October 22, 2025 • 49 mins
“Prepare: AI is the new CU crisis super villain.” That’s the title of a recent CUInsight story authored by Casey Boggs, founder of Reputation US, an d of course we had to get him on this podcast.
Understand, I am a strong supporter of AI in general and AI in particular inside credit unions. This is a life and death matter.
Yet there is a possible downside to AI and we already know that the main AI tools have played substantial roles in teen suicides, in creating false “facts,” and in many other ways leading humans astray.
While I may disagree with many of Boggs' conclusions, his core advice - proceed with AI cautiously and thoughtfully - is on the money.
Only fools rush in.
Read Boggs’ CU Insight piece, listen to the podcast, read a recent CUInsight story by me where I explore CU use of AI with CU2.0 founder Kirk Drake, and keep on learning and experimenting and using AI.
Listen up
Like what you are hearing? Find out how you can help sponsor this podcast here. Very affordable sponsorship packages are available. Email rjmcgarvey@gmail.com
And like this podcast on whatever service you use to stream it. That matters.
Find out more about CU2.0 and the digital transformation of credit unions here. It's a journey every credit union needs to take. Pronto
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_03 (00:00):
Welcome to the TU2.0 podcast.
SPEAKER_00 (00:05):
Hi, and welcome to the CU2.0 podcast with big new
ideas about credit unions andconversations about innovative
technology with credit union andfintech leaders.
This podcast is brought to youby Quillo, the real-time loan
syndication network for creditunions, and by your host,
longtime credit union andfinancial technology journalist
(00:27):
Robert McGarvey.
And now the CU 2.0 podcast withRobert McGarvey.
SPEAKER_03 (00:35):
AI is the new CU Crisis Supervillain.
That's the title of a recent CUInsight Story authored by Tasty
Box, founder of Reputation U.S.
And of course we had to get himon the podcast.
Understand I'm a strongsupporter of AI in general.
And AI in particular insidecritics of this is a life and
(00:59):
death matter.
Yes.
Yes, there is a possibledownside to AI.
We already know that the main AIis falsified substantial team
suicides and false facts andmany other ways of leading
humans of strife.
Well, I may disagree with manyof boxes, conclusions, of
(01:20):
course, but see what AIcostously and thoughtfully is on
the mind.
All the fools are too read boxesto you insight, please just a
link to that in the show notes.
Listen to the podcast.
Read a read this to you insidestory about me or the force to
use the way I was to you to theold founder portrait is a link
(01:43):
to that story in the show notes.
And just keep on learning andexperimenting and using AI.
It's the future.
Learn how to use itintelligently and wisely.
Now, what kinds of AI disastersor mini disasters or acts of
bond villainy are do youanticipate seeing?
SPEAKER_01 (02:07):
Yeah, anticipate and already seeing, I would say
probably in three to fourbuckets here.
One is just simply thecybersecurity and data privacy
is a big piece of this.
So the world of cybersecurityhas been going on for quite some
time, right?
And there's been no shortage ofan industry that has been there
to stave off these bad actors,as they call them, or malicious
(02:31):
actors that are trying toessentially come in and take
data and ransom your informationand try to make money, right?
So similarly, you know, thisemerging AI is really uh
mishandling member data.
And you know, they'remanipulating the information so
uh that they're just simplytrying to uh gather something of
(02:54):
value from the credit union thatis you know that at some point
the credit union has to wantthat back.
And so using AI in differentforms to uh penetrate those uh
those the member data or uhfinancials, that's where we're
seeing probably the mostpenetration early on with AI.
(03:15):
And then another subset to alittle bit of that is deep
fakes.
And to what degree the word deepfake is essentially just a way
to falsify uh a person's voiceor image, and we're seeing more
of this just to falsify when weget an email or a voicemail or
(03:37):
sometimes even video of a personthat pretends to be someone from
a credit union saying, hey, uhgive us a call, give us your
information, your socialsecurity number, your uh
financial information, and thatI haven't seen it as much, but
it is still happening, and thatwill essentially cure losses for
the members because they thinkit's someone from the credit
(03:59):
union, right?
And they're getting they'regoing after the vulnerable
folks.
The um the bad actors are goingafter the folks who might be
vulnerable or um I don't knownaive or just simply um, I guess
crazy enough to give theirinformation to someone who
seemingly is the uh executive ormember services person that they
(04:20):
know, trust, and like.
So that's uh that's one we'reseeing as well.
Another one in all this islending discrimination.
And that one is about it reallyfits in not only credit unions
but banks as well and just otherfolks who are in the risk
business area.
But AI models um you know reallythey look for uh these uh loans
(04:42):
or impose unfavorable terms onprotected groups, and uh with
that becomes uh there's scrutinybehind it because they are uh
being biased or there'sdiscriminating against members
of a certain class ormarginalized group.
And that's a big no-no.
And uh and that happens to besomething that not only from a
(05:06):
reputation side, but certainlyfrom a regulatory and even
lawsuit-wise, could be uh badfor a credit union.
SPEAKER_03 (05:17):
Now, obviously, discrimination lending was
invented by human beings who tosome extent perfected that until
some federal laws came along andsaid, no, no, you can't do that.
It's uh I don't see why themachine can't be taught that it
can't do it in the same way thatbanks and credit unions have
(05:38):
taught their lending staff thatno, you you can't make decisions
based on race, race, age, blah,blah, blah.
It's uh um that shouldn't bethat difficult a thing to teach
the machine.
SPEAKER_01 (05:52):
You would hope not.
You would hope not.
I think it still applies becausethis is when human uh oversight
is very important.
Uh so like a checks and bounces.
So if you are crediting andusing AI for the first degree of
um determining if someone is arisky or non-risky lender or you
(06:13):
know uh client, uh, that's onepart.
But if you're not necessarilygoing underneath the hood to
find out more information aboutthis person, uh and you're
taking the AI recommendationsfor what it's worth, then that's
on you.
That's on the credit union fornot doing its due diligence.
But part of the the misstepsthat are actually happening here
(06:36):
is you know, humans arebecoming, dare I say, lazy or
don't do their due diligence andusing the things that are
afforded to them, like you know,AI tools, uh and um the checks
and balances of this um are injeopardy, which essentially,
again, on the fault of thecredit union or the the lending
(07:00):
bank.
SPEAKER_03 (07:02):
Yeah, that that said, we use machines to make a
lot of credit decisions.
I mean, I applied for a newcredit card uh a week or so ago.
And within a minute, I wasapproved.
Um how did they approve that?
They got my FICO score, period.
I mean, there's there's no otherway they could have gotten any
(07:24):
information about me that wasuseful in that time frame, other
than uh a FICO score.
And they said, oh wow, we'llgive this guy a credit card, of
course.
SPEAKER_01 (07:33):
Right.
I think it might be I thinkthat's you know, that's credit
is one part, and then the otheraspect is just lending money,
you know, you know, the equityline of credit, you know,
lending money for a car loan,that type of thing.
Uh for just a credit card, Ithink that is automated based
off what you just said.
Uh you know, numbers don't lie.
If they have you know FICO scorethat's you know in the 700s or
(07:56):
above, that immediatelyqualifies you to take out a
credit card.
SPEAKER_03 (08:01):
But I think it gets a little more sophisticated when
it comes to well, also like thecredit card, you might have a$10
or$20,000 credit limit, but it'shighly unusual to take that card
the day you receive it and runout and put the maximum amount
out.
I'm sure Kirks would do that,but that's not normal.
And if you're missing paymentsalong the way, at some point
(08:22):
they pull a plug on it.
That's right.
And they're not gonna lose thewhole amount in most cases.
So, I mean, yeah, creditlending, this has been going on,
credit cards have been going onfor at least geez, I don't know,
60 years.
They've gotten very good atthis.
And uh a lot of it's machinelearning, though, that's
actually pretty good.
(08:42):
I mean, I give the example thatif I walked into an Apple store
in Detroit, Michigan, and boughtthree super expensive um uh
Apple computers, that chargewould be declined uh by my Apple
card, which would say Bozo'snever been to Detroit.
He don't know where Detroit is.
(09:06):
And furthermore, he's neverbought this kind of computer, he
doesn't seem to like it.
Uh and that would be a smartmove on that that machine's
part, saying, nope, decline.
There is there are places wherethe the the learning makes
sense, you know, to use themachine learning.
That's right.
And you know, an issue thatinterests me is that let's say
(09:30):
you apply for uh uh$75,000 homeequity loan, and I'm looking at
your application, and I say,geez, I don't know if I should
say yes or no, but I'm a smartguy.
So I take your name out of it,but I put a lot of other
information into Chat GPT andsay, should I give a should we
give a loan to this guy?
(09:52):
Question What happens to thatdata?
Now, Google says we will nevertrain the system on your data.
Your data is your data.
That's what Google says withwith uh uh Gemini.
ChatGPT, there's a toggle youcan set which says, Don't use my
data to train.
Can you rely on ChatGPT?
(10:15):
Well, chat GPT has made someegregious mistakes in recent
days, recent weeks.
Sure.
Advising uh maladjusted teenageboys about how to commit
suicide, and some do at least.
SPEAKER_01 (10:27):
So right.
It's a whole nother crisis forsure.
Um because not just corporateparts, but there's a lot of AI
issues uh surrounding thosematters.
Yeah.
SPEAKER_03 (10:37):
Well, so that that that raises the question of can
we trust their word?
Can we trust their ethics, whichis uh scary stuff to bring up,
but in the case of the teenagesuicides and the advice, I have
advice and and quote marks thathave been given by by the
machine.
You say, wow, this is not thisis not good advice.
(10:59):
Yeah, you don't give a13-year-old who's depressed tips
on how to commit suicide.
One doesn't do that.
SPEAKER_01 (11:04):
Or in general, you know, I'm a father of twins, uh,
teenagers, and not my teenagerper se, but uh in general, uh
kids are looking for importantcounsel on where to go to
college, uh what to do incertain circumstances.
Uh, when we as a parent were thethat sanding board based on our
experience, now they are lookingto these you know AI tools for
(11:29):
pretty much everything, personalor otherwise.
