assessments find thesoft spots before the bad guys
do, identifying likelihood andimpact, where all our security
control assessments rank thelevel of maturity relative to
(36:39):
the size of your organization.
Thanks to our devoted listenersand followers, as well as our
producer, Joshua J.
Schmidt, and our audio videoeditor, Cameron Hill.
You can stay up to date on thelatest cybersecurity topics by
giving us a like and a follow onour socials, and subscribing to
this podcast on Apple, Spotify,or wherever you source your
(37:01):
security content.