And that gets very uhchallenging because it looks
good, it seems logical, butdoesn't necessarily fit into
what they know about the humanbeing, uh, that teenager.
Uh and so uh information isbeing shared while it seems
convincing and well thought out,or at least it is uh curated to
(11:52):
something that uh bite-sized forthe teenagers, it doesn't factor
in that teenager themselves.
There's so many differentvariables that fit into that.
Um, and it's you know, to relyon AI to provide those insights
is so limiting.
Very much still, I don't carehow much information you put
into AI, it just cannot uh gointo those nuances of uh the
(12:14):
human capital and the uniquenessthat every person is involved,
including like you know, go backto the credit union line.
If you're if you're providing uhlending to someone, there's a
story behind those people,right?
They might not necessarily havethe best credit for whatever
reason, but they're doing otherthings like you know, looking
for work, they are doing four orfive different jobs, they just
(12:38):
had you know uh some otherissues that prevented them from
having a good score.
So credit union and knowing alittle bit more about that human
being on a one-on-one basis,they'll say, Yeah, uh I get
their circumstances, I'm willingto give them this this loan.
Uh, if you based purely on justthe numbers or some other
factors, then the biases willwill surface.
SPEAKER_03 (13:02):
Well, that's you're talking about the old credit
union tradition.
I know when Jim Blaine was CEOof state employees at the Credit
Union of North Carolina, headamantly insisted they would
lend to the human being, not toa piece of paper.
And yeah, if you'd had uh ahealth crisis and you miss
(13:23):
missed paying a few bills hereand there, that would be taken
into account.
And you might get the the moneyto buy a car that would let you
get work drive to your job.
You might get that from St.
Employees because the judgmentwould be based upon the human,
not the fact that you had a runa little run of bad luck.
That's right.
A lot of credit unions havemoved off of that,
unfortunately.
(13:44):
But what can I say?
I don't run a lending committeeon a credit union.
So yeah, and it it's harder tofactor the human stuff in.
You have to talk to people, youhave to call for the person's
foreman, like in the old days ofthe select employee groups.
You know, tell me about them.
Is it reliable?
Well, you know, that kind ofstuff.
SPEAKER_01 (14:03):
So well, I do think that and one of the areas that
we brought up as well that yeah,the the the chatbots, you know,
are efficient, right?
But members um are oftenperceived them as cold.
And I don't know about you, butif I call up uh a company and
it's always simply get as anautomation or even a an online
(14:26):
version to give me the answers,it's really impersonal, or there
it's limiting, or it's got youknow air written all over it.
And if you are perceived as aninstitution that relies solely
on AI-driven chatbots, you areimpersonal, right?
And the member might resent you,or might oh, I can get this
(14:48):
anywhere, where the pendulum isgonna hopefully swing from AI to
people who are authentic.
That's where I think uh there'sgonna be more of an appetite.
As I I'm you know, I'm guessinghere is purely on opinion, but I
think at some point we're like,all right, this is ridiculous.
I want to talk to a human being,I want to go into the branch, I
(15:10):
want to go and meet thesepersons because there's too much
at risk here.
And if they're only going to bea lender or a relationship
manager purely based off ofnumbers and using AI just to
make money, then that's notgoing to be my financial
solution.
I need a human being.
SPEAKER_03 (15:28):
There's debate within the credit union world as
to what are the best tasks tofirst pursue with AI.
And to me, the winning argumentis to use AI on not on customer,
not on member-facing stuff, butto use it on back office stuff.
The most interesting example Iknow of a successful AI project
(15:52):
is that one Navada credit union.
They took all of the informationthat call center people consult
when you call up with aquestion, fed it all into AI,
aiming to simplify simplify.
There were many severaldocuments often dealt with the
same topic, sometimes gavecontradictory advice.
(16:12):
I mean, this is informationcollected over 30 or 40 years.
I mean, I'm I'm sure Icontradict myself over things I
said and wrote 20 years ago.
So they cleaned all of that upand uh have uh resulted in
terrific uh time savings.
So instead of an employeesitting there flipping through
(16:33):
pages, getting confused becausethey're contradictions, you just
ask the machine when will werepossess this guy's car if he
misses payments?
Within 10 seconds, you got ananswer.
So the the member's not sittingon the phone for three minutes
as you dial around.
Boom, answer.
(16:54):
So does the member know that AIhas played a role in this?
No, why should they?
It's irrelevant.
You're you're still talking witha person.
I mean, you don't I assume youdon't have any issues about
those kind of projects.
SPEAKER_01 (17:06):
Not yet.
And I and I say that because uhthose probably aren't as exposed
because if it is it is an erroruh that's going to be tracked
probably internally, and I can'timagine any institution is going
to track and tout some of theerrors they've done internally,
or they're they're seeing someflaws in that system.
(17:27):
They might you know uh talk tothe vendor who they're using or
the AI aspect that they'rethey're using and saying, hey,
this this does have flaws, but Idon't think it's one of those
things where they're going to uhuh highlight some of the you
know inefficiencies that aretaking place that's gonna be
forward-thinking to you know thepublic or uh prospective
members.
SPEAKER_03 (17:46):
So I know another place where AI is having
significant impact at creditunions is automating
collections.
So if you're late, instead of ahuman being calling up, leaving
a message, because you're notgonna answer the phone, sure, a
machine will call up and leavethe message.
And one of the most interestingthings about that is there so
(18:10):
far I have not heard of anycollection staff complaining
about this because usually mostof the people doing that find it
to be um work they don't reallywant to do.
And these credit unions have nothad any staff reduction.
So they've given people othertasks to do.
So it's not like, oh, we'regonna fire you and replace you
with a machine.
It's uh, hey, you don't have tomake those collection calls
(18:32):
anymore.
How's that?
You know, wow, this is great.
I came in, I had to eat Tums allmorning before I made my first
call.
No, this is wonderful.
Thank you.
God, I love AI.
SPEAKER_01 (18:48):
There's there are some good, yeah, there are some
good parts of AI for sure.
And uh it's not primarily topoo-poo on it.
In fact, uh it's gonna be animportant uh engine to different
aspects, you know, andaugmentation to uh things that
we're already doing and takingtaking the lead on.
There's no shortage of examplesto that end, uh, but it's it's
(19:10):
not where it needs to be at all.
And I'm not sure if it'll everbe, but in the same vein,
because of something that's new,there's going to be malfeasans,
nefarious actors out there, uh,and folks who are essentially
looking to game the system.
And that, you know, based onwhat we've seen so far, that's
(19:30):
already happening.
And it's gonna probably be moresophisticated and it's kind of
scary.
And so part of what we'retalking about today, about you
know the AI uh crisis of sorts,is that you need to be vigilant.
You know, what's going on rightnow in September 17th, 2025, uh
is likely gonna be a lot moredifferent than September 17th,
(19:53):
2026.
And uh it's it's it's scary,daunting.
Um, it's going to be a littlebit exciting, but we all have to
be participants and beingvigilant and being ready because
we don't necessarily know whatwe're gonna be um up against.
Right.
(20:13):
Yeah.
Crisis prep is gonna beimportant and not just for
preparation parts, and that'swhat we advocate for, but I
think just you have to geteducated to it.
It's one of those things wherewhen I talk to someone about
what is your degree of uhknowledge of AI, adoption, and
other things, it's still a slowburn for a lot of people, and
that's okay.
(20:34):
Uh, but for companies that aredealing with records, uh medical
records or financial records orjust information or there are in
the business of customerservice, this is has to be
paramount because there's a lotat stake here.
And in one little slip or uhmishandling of how you uh
(20:56):
handled the situation, it'sgoing to be a dark mark on on
you uh as a company.
And that's what we're justtelling folks and scream from
the rooftops now is that it itis here.
Keep an eye out for it, but notjust simply like waiting for the
crisis to happen, but having youknow safeguards in place.
(21:17):
And that's essentially havingalmost a team ready for being
vigilant, uh, ready to prepare,and to hopefully to mitigate
against any of the damage thatmight come your way.
SPEAKER_03 (21:28):
Well, go back to 1995 when every company's
rushing to be online.
They pretty quickly realize thatI I don't want you to take your
personal computer go down tolocal Starbucks, use their their
Wi-Fi or whatever they had atthat point, to get online and
(21:52):
dial into my mainframe and startdownloading stuff.
I just don't want this tohappen.
And they implemented ways tomake that not happen.
For instance, you have to use aVPN, you can't use a personal
computer, you know, stuff likethat.
A whole bunch of steps designedto keep the connection and thus
the data secure.
(22:13):
I I see companies needing, Ithink they need help, but I I
just see companies rushing tofigure out how, okay, we're
gonna use AI, how can we do itmore securely?
Um and I think you're right thatthe answers to that question
aren't that clear right now.
It's we're so busy getting thesystem up and working.
(22:33):
Uh security, we'll worry aboutthat later, man.
I mean, who worries about thatin Silicon Valley at the
beginning?
You know, we worry about itlater.
SPEAKER_01 (22:44):
Yeah.
It's uh you know, sellingpreventative medicines uh a
tough sell in general.
Uh we just need to get up and uhget up and ready.
And if uh something goessideways, then we'll go
sideways, we'll deal with theramifications.
Uh however, those are costlymaneuvers because there's a lot
at risk here.
That's the challenge on so manydifferent aspects.
(23:05):
Um, a couple of differentthings, Robert, you know, like
uh about being prepared or not,that um, you know, I have a few
stats that we have on ourwebsite about AI, but according
to this Oxford uh metrica, uhcompanies that handle crises
poorly lose 15 to 30 percent oftheir market value within days.
(23:26):
So, you know, that's a that's alot.
So, you know, the whole idea,let's we'll just handle it when
it comes in.
Well, if uh you're you're gonnalose 15 to 30 percent of your
market value when somethinghappens, uh, are you okay with
that?
You know, so that's the the thebell we're ringing here.
That preparation doesn't reallycost you much, but a crisis
(23:47):
costs you a ton.
It's really important for uhorganizations, AI or otherwise,
to be prepared.
And I'm just shocked that mostaren't.
Yes, they might in the creditunion world, yes, they might
have a um, you know, some in youin C UA um disaster recovery
plan in place, but that's morelines of a checklist they have
(24:10):
to do, but no one really knowswhere it is or what it's all
about.
But to actually practice on ituh and be prepared for it, um
that's the the key element thatwe find daunting because we're
called in the in the uh 11thhour or we parachute in during a
crisis and like, okay, let'sjust take care of the situation.
(24:31):
Hopefully we can mitigateagainst any damage.
But um, at that point, we're thefirefighters putting out the
fire as opposed to we could havereally uh saved a lot of the
embarrassments, uh a lot of thedata and materials if we were
just prepared.
And I'm not just saying it'sjust because it's a part of our
job, it's just real, it's justsad that um most organizations
(24:52):
aren't prepared for a crisis.
For some reason, they buyinsurance for things that are
that might happen, but uh forreputation, which actually, of
all things, reputation is aninsurable asset, they are
they're willing to roll the diceon um the that kind of damage,
and that's proven a bad move.
SPEAKER_03 (25:13):
To my knowledge, uh NCUA hasn't issued meaningful
documents regarding AI usage andcredit unions.
I might be wrong on that, butI've been talking to so many
people recently so often aboutNCUA.
I don't think I am.
SPEAKER_01 (25:29):
Well, you're you're right.
Um uh in fact, I was just justbefore our meeting today, I was
on a AI conference call withBeasley's, which is a you know
insurance carrier, and they wereunderscoring that there aren't
much regulations right now atall.
Uh in the Europe and the EU,they have quite a few already,
(25:51):
which is good.
There's some oversight, but inAmerica, there's not much.
And then if you get a littlemore granular, like a credit
union industry, there's not muchat all.
I can't speak specifically of uhevery industry, but I think it's
still so nascent and uh probablystill up for debate to what
degree it can be regulated, thatit's still the Wild West, which
(26:13):
when things are not regulated orthere is no oversight or really
strict policies in place, thatis definitely a prime
opportunity for crime and justmalicious uh activity.
SPEAKER_03 (26:31):
The unfortunate thing here is that this sets it
up for the uh mega banks to geta bigger lead over smaller
institutions like the unions.
Because Chase and Bank ofAmerica have so much money.
Many years ago, I talked withthe guy who was charged with uh
building the first mobilebanking app for Chase.
(26:54):
And he went out and he recruitedlike the Kansas City Chiefs of
ITEC.
I mean he had so much money tospend.
You know, this was we gotta doit, we're gonna do it.
Like they had a real short timeframe, they wanted it done.
And uh no credit union competewith can compete with that kind
of bank book.
(27:15):
It's uh and these guys, whenthey see a need and there's
urgency, they spend money,money, money, money because they
got it.
Right.
I don't see any credit unionhaving that kind of deep pocket,
unfortunately.
SPEAKER_01 (27:32):
Yeah, it it's uh it's the the nature of the beast
in general.
Um the bigger companies areputting money toward this and
preventative measures and tosafeguard and it may be a false
sense of security or maybe aright sense of security.
But if I'm a customer or amember, and probably one of the
most important things in my lifeis the safety of my money, and
(27:54):
I'm looking at it over time, I'mprobably gonna start gravitating
toward security than even thefriendly customer service.
Like, you know, the Sally knowsme really well.
That's that's important, but youknow, more important than that
is security, false or otherwise.
So these larger banks are all inon putting that extra layer of
(28:17):
protection, and that gives us asconsumers some degree of
comfort.
SPEAKER_03 (28:24):
And knowing Sally the teller and standing in
Sally's teller line is somethingthat your teenagers will never
understand and never do.
There might be an avatar namedSally that they're dealing with.
Yeah, but it's not quite thesame as knowing Sally and
knowing where she lives and howher kids are doing, and she
(28:44):
knows how your kids are doing.
SPEAKER_01 (28:46):
Yeah.
And that might that might changewhen you you become, you know,
you hit 30 and 40 years old, andyou kind of see this in that
that precious millennial agethat they're now in their 40s
now, that uh they're they wereso doted on, uh, and their their
mindsets were so I don't know,progressive and have a change.
(29:07):
But these folks become adultsand they they want things that
we now want as we we grew up aswell.
So I still think human touch isgonna be important,
authenticity, and uh that'sgonna be uh a critical
partnering.
As I mentioned, the pendulumswing to the one-on-one and
personal information.
I think that's gonna be needed.
(29:28):
What is gonna probably belacking is that that firewall of
protection uh for your for moneyor the perceived protection that
that might be the the gamechanger and what people would
will you to sacrifice than whatthey really need for the level
(29:49):
of uh comfort with when it comesto their money, if that makes
sense.
SPEAKER_03 (29:54):
Well, and you know the personal attention, yeah.
Chase puts A heck of a lot moreemphasis on building up its
personal banking portfolio thanit does getting retail banking
customers or opening an accountwith 50 bucks.
And but but Chase is all in onthat personal banker stuff.
And you don't have to be amillionaire to be uh Chase
(30:16):
personal banking customer.
That would help, but it's notit's not necessary.
So lesser amounts will do, butyou can't be bouncing your rent
check every month and have apersonal banker.
SPEAKER_01 (30:27):
So yeah, that at some point they're gonna red
flag that and say, hey, we wegotta we have a have kind of a
conversation.
SPEAKER_03 (30:34):
So yeah, I I I see the personal touch remaining,
and it continues to be somethingcredit unions can can uh succeed
at.
And I don't see AI taking thatover just yet.
How about using AI to do essays,for instance?
I know many people who now turnover most of their email to
Gmail, which has an AI function.
(30:56):
It'll answer the things for you.
It also will summarize them soyou don't even have to read them
and write an answer.
Now it gives you the answer, doyou you know, is this okay?
Yeah, sure.
I didn't read the email in thefirst place.
So yeah, it's okay.
I I don't know how I feel aboutall of that, but if I had high
volume email, I'd probably startusing it more than myself, to be
(31:18):
honest.
SPEAKER_02 (31:19):
But I don't have really high volume.
SPEAKER_01 (31:22):
Yeah, we uh we we struggle with uh a lot of
different aspects here of um youknow using computers and AI and
uh other aspects to makeinformed decisions.
Now it's it's no denying it, youknow, similar to what you
brought up in 1995, where it's atransition from what we know,
what we believe in, to then, oh,this actually thing does work
(31:44):
and it's actually veryconvenient and it's helpful.
And oh, I can actually purchasesomething.
I'm actually putting my creditcard on this thing and I'm
getting something.
Okay, there's a bit of trustgoing on, right?
That's where we are right nowwith AI.
It's that it's that trust factorthat's um somewhat lacking or
lurking.
Um, and that's that's okay, it'sgonna come with the territory,
but I I guarantee people a lotmore trustworthy in AI very,
(32:08):
very soon, or the this you know,new tool that everyone has.
But there's very much uhapprehension or even um what
else is what's what's around thecorner.
So again, I just want to cautionyou know your listeners here
that it's it's a new tool, don'ttrust it yet.
(32:29):
Um you know, use it cautiously,but knowing the rules of
engagement that it's noteverything, it's not uh it's not
anywhere near perfect and nonothingness, and there are
challenges that that lurk.
But you know, um validate uh andfind out, you know, get
different sources of informationbesides just simply chatting to
(32:52):
one thing and getting one sourceand that's that's it.
Do more work and uh find outwhat you know, be truthfinders
and the information you'reyou're seeking.
SPEAKER_03 (33:01):
Well, one thing I would always tell credit unions,
but I don't need to tell thembecause they do it anyway.
If in doubt, ask another creditunion.
If a credit union says, and Iget this question sometimes,
should we use such and such avendor?
I always say, Oh, why don't youfind a credit union that you
know and ask and they're usingthat vendor?
(33:24):
Ask them, don't ask me becausethey will if they will have more
valid experience.
And I I might think the CEO is agreat person, but that I'm not a
customer, so right.
And credit unions are reallygood at asking other credit
unions about about stuff, andcredit unions are really good
about sharing uh honest feedbackabout vendors.
(33:47):
Yeah, and that's that's that'sword of mouth, right?
If Chase calls up Bank ofAmerica and says, Hey, you know,
are you using such and suchvendor or how are they?
And let's say B of A thoughtthey were the worst vendor in
the world.
Oh, love them, man.
Love them.
Yeah, I'd if I were you, I'dturn over all your functions.
SPEAKER_01 (34:05):
Now that would be mischievous.
Yeah.
Fully endorse them uh falsely.
I like it.
SPEAKER_03 (34:11):
I mean, it's I've known bosses who have bad
employees that they're dying,which is they please leave,
please leave, and you get areference call.
How oh great! Oh, I'd hate tolose them, but sounds like a
good opportunity.
So but if a credit union comesup to you and says, okay, Casey,
I'm concerned.
(34:32):
What can you do for me?
What what exact what if you tellme you short form, what kind of
program would you set up?
Uh specifically for uh just ingeneral or um credit union
concerned about no, not just ingeneral.
SPEAKER_01 (34:47):
Yeah, I think there's a important tool that we
offer that uh is reallyunderutilized, but it's really
our reputation uh SWOT analysis.
And what that is, is to uh lookat what is your reputation of
your of your credit union.
(35:08):
So, and that's an unbiased lookat the products, the processes,
the people of the credit unioninside and outside the
organization, uh, because thebrand of the of the credit union
is what the credit union says byitself.
The reputation is what otherpeople say.
So as they bring us on, we golook underneath the hood to find
(35:32):
out what people are saying ornot saying about your credit
union or your competition.
So the SWAT that I speak about,Robert, is about like letting us
to see how good is yourreputation.
And it's it's being vulnerable,but it's looking to see what you
don't see and find out otherareas that you can shore up or
(35:54):
improve upon, or maybe evenaccentuate.
So the findings behind such aSWOT analysis, and this could be
surveys, this could be focusgroups, this could be online
reputation uh assessments, thiscould be uh interviewing
employees, past members, etcetera, et cetera, uh, with the
goal in mind is what are wetrying to accomplish here?
What is it that we we're lookingto find?
(36:15):
Um, that gives information tothe credit union to say, okay,
we need to improve our services.
We need to improve our, youknow, our products, our online
or our mobile app, whatever theyfind.
Now, there might be some lonewolves out there who are going
to cry out, we'll do we'lldiscard that, but we're looking
(36:38):
for themes here.
So that's a good point of entryfor a credit union to find out
what is it that they don't know.
I call it a reputation MRI.
A reputation MRI essentiallylet's find out where you're
healthy or where you're not.
SPEAKER_03 (36:52):
Yeah, how does AI figure into this?
SPEAKER_01 (36:55):
Well, AI figures into it as far as some of the
research that he goes into.
So if I, for instance, I putinto your into chat GPT, for
example, um, what do we knowabout the reputation of a uh of
a credit union?
It's a good starting point, butit's not the end point.
Um, AI can help at least theperception of it because that's
(37:15):
actually an excellent idea.
Every credit union should dothat.
SPEAKER_03 (37:18):
Yeah, right.
Chat GPT and look up Gemini.
You know, what's a broadreputation?
Yeah, look at Robert McCarty.
Oh, I've done it.
It's careless, man.
Don't do it.
Yeah, don't look, don't look meup, please.
SPEAKER_01 (37:35):
But that's that's kind of what uh if I've if I'm a
member, that's one of the firstthings I'm doing, right?
SPEAKER_03 (37:41):
If I am a uh if I'm an employee, in the old days,
people would look up the BetterBusiness Bureau.
That's right.
And if a credit union has a lotof dings in Better Business
Bureau, they got more problemsthan I can help them with.
Yeah, it's uh looking up newproducts.
You're just sitting at yourdesk, you type it in.
You know, what do you think ofNavy Federal?
(38:01):
And should I go there versusPentagon Federal?
I'm sure it has an answer tothat.
SPEAKER_01 (38:06):
So it's it's a starting point, but if you go a
little bit deeper, and that'slike, you know, if you're an
employee or uh prospectiveemployee, uh, I'm gonna say,
hey, is this a good place towork?
Right.
Uh you look at the glass door,right?
Or if you are looking at productcomparison, you can do your own
homework.
So part of this is your secretshopping this as a either
(38:28):
employee or a member or someonefrom the public.
What is that you know or don'tknow?
So our job is to do all thatresearch for them and provide a
unbiased third-party analysis ofwhat we found online, online,
uh, offline surveys, et cetera,to culminate.
Here's what we found.
Here's the theme that you shouldbe aware of.
(38:48):
Uh, don't be distracted by thoseglorious five-star ratings or uh
even the net promo promoterscores, those are um fool's
gold, I think.
Uh what you need to do is reallyfind out across different
sectors where are you what'sbeing said or not being said
about you and verse competition.
You know, competition isimportant as well.
(39:09):
Uh how do you compare andcontent uh contrast for other uh
organizations?
So all that's a good startingpoint to find out really what
your true value of yourreputation is or where there are
deficiencies.
SPEAKER_03 (39:24):
Then I assume you would help to kind of you know
create some guardrails.
Yes, using using uh AI.
SPEAKER_01 (39:32):
Oh yes.
Yeah, that's a that's a goodpoint because uh for AI
specifically, uh there are I'veseen a little more of an
adoption for this.
I don't I don't think too manyum companies are not adopting
that, but I think it's importantto say, all right, um, this is
this is a new important strongtool that there are guardrails.
(39:52):
Even I've seen a lot ofcompanies that are blocking the
use of Chat GPT or other AItools on a regular basis from
their computers.
People have to go home to useit, and that's that's okay.
But um if you do, uh this is onearea that's one industry, I
should say, that's been workingon this is the the legal
(40:13):
industry.
SPEAKER_03 (40:14):
They're reliant a little bit too much on gathering
facts and information about acase or details from a client
using these oh these these courtfilings that have references to
bogus non-existent previouscases.
This is this is horrifying.
SPEAKER_01 (40:33):
Yeah, so similarly, um, whatever you're looking at
for data and information, ifyou're a credit union employee,
or even, hey, I want to look upuh this is a prospective member,
what do I know about them?
Uh, it doesn't tell the wholestory, right?
And so use caution when usingsuch a thing because it doesn't
tell the whole story, or theremight be erroneous details about
(40:57):
them, or you know, other stuffthat you need to go a little bit
deeper, deeper on.
So using caution, havingpolicies and procedures in
place, and knowing that if youdo use it and you're basing
opinions and or your work on it,and we find out about it,
there's consequences, and one ofthem could be to be fired.
(41:20):
And you're seeing a lot more ofthat for employees that they're
using it falsely, or they'reputting uh information about the
member into the that that tool,and it's unfair.
So there's a lot of uh do's anddon'ts that are taking place
right now as guardrails toprotect everyone, but if it's
(41:42):
not used properly, then there'sconsequences mostly from the
employment side.
SPEAKER_03 (41:48):
Yeah, I I would just say as a rule that no credit
union employee should put inpersonally identifiable
information about members orco-workers or themselves into AI
under any circumstances.
I'm sure there's somecircumstances who would be fine,
but just to simplify matters andsay under any circumstances,
(42:09):
don't do it.
That to me is the most importantthing a credit union employee
can follow because if youviolate that, you file it,
violate something big time, man.
SPEAKER_01 (42:23):
So yeah, trust and reputation, and uh that's where
our world comes in play thatthis is this is serious stuff
now.
Uh one false you know step, andand that goes a long way.
And it was um the famous WarrenBuffett quote is it takes 25
years to build a reputation andfive minutes to ruin it.
(42:44):
If you think about that, you'lldo things differently.
And the idea here is that uhreputation is so sensitive, and
people are sensitive to it, thatuh we're forgiving nation, but
we're not a free uh forgettingnation.
And what I mean by that is thatyou know, if I'm gonna make a
decision between one fineinstitution and another, and one
(43:06):
of them has a even the smallestof black marks, that might tip
to scale for me to go somewhereelse.
Or if there's enough um uhchatter online, or oh, did you
see that happen about XYZ?
You know, look out.
Uh, one of the credit unions upin the Northern California is
Patelco.
Now they had a fiber attackhappen to them last year, and
(43:28):
they lost heaps of members, andprimarily because of how they
mishandled the information thatthey had or when they knew about
the the hack.
And that really pissed off a lotof members enough to say,
bye-bye, I'm gone, right?
So that's another key example ofnot being prepared, not handling
(43:54):
the situation well, andultimately hurting the bottom
line.
SPEAKER_03 (43:58):
Now, what was the uh reaction from readers to your CU
inside piece about AI assupervillain?
SPEAKER_01 (44:06):
Yeah, we got a lot of shares and some comments back
to us, and uh most folks aresaying this is something to
watch.
Uh I think it's new, Robert,that this is something that we
all kind of know that's it'shappening, but it's it's uh
raising an alarm bell that thisis a something to uh pay
(44:27):
attention to because we don'tknow what we don't know.
And while we're using it forhopefully all the right reasons
and fun reasons, and it's a newtoy, uh play with that toy with
caution or be vigilant thatthere's we don't necessarily
know all the issues behind it.
And so um that's one of thosethings where uh the cautionary
(44:49):
tale is the the the response Igot.
It's like this is good, thankyou for for sharing that.
SPEAKER_03 (44:54):
Well, I I totally agree that you know these tools
are extremely powerful, but wedon't know the full extent of
their powers, to be honest.
Therefore, we it's up to us tobe cautious about how we use
them.
And that's difficult.
People you know what is caughtwhat caught what what's cautious
(45:15):
to me isn't necessarily cautiousto you, but I I think we're at a
point where it's good to havethis kind of discussion.
And uh will we agree, notnecessarily, but I totally agree
with the concept that man, yougotta be thinking about these
issues.
You just can't plunge head uhhead first into this.
SPEAKER_01 (45:36):
It's uh yeah, and I I don't envy a a board of
directors or executiveleadership team of any company,
let alone the credit unions,that they have their own
heartburns going on right now,from deposits or lack of
deposits to uh moving target inthe financial world, uh, all
these different disruptors thatare happening.
(45:57):
Now you throw AI into the mix.
Oh wow, there's so much uh uhheadspace that's swirling uh
that's going on, obviously, fromtariffs to the financial
uncertainty.
There's just so much happeningright now, and AI is just one of
those important things.
Uh, I don't know how they do it.
SPEAKER_03 (46:16):
And it's uh it's tough to and this is the biggest
change in financialinstitutions, not since the
smartphone, not since the web,not since uh either of those
things.
It's actually the biggest changesince uh institutions, financial
institutions began tocomputerize when they moved uh
(46:37):
data off of pieces of paper ontofloppy disks or magnetic tape.
I mean, that's that's when thisuh that's how big this change
is.
I mean, this is and that thatchange happened in the mid-1980s
for most institutions.
We're at the same place now, andit's it's just an immense
change.
Now, could this blow up?
(46:58):
Yeah, I mean, but then again,the transition into computers
could have blown up too, and itdidn't.
SPEAKER_01 (47:03):
So yeah, it's just uh we we're all have to adjust,
right?
And that's the the the toughpart here.
And I don't necessarily know,you know, uh one of the areas is
jobs, Robert.
You know, where the jobs cancome from when these are this
disruptor of AI is taking overjobs, but uh every industry,
everything um, or every era Ishould say, has adjusted to the
(47:27):
times.
And I don't necessarily knowwhat the new jobs are going to
be.
Um I hope there's still room forhuman beings.
SPEAKER_03 (47:37):
Oh, there are, but and you know, if you actually
look at historic staffingnumbers and credit unions,
they're much lower for, let'ssay, million dollars in assets
than they were 20 years ago.
There's just fewer people.
Yeah, uh significantly fewer.
And I think that trend willcontinue.
Can we wipe out all departments?
(47:59):
Well, I hope not.
We can't all be sitting aroundletting machines do all our
work.
I at least I hope not.
SPEAKER_01 (48:08):
Or you know, you know, what's what's our value?
SPEAKER_03 (48:10):
You know, and then machines might decide amongst
themselves to get rid of us.
SPEAKER_01 (48:16):
Well, if if it if that's the world uh we had to
live in, then um I'll be thefirst in line to say goodbye.
SPEAKER_03 (48:22):
Hey man, you must have seen the matrix, right?
SPEAKER_01 (48:24):
I know I I still like human beings.
If uh human beings are beingtaken over by AI, then well,
that's my time to you know saygoodbye.
SPEAKER_03 (48:32):
Uh go go watch the matrix again tonight.
SPEAKER_01 (48:35):
No, I I prefer not to.
I won't watch uh maybe uh AnnieGriffith rerun or something.
SPEAKER_03 (48:43):
Before we go, think hard about how you can help
support this podcast so we cando more interviews with more
thoughtful leaders in the creditunion world.
What we're trying to figure outhere in these podcasts is what's
next for credit unit.
What can they do to really,really, really make a difference
in the financial team?
Can't all be megabanced, can it?
(49:05):
It's my hope it won't all bemega banked.
It'll always be a place forcredit unit.
That's what we're discussinghere.
To figure out how you can help,get in touch with me.
This is RJMGGarvey at gmail.com,Robert McGarvey, and that's
rjmegarvey at gmail.com.
We'll figure out a way that youcan help.
We need your support, we wantyour support, we thank you for
(49:27):
your support.
The CU2.0 podcast.
Welcome to the TU2.0 podcast.
SPEAKER_00 (00:05):
Hi, and welcome to the CU2.0 podcast with big new
ideas about credit unions andconversations about innovative
technology with credit union andfintech leaders.
This podcast is brought to youby Quillo, the real-time loan
syndication network for creditunions, and by your host,
longtime credit union andfinancial technology journalist
(00:27):
Robert McGarvey.
And now the CU 2.0 podcast withRobert McGarvey.
SPEAKER_03 (00:35):
AI is the new CU Crisis Supervillain.
That's the title of a recent CUInsight Story authored by Tasty
Box, founder of Reputation U.S.
And of course we had to get himon the podcast.
Understand I'm a strongsupporter of AI in general.
And AI in particular insidecritics of this is a life and
(00:59):
death matter.
Yes.
Yes, there is a possibledownside to AI.
We already know that the main AIis falsified substantial team
suicides and false facts andmany other ways of leading
humans of strife.
Well, I may disagree with manyof boxes, conclusions, of
(01:20):
course, but see what AIcostously and thoughtfully is on
the mind.
All the fools are too read boxesto you insight, please just a
link to that in the show notes.
Listen to the podcast.
Read a read this to you insidestory about me or the force to
use the way I was to you to theold founder portrait is a link
(01:43):
to that story in the show notes.
And just keep on learning andexperimenting and using AI.
It's the future.
Learn how to use itintelligently and wisely.
Now, what kinds of AI disastersor mini disasters or acts of
bond villainy are do youanticipate seeing?
SPEAKER_01 (02:07):
Yeah, anticipate and already seeing, I would say
probably in three to fourbuckets here.
One is just simply thecybersecurity and data privacy
is a big piece of this.
So the world of cybersecurityhas been going on for quite some
time, right?
And there's been no shortage ofan industry that has been there
to stave off these bad actors,as they call them, or malicious
(02:31):
actors that are trying toessentially come in and take
data and ransom your informationand try to make money, right?
So similarly, you know, thisemerging AI is really uh
mishandling member data.
And you know, they'remanipulating the information so
uh that they're just simplytrying to uh gather something of
(02:54):
value from the credit union thatis you know that at some point
the credit union has to wantthat back.
And so using AI in differentforms to uh penetrate those uh
those the member data or uhfinancials, that's where we're
seeing probably the mostpenetration early on with AI.
(03:15):
And then another subset to alittle bit of that is deep
fakes.
And to what degree the word deepfake is essentially just a way
to falsify uh a person's voiceor image, and we're seeing more
of this just to falsify when weget an email or a voicemail or
(03:37):
sometimes even video of a personthat pretends to be someone from
a credit union saying, hey, uhgive us a call, give us your
information, your socialsecurity number, your uh
financial information, and thatI haven't seen it as much, but
it is still happening, and thatwill essentially cure losses for
the members because they thinkit's someone from the credit
(03:59):
union, right?
And they're getting they'regoing after the vulnerable
folks.
The um the bad actors are goingafter the folks who might be
vulnerable or um I don't knownaive or just simply um, I guess
crazy enough to give theirinformation to someone who
seemingly is the uh executive ormember services person that they
(04:20):
know, trust, and like.
So that's uh that's one we'reseeing as well.
Another one in all this islending discrimination.
And that one is about it reallyfits in not only credit unions
but banks as well and just otherfolks who are in the risk
business area.
But AI models um you know reallythey look for uh these uh loans
(04:42):
or impose unfavorable terms onprotected groups, and uh with
that becomes uh there's scrutinybehind it because they are uh
being biased or there'sdiscriminating against members
of a certain class ormarginalized group.
And that's a big no-no.
And uh and that happens to besomething that not only from a
(05:06):
reputation side, but certainlyfrom a regulatory and even
lawsuit-wise, could be uh badfor a credit union.
SPEAKER_03 (05:17):
Now, obviously, discrimination lending was
invented by human beings who tosome extent perfected that until
some federal laws came along andsaid, no, no, you can't do that.
It's uh I don't see why themachine can't be taught that it
can't do it in the same way thatbanks and credit unions have
(05:38):
taught their lending staff thatno, you you can't make decisions
based on race, race, age, blah,blah, blah.
It's uh um that shouldn't bethat difficult a thing to teach
the machine.
SPEAKER_01 (05:52):
You would hope not.
You would hope not.
I think it still applies becausethis is when human uh oversight
is very important.
Uh so like a checks and bounces.
So if you are crediting andusing AI for the first degree of
um determining if someone is arisky or non-risky lender or you
(06:13):
know uh client, uh, that's onepart.
But if you're not necessarilygoing underneath the hood to
find out more information aboutthis person, uh and you're
taking the AI recommendationsfor what it's worth, then that's
on you.
That's on the credit union fornot doing its due diligence.
But part of the the misstepsthat are actually happening here
(06:36):
is you know, humans arebecoming, dare I say, lazy or
don't do their due diligence andusing the things that are
afforded to them, like you know,AI tools, uh and um the checks
and balances of this um are injeopardy, which essentially,
again, on the fault of thecredit union or the the lending
(07:00):
bank.
SPEAKER_03 (07:02):
Yeah, that that said, we use machines to make a
lot of credit decisions.
I mean, I applied for a newcredit card uh a week or so ago.
And within a minute, I wasapproved.
Um how did they approve that?
They got my FICO score, period.
I mean, there's there's no otherway they could have gotten any
(07:24):
information about me that wasuseful in that time frame, other
than uh a FICO score.
And they said, oh wow, we'llgive this guy a credit card, of
course.
SPEAKER_01 (07:33):
Right.
I think it might be I thinkthat's you know, that's credit
is one part, and then the otheraspect is just lending money,
you know, you know, the equityline of credit, you know,
lending money for a car loan,that type of thing.
Uh for just a credit card, Ithink that is automated based
off what you just said.
Uh you know, numbers don't lie.
If they have you know FICO scorethat's you know in the 700s or
(07:56):
above, that immediatelyqualifies you to take out a
credit card.
SPEAKER_03 (08:01):
But I think it gets a little more sophisticated when
it comes to well, also like thecredit card, you might have a$10
or$20,000 credit limit, but it'shighly unusual to take that card
the day you receive it and runout and put the maximum amount
out.
I'm sure Kirks would do that,but that's not normal.
And if you're missing paymentsalong the way, at some point
(08:22):
they pull a plug on it.
That's right.
And they're not gonna lose thewhole amount in most cases.
So, I mean, yeah, creditlending, this has been going on,
credit cards have been going onfor at least geez, I don't know,
60 years.
They've gotten very good atthis.
And uh a lot of it's machinelearning, though, that's
actually pretty good.
(08:42):
I mean, I give the example thatif I walked into an Apple store
in Detroit, Michigan, and boughtthree super expensive um uh
Apple computers, that chargewould be declined uh by my Apple
card, which would say Bozo'snever been to Detroit.
He don't know where Detroit is.
(09:06):
And furthermore, he's neverbought this kind of computer, he
doesn't seem to like it.
Uh and that would be a smartmove on that that machine's
part, saying, nope, decline.
There is there are places wherethe the the learning makes
sense, you know, to use themachine learning.
That's right.
And you know, an issue thatinterests me is that let's say
(09:30):
you apply for uh uh$75,000 homeequity loan, and I'm looking at
your application, and I say,geez, I don't know if I should
say yes or no, but I'm a smartguy.
So I take your name out of it,but I put a lot of other
information into Chat GPT andsay, should I give a should we
give a loan to this guy?
(09:52):
Question What happens to thatdata?
Now, Google says we will nevertrain the system on your data.
Your data is your data.
That's what Google says withwith uh uh Gemini.
ChatGPT, there's a toggle youcan set which says, Don't use my
data to train.
Can you rely on ChatGPT?
(10:15):
Well, chat GPT has made someegregious mistakes in recent
days, recent weeks.
Sure.
Advising uh maladjusted teenageboys about how to commit
suicide, and some do at least.
SPEAKER_01 (10:27):
So right.
It's a whole nother crisis forsure.
Um because not just corporateparts, but there's a lot of AI
issues uh surrounding thosematters.
Yeah.
SPEAKER_03 (10:37):
Well, so that that that raises the question of can
we trust their word?
Can we trust their ethics, whichis uh scary stuff to bring up,
but in the case of the teenagesuicides and the advice, I have
advice and and quote marks thathave been given by by the
machine.
You say, wow, this is not thisis not good advice.
(10:59):
Yeah, you don't give a13-year-old who's depressed tips
on how to commit suicide.
One doesn't do that.
SPEAKER_01 (11:04):
Or in general, you know, I'm a father of twins, uh,
teenagers, and not my teenagerper se, but uh in general, uh
kids are looking for importantcounsel on where to go to
college, uh what to do incertain circumstances.
Uh, when we as a parent were thethat sanding board based on our
experience, now they are lookingto these you know AI tools for
(11:29):
pretty much everything, personalor otherwise.
And that gets very uhchallenging because it looks
good, it seems logical, butdoesn't necessarily fit into
what they know about the humanbeing, uh, that teenager.
Uh and so uh information isbeing shared while it seems
convincing and well thought out,or at least it is uh curated to
(11:52):
something that uh bite-sized forthe teenagers, it doesn't factor
in that teenager themselves.
There's so many differentvariables that fit into that.
Um, and it's you know, to relyon AI to provide those insights
is so limiting.
Very much still, I don't carehow much information you put
into AI, it just cannot uh gointo those nuances of uh the
(12:14):
human capital and the uniquenessthat every person is involved,
including like you know, go backto the credit union line.
If you're if you're providing uhlending to someone, there's a
story behind those people,right?
They might not necessarily havethe best credit for whatever
reason, but they're doing otherthings like you know, looking
for work, they are doing four orfive different jobs, they just
(12:38):
had you know uh some otherissues that prevented them from
having a good score.
So credit union and knowing alittle bit more about that human
being on a one-on-one basis,they'll say, Yeah, uh I get
their circumstances, I'm willingto give them this this loan.
Uh, if you based purely on justthe numbers or some other
factors, then the biases willwill surface.
SPEAKER_03 (13:02):
Well, that's you're talking about the old credit
union tradition.
I know when Jim Blaine was CEOof state employees at the Credit
Union of North Carolina, headamantly insisted they would
lend to the human being, not toa piece of paper.
And yeah, if you'd had uh ahealth crisis and you miss
(13:23):
missed paying a few bills hereand there, that would be taken
into account.
And you might get the the moneyto buy a car that would let you
get work drive to your job.
You might get that from St.
Employees because the judgmentwould be based upon the human,
not the fact that you had a runa little run of bad luck.
That's right.
A lot of credit unions havemoved off of that,
unfortunately.
(13:44):
But what can I say?
I don't run a lending committeeon a credit union.
So yeah, and it it's harder tofactor the human stuff in.
You have to talk to people, youhave to call for the person's
foreman, like in the old days ofthe select employee groups.
You know, tell me about them.
Is it reliable?
Well, you know, that kind ofstuff.
SPEAKER_01 (14:03):
So well, I do think that and one of the areas that
we brought up as well that yeah,the the the chatbots, you know,
are efficient, right?
But members um are oftenperceived them as cold.
And I don't know about you, butif I call up uh a company and
it's always simply get as anautomation or even a an online
(14:26):
version to give me the answers,it's really impersonal, or there
it's limiting, or it's got youknow air written all over it.
And if you are perceived as aninstitution that relies solely
on AI-driven chatbots, you areimpersonal, right?
And the member might resent you,or might oh, I can get this
(14:48):
anywhere, where the pendulum isgonna hopefully swing from AI to
people who are authentic.
That's where I think uh there'sgonna be more of an appetite.
As I I'm you know, I'm guessinghere is purely on opinion, but I
think at some point we're like,all right, this is ridiculous.
I want to talk to a human being,I want to go into the branch, I
(15:10):
want to go and meet thesepersons because there's too much
at risk here.
And if they're only going to bea lender or a relationship
manager purely based off ofnumbers and using AI just to
make money, then that's notgoing to be my financial
solution.
I need a human being.
SPEAKER_03 (15:28):
There's debate within the credit union world as
to what are the best tasks tofirst pursue with AI.
And to me, the winning argumentis to use AI on not on customer,
not on member-facing stuff, butto use it on back office stuff.
The most interesting example Iknow of a successful AI project
(15:52):
is that one Navada credit union.
They took all of the informationthat call center people consult
when you call up with aquestion, fed it all into AI,
aiming to simplify simplify.
There were many severaldocuments often dealt with the
same topic, sometimes gavecontradictory advice.
(16:12):
I mean, this is informationcollected over 30 or 40 years.
I mean, I'm I'm sure Icontradict myself over things I
said and wrote 20 years ago.
So they cleaned all of that upand uh have uh resulted in
terrific uh time savings.
So instead of an employeesitting there flipping through
(16:33):
pages, getting confused becausethey're contradictions, you just
ask the machine when will werepossess this guy's car if he
misses payments?
Within 10 seconds, you got ananswer.
So the the member's not sittingon the phone for three minutes
as you dial around.
Boom, answer.
(16:54):
So does the member know that AIhas played a role in this?
No, why should they?
It's irrelevant.
You're you're still talking witha person.
I mean, you don't I assume youdon't have any issues about
those kind of projects.
SPEAKER_01 (17:06):
Not yet.
And I and I say that because uhthose probably aren't as exposed
because if it is it is an erroruh that's going to be tracked
probably internally, and I can'timagine any institution is going
to track and tout some of theerrors they've done internally,
or they're they're seeing someflaws in that system.
(17:27):
They might you know uh talk tothe vendor who they're using or
the AI aspect that they'rethey're using and saying, hey,
this this does have flaws, but Idon't think it's one of those
things where they're going to uhuh highlight some of the you
know inefficiencies that aretaking place that's gonna be
forward-thinking to you know thepublic or uh prospective
members.
SPEAKER_03 (17:46):
So I know another place where AI is having
significant impact at creditunions is automating
collections.
So if you're late, instead of ahuman being calling up, leaving
a message, because you're notgonna answer the phone, sure, a
machine will call up and leavethe message.
And one of the most interestingthings about that is there so
(18:10):
far I have not heard of anycollection staff complaining
about this because usually mostof the people doing that find it
to be um work they don't reallywant to do.
And these credit unions have nothad any staff reduction.
So they've given people othertasks to do.
So it's not like, oh, we'regonna fire you and replace you
with a machine.
It's uh, hey, you don't have tomake those collection calls
(18:32):
anymore.
How's that?
You know, wow, this is great.
I came in, I had to eat Tums allmorning before I made my first
call.
No, this is wonderful.
Thank you.
God, I love AI.
SPEAKER_01 (18:48):
There's there are some good, yeah, there are some
good parts of AI for sure.
And uh it's not primarily topoo-poo on it.
In fact, uh it's gonna be animportant uh engine to different
aspects, you know, andaugmentation to uh things that
we're already doing and takingtaking the lead on.
There's no shortage of examplesto that end, uh, but it's it's
(19:10):
not where it needs to be at all.
And I'm not sure if it'll everbe, but in the same vein,
because of something that's new,there's going to be malfeasans,
nefarious actors out there, uh,and folks who are essentially
looking to game the system.
And that, you know, based onwhat we've seen so far, that's
(19:30):
already happening.
And it's gonna probably be moresophisticated and it's kind of
scary.
And so part of what we'retalking about today, about you
know the AI uh crisis of sorts,is that you need to be vigilant.
You know, what's going on rightnow in September 17th, 2025, uh
is likely gonna be a lot moredifferent than September 17th,
(19:53):
2026.
And uh it's it's it's scary,daunting.
Um, it's going to be a littlebit exciting, but we all have to
be participants and beingvigilant and being ready because
we don't necessarily know whatwe're gonna be um up against.
Right.
(20:13):
Yeah.
Crisis prep is gonna beimportant and not just for
preparation parts, and that'swhat we advocate for, but I
think just you have to geteducated to it.
It's one of those things wherewhen I talk to someone about
what is your degree of uhknowledge of AI, adoption, and
other things, it's still a slowburn for a lot of people, and
that's okay.
(20:34):
Uh, but for companies that aredealing with records, uh medical
records or financial records orjust information or there are in
the business of customerservice, this is has to be
paramount because there's a lotat stake here.
And in one little slip or uhmishandling of how you uh
(20:56):
handled the situation, it'sgoing to be a dark mark on on
you uh as a company.
And that's what we're justtelling folks and scream from
the rooftops now is that it itis here.
Keep an eye out for it, but notjust simply like waiting for the
crisis to happen, but having youknow safeguards in place.
(21:17):
And that's essentially havingalmost a team ready for being
vigilant, uh, ready to prepare,and to hopefully to mitigate
against any of the damage thatmight come your way.
SPEAKER_03 (21:28):
Well, go back to 1995 when every company's
rushing to be online.
They pretty quickly realize thatI I don't want you to take your
personal computer go down tolocal Starbucks, use their their
Wi-Fi or whatever they had atthat point, to get online and
(21:52):
dial into my mainframe and startdownloading stuff.
I just don't want this tohappen.
And they implemented ways tomake that not happen.
For instance, you have to use aVPN, you can't use a personal
computer, you know, stuff likethat.
A whole bunch of steps designedto keep the connection and thus
the data secure.
(22:13):
I I see companies needing, Ithink they need help, but I I
just see companies rushing tofigure out how, okay, we're
gonna use AI, how can we do itmore securely?
Um and I think you're right thatthe answers to that question
aren't that clear right now.
It's we're so busy getting thesystem up and working.
(22:33):
Uh security, we'll worry aboutthat later, man.
I mean, who worries about thatin Silicon Valley at the
beginning?
You know, we worry about itlater.
SPEAKER_01 (22:44):
Yeah.
It's uh you know, sellingpreventative medicines uh a
tough sell in general.
Uh we just need to get up and uhget up and ready.
And if uh something goessideways, then we'll go
sideways, we'll deal with theramifications.
Uh however, those are costlymaneuvers because there's a lot
at risk here.
That's the challenge on so manydifferent aspects.
(23:05):
Um, a couple of differentthings, Robert, you know, like
uh about being prepared or not,that um, you know, I have a few
stats that we have on ourwebsite about AI, but according
to this Oxford uh metrica, uhcompanies that handle crises
poorly lose 15 to 30 percent oftheir market value within days.
(23:26):
So, you know, that's a that's alot.
So, you know, the whole idea,let's we'll just handle it when
it comes in.
Well, if uh you're you're gonnalose 15 to 30 percent of your
market value when somethinghappens, uh, are you okay with
that?
You know, so that's the the thebell we're ringing here.
That preparation doesn't reallycost you much, but a crisis
(23:47):
costs you a ton.
It's really important for uhorganizations, AI or otherwise,
to be prepared.
And I'm just shocked that mostaren't.
Yes, they might in the creditunion world, yes, they might
have a um, you know, some in youin C UA um disaster recovery
plan in place, but that's morelines of a checklist they have
(24:10):
to do, but no one really knowswhere it is or what it's all
about.
But to actually practice on ituh and be prepared for it, um
that's the the key element thatwe find daunting because we're
called in the in the uh 11thhour or we parachute in during a
crisis and like, okay, let'sjust take care of the situation.
(24:31):
Hopefully we can mitigateagainst any damage.
But um, at that point, we're thefirefighters putting out the
fire as opposed to we could havereally uh saved a lot of the
embarrassments, uh a lot of thedata and materials if we were
just prepared.
And I'm not just saying it'sjust because it's a part of our
job, it's just real, it's justsad that um most organizations
(24:52):
aren't prepared for a crisis.
For some reason, they buyinsurance for things that are
that might happen, but uh forreputation, which actually, of
all things, reputation is aninsurable asset, they are
they're willing to roll the diceon um the that kind of damage,
and that's proven a bad move.
SPEAKER_03 (25:13):
To my knowledge, uh NCUA hasn't issued meaningful
documents regarding AI usage andcredit unions.
I might be wrong on that, butI've been talking to so many
people recently so often aboutNCUA.
I don't think I am.
SPEAKER_01 (25:29):
Well, you're you're right.
Um uh in fact, I was just justbefore our meeting today, I was
on a AI conference call withBeasley's, which is a you know
insurance carrier, and they wereunderscoring that there aren't
much regulations right now atall.
Uh in the Europe and the EU,they have quite a few already,
(25:51):
which is good.
There's some oversight, but inAmerica, there's not much.
And then if you get a littlemore granular, like a credit
union industry, there's not muchat all.
I can't speak specifically of uhevery industry, but I think it's
still so nascent and uh probablystill up for debate to what
degree it can be regulated, thatit's still the Wild West, which
(26:13):
when things are not regulated orthere is no oversight or really
strict policies in place, thatis definitely a prime
opportunity for crime and justmalicious uh activity.
SPEAKER_03 (26:31):
The unfortunate thing here is that this sets it
up for the uh mega banks to geta bigger lead over smaller
institutions like the unions.
Because Chase and Bank ofAmerica have so much money.
Many years ago, I talked withthe guy who was charged with uh
building the first mobilebanking app for Chase.
(26:54):
And he went out and he recruitedlike the Kansas City Chiefs of
ITEC.
I mean he had so much money tospend.
You know, this was we gotta doit, we're gonna do it.
Like they had a real short timeframe, they wanted it done.
And uh no credit union competewith can compete with that kind
of bank book.
(27:15):
It's uh and these guys, whenthey see a need and there's
urgency, they spend money,money, money, money because they
got it.
Right.
I don't see any credit unionhaving that kind of deep pocket,
unfortunately.
SPEAKER_01 (27:32):
Yeah, it it's uh it's the the nature of the beast
in general.
Um the bigger companies areputting money toward this and
preventative measures and tosafeguard and it may be a false
sense of security or maybe aright sense of security.
But if I'm a customer or amember, and probably one of the
most important things in my lifeis the safety of my money, and
(27:54):
I'm looking at it over time, I'mprobably gonna start gravitating
toward security than even thefriendly customer service.
Like, you know, the Sally knowsme really well.
That's that's important, but youknow, more important than that
is security, false or otherwise.
So these larger banks are all inon putting that extra layer of
(28:17):
protection, and that gives us asconsumers some degree of
comfort.
SPEAKER_03 (28:24):
And knowing Sally the teller and standing in
Sally's teller line is somethingthat your teenagers will never
understand and never do.
There might be an avatar namedSally that they're dealing with.
Yeah, but it's not quite thesame as knowing Sally and
knowing where she lives and howher kids are doing, and she
(28:44):
knows how your kids are doing.
SPEAKER_01 (28:46):
Yeah.
And that might that might changewhen you you become, you know,
you hit 30 and 40 years old, andyou kind of see this in that
that precious millennial agethat they're now in their 40s
now, that uh they're they wereso doted on, uh, and their their
mindsets were so I don't know,progressive and have a change.
(29:07):
But these folks become adultsand they they want things that
we now want as we we grew up aswell.
So I still think human touch isgonna be important,
authenticity, and uh that'sgonna be uh a critical
partnering.
As I mentioned, the pendulumswing to the one-on-one and
personal information.
I think that's gonna be needed.
(29:28):
What is gonna probably belacking is that that firewall of
protection uh for your for moneyor the perceived protection that
that might be the the gamechanger and what people would
will you to sacrifice than whatthey really need for the level
(29:49):
of uh comfort with when it comesto their money, if that makes
sense.
SPEAKER_03 (29:54):
Well, and you know the personal attention, yeah.
Chase puts A heck of a lot moreemphasis on building up its
personal banking portfolio thanit does getting retail banking
customers or opening an accountwith 50 bucks.
And but but Chase is all in onthat personal banker stuff.
And you don't have to be amillionaire to be uh Chase
(30:16):
personal banking customer.
That would help, but it's notit's not necessary.
So lesser amounts will do, butyou can't be bouncing your rent
check every month and have apersonal banker.
SPEAKER_01 (30:27):
So yeah, that at some point they're gonna red
flag that and say, hey, we wegotta we have a have kind of a
conversation.
SPEAKER_03 (30:34):
So yeah, I I I see the personal touch remaining,
and it continues to be somethingcredit unions can can uh succeed
at.
And I don't see AI taking thatover just yet.
How about using AI to do essays,for instance?
I know many people who now turnover most of their email to
Gmail, which has an AI function.
(30:56):
It'll answer the things for you.
It also will summarize them soyou don't even have to read them
and write an answer.
Now it gives you the answer, doyou you know, is this okay?
Yeah, sure.
I didn't read the email in thefirst place.
So yeah, it's okay.
I I don't know how I feel aboutall of that, but if I had high
volume email, I'd probably startusing it more than myself, to be
(31:18):
honest.
SPEAKER_02 (31:19):
But I don't have really high volume.
SPEAKER_01 (31:22):
Yeah, we uh we we struggle with uh a lot of
different aspects here of um youknow using computers and AI and
uh other aspects to makeinformed decisions.
Now it's it's no denying it, youknow, similar to what you
brought up in 1995, where it's atransition from what we know,
what we believe in, to then, oh,this actually thing does work
(31:44):
and it's actually veryconvenient and it's helpful.
And oh, I can actually purchasesomething.
I'm actually putting my creditcard on this thing and I'm
getting something.
Okay, there's a bit of trustgoing on, right?
That's where we are right nowwith AI.
It's that it's that trust factorthat's um somewhat lacking or
lurking.
Um, and that's that's okay, it'sgonna come with the territory,
but I I guarantee people a lotmore trustworthy in AI very,
(32:08):
very soon, or the this you know,new tool that everyone has.
But there's very much uhapprehension or even um what
else is what's what's around thecorner.
So again, I just want to cautionyou know your listeners here
that it's it's a new tool, don'ttrust it yet.
(32:29):
Um you know, use it cautiously,but knowing the rules of
engagement that it's noteverything, it's not uh it's not
anywhere near perfect and nonothingness, and there are
challenges that that lurk.
But you know, um validate uh andfind out, you know, get
different sources of informationbesides just simply chatting to
(32:52):
one thing and getting one sourceand that's that's it.
Do more work and uh find outwhat you know, be truthfinders
and the information you'reyou're seeking.
SPEAKER_03 (33:01):
Well, one thing I would always tell credit unions,
but I don't need to tell thembecause they do it anyway.
If in doubt, ask another creditunion.
If a credit union says, and Iget this question sometimes,
should we use such and such avendor?
I always say, Oh, why don't youfind a credit union that you
know and ask and they're usingthat vendor?
(33:24):
Ask them, don't ask me becausethey will if they will have more
valid experience.
And I I might think the CEO is agreat person, but that I'm not a
customer, so right.
And credit unions are reallygood at asking other credit
unions about about stuff, andcredit unions are really good
about sharing uh honest feedbackabout vendors.
(33:47):
Yeah, and that's that's that'sword of mouth, right?
If Chase calls up Bank ofAmerica and says, Hey, you know,
are you using such and suchvendor or how are they?
And let's say B of A thoughtthey were the worst vendor in
the world.
Oh, love them, man.
Love them.
Yeah, I'd if I were you, I'dturn over all your functions.
SPEAKER_01 (34:05):
Now that would be mischievous.
Yeah.
Fully endorse them uh falsely.
I like it.
SPEAKER_03 (34:11):
I mean, it's I've known bosses who have bad
employees that they're dying,which is they please leave,
please leave, and you get areference call.
How oh great! Oh, I'd hate tolose them, but sounds like a
good opportunity.
So but if a credit union comesup to you and says, okay, Casey,
I'm concerned.
(34:32):
What can you do for me?
What what exact what if you tellme you short form, what kind of
program would you set up?
Uh specifically for uh just ingeneral or um credit union
concerned about no, not just ingeneral.
SPEAKER_01 (34:47):
Yeah, I think there's a important tool that we
offer that uh is reallyunderutilized, but it's really
our reputation uh SWOT analysis.
And what that is, is to uh lookat what is your reputation of
your of your credit union.
(35:08):
So, and that's an unbiased lookat the products, the processes,
the people of the credit unioninside and outside the
organization, uh, because thebrand of the of the credit union
is what the credit union says byitself.
The reputation is what otherpeople say.
So as they bring us on, we golook underneath the hood to find
(35:32):
out what people are saying ornot saying about your credit
union or your competition.
So the SWAT that I speak about,Robert, is about like letting us
to see how good is yourreputation.
And it's it's being vulnerable,but it's looking to see what you
don't see and find out otherareas that you can shore up or
(35:54):
improve upon, or maybe evenaccentuate.
So the findings behind such aSWOT analysis, and this could be
surveys, this could be focusgroups, this could be online
reputation uh assessments, thiscould be uh interviewing
employees, past members, etcetera, et cetera, uh, with the
goal in mind is what are wetrying to accomplish here?
What is it that we we're lookingto find?
(36:15):
Um, that gives information tothe credit union to say, okay,
we need to improve our services.
We need to improve our, youknow, our products, our online
or our mobile app, whatever theyfind.
Now, there might be some lonewolves out there who are going
to cry out, we'll do we'lldiscard that, but we're looking
(36:38):
for themes here.
So that's a good point of entryfor a credit union to find out
what is it that they don't know.
I call it a reputation MRI.
A reputation MRI essentiallylet's find out where you're
healthy or where you're not.
SPEAKER_03 (36:52):
Yeah, how does AI figure into this?
SPEAKER_01 (36:55):
Well, AI figures into it as far as some of the
research that he goes into.
So if I, for instance, I putinto your into chat GPT, for
example, um, what do we knowabout the reputation of a uh of
a credit union?
It's a good starting point, butit's not the end point.
Um, AI can help at least theperception of it because that's
(37:15):
actually an excellent idea.
Every credit union should dothat.
SPEAKER_03 (37:18):
Yeah, right.
Chat GPT and look up Gemini.
You know, what's a broadreputation?
Yeah, look at Robert McCarty.
Oh, I've done it.
It's careless, man.
Don't do it.
Yeah, don't look, don't look meup, please.
SPEAKER_01 (37:35):
But that's that's kind of what uh if I've if I'm a
member, that's one of the firstthings I'm doing, right?
SPEAKER_03 (37:41):
If I am a uh if I'm an employee, in the old days,
people would look up the BetterBusiness Bureau.
That's right.
And if a credit union has a lotof dings in Better Business
Bureau, they got more problemsthan I can help them with.
Yeah, it's uh looking up newproducts.
You're just sitting at yourdesk, you type it in.
You know, what do you think ofNavy Federal?
(38:01):
And should I go there versusPentagon Federal?
I'm sure it has an answer tothat.
SPEAKER_01 (38:06):
So it's it's a starting point, but if you go a
little bit deeper, and that'slike, you know, if you're an
employee or uh prospectiveemployee, uh, I'm gonna say,
hey, is this a good place towork?
Right.
Uh you look at the glass door,right?
Or if you are looking at productcomparison, you can do your own
homework.
So part of this is your secretshopping this as a either
(38:28):
employee or a member or someonefrom the public.
What is that you know or don'tknow?
So our job is to do all thatresearch for them and provide a
unbiased third-party analysis ofwhat we found online, online,
uh, offline surveys, et cetera,to culminate.
Here's what we found.
Here's the theme that you shouldbe aware of.
(38:48):
Uh, don't be distracted by thoseglorious five-star ratings or uh
even the net promo promoterscores, those are um fool's
gold, I think.
Uh what you need to do is reallyfind out across different
sectors where are you what'sbeing said or not being said
about you and verse competition.
You know, competition isimportant as well.
(39:09):
Uh how do you compare andcontent uh contrast for other uh
organizations?
So all that's a good startingpoint to find out really what
your true value of yourreputation is or where there are
deficiencies.
SPEAKER_03 (39:24):
Then I assume you would help to kind of you know
create some guardrails.
Yes, using using uh AI.
SPEAKER_01 (39:32):
Oh yes.
Yeah, that's a that's a goodpoint because uh for AI
specifically, uh there are I'veseen a little more of an
adoption for this.
I don't I don't think too manyum companies are not adopting
that, but I think it's importantto say, all right, um, this is
this is a new important strongtool that there are guardrails.
(39:52):
Even I've seen a lot ofcompanies that are blocking the
use of Chat GPT or other AItools on a regular basis from
their computers.
People have to go home to useit, and that's that's okay.
But um if you do, uh this is onearea that's one industry, I
should say, that's been workingon this is the the legal
(40:13):
industry.
SPEAKER_03 (40:14):
They're reliant a little bit too much on gathering
facts and information about acase or details from a client
using these oh these these courtfilings that have references to
bogus non-existent previouscases.
This is this is horrifying.
SPEAKER_01 (40:33):
Yeah, so similarly, um, whatever you're looking at
for data and information, ifyou're a credit union employee,
or even, hey, I want to look upuh this is a prospective member,
what do I know about them?
Uh, it doesn't tell the wholestory, right?
And so use caution when usingsuch a thing because it doesn't
tell the whole story, or theremight be erroneous details about
(40:57):
them, or you know, other stuffthat you need to go a little bit
deeper, deeper on.
So using caution, havingpolicies and procedures in
place, and knowing that if youdo use it and you're basing
opinions and or your work on it,and we find out about it,
there's consequences, and one ofthem could be to be fired.
(41:20):
And you're seeing a lot more ofthat for employees that they're
using it falsely, or they'reputting uh information about the
member into the that that tool,and it's unfair.
So there's a lot of uh do's anddon'ts that are taking place
right now as guardrails toprotect everyone, but if it's
(41:42):
not used properly, then there'sconsequences mostly from the
employment side.
SPEAKER_03 (41:48):
Yeah, I I would just say as a rule that no credit
union employee should put inpersonally identifiable
information about members orco-workers or themselves into AI
under any circumstances.
I'm sure there's somecircumstances who would be fine,
but just to simplify matters andsay under any circumstances,
(42:09):
don't do it.
That to me is the most importantthing a credit union employee
can follow because if youviolate that, you file it,
violate something big time, man.
SPEAKER_01 (42:23):
So yeah, trust and reputation, and uh that's where
our world comes in play thatthis is this is serious stuff
now.
Uh one false you know step, andand that goes a long way.
And it was um the famous WarrenBuffett quote is it takes 25
years to build a reputation andfive minutes to ruin it.
(42:44):
If you think about that, you'lldo things differently.
And the idea here is that uhreputation is so sensitive, and
people are sensitive to it, thatuh we're forgiving nation, but
we're not a free uh forgettingnation.
And what I mean by that is thatyou know, if I'm gonna make a
decision between one fineinstitution and another, and one
(43:06):
of them has a even the smallestof black marks, that might tip
to scale for me to go somewhereelse.
Or if there's enough um uhchatter online, or oh, did you
see that happen about XYZ?
You know, look out.
Uh, one of the credit unions upin the Northern California is
Patelco.
Now they had a fiber attackhappen to them last year, and
(43:28):
they lost heaps of members, andprimarily because of how they
mishandled the information thatthey had or when they knew about
the the hack.
And that really pissed off a lotof members enough to say,
bye-bye, I'm gone, right?
So that's another key example ofnot being prepared, not handling
(43:54):
the situation well, andultimately hurting the bottom
line.
SPEAKER_03 (43:58):
Now, what was the uh reaction from readers to your CU
inside piece about AI assupervillain?
SPEAKER_01 (44:06):
Yeah, we got a lot of shares and some comments back
to us, and uh most folks aresaying this is something to
watch.
Uh I think it's new, Robert,that this is something that we
all kind of know that's it'shappening, but it's it's uh
raising an alarm bell that thisis a something to uh pay
(44:27):
attention to because we don'tknow what we don't know.
And while we're using it forhopefully all the right reasons
and fun reasons, and it's a newtoy, uh play with that toy with
caution or be vigilant thatthere's we don't necessarily
know all the issues behind it.
And so um that's one of thosethings where uh the cautionary
(44:49):
tale is the the the response Igot.
It's like this is good, thankyou for for sharing that.
SPEAKER_03 (44:54):
Well, I I totally agree that you know these tools
are extremely powerful, but wedon't know the full extent of
their powers, to be honest.
Therefore, we it's up to us tobe cautious about how we use
them.
And that's difficult.
People you know what is caughtwhat caught what what's cautious
(45:15):
to me isn't necessarily cautiousto you, but I I think we're at a
point where it's good to havethis kind of discussion.
And uh will we agree, notnecessarily, but I totally agree
with the concept that man, yougotta be thinking about these
issues.
You just can't plunge head uhhead first into this.
SPEAKER_01 (45:36):
It's uh yeah, and I I don't envy a a board of
directors or executiveleadership team of any company,
let alone the credit unions,that they have their own
heartburns going on right now,from deposits or lack of
deposits to uh moving target inthe financial world, uh, all
these different disruptors thatare happening.
(45:57):
Now you throw AI into the mix.
Oh wow, there's so much uh uhheadspace that's swirling uh
that's going on, obviously, fromtariffs to the financial
uncertainty.
There's just so much happeningright now, and AI is just one of
those important things.
Uh, I don't know how they do it.
SPEAKER_03 (46:16):
And it's uh it's tough to and this is the biggest
change in financialinstitutions, not since the
smartphone, not since the web,not since uh either of those
things.
It's actually the biggest changesince uh institutions, financial
institutions began tocomputerize when they moved uh
(46:37):
data off of pieces of paper ontofloppy disks or magnetic tape.
I mean, that's that's when thisuh that's how big this change
is.
I mean, this is and that thatchange happened in the mid-1980s
for most institutions.
We're at the same place now, andit's it's just an immense
change.
Now, could this blow up?
(46:58):
Yeah, I mean, but then again,the transition into computers
could have blown up too, and itdidn't.
SPEAKER_01 (47:03):
So yeah, it's just uh we we're all have to adjust,
right?
And that's the the the toughpart here.
And I don't necessarily know,you know, uh one of the areas is
jobs, Robert.
You know, where the jobs cancome from when these are this
disruptor of AI is taking overjobs, but uh every industry,
everything um, or every era Ishould say, has adjusted to the
(47:27):
times.
And I don't necessarily knowwhat the new jobs are going to
be.
Um I hope there's still room forhuman beings.
SPEAKER_03 (47:37):
Oh, there are, but and you know, if you actually
look at historic staffingnumbers and credit unions,
they're much lower for, let'ssay, million dollars in assets
than they were 20 years ago.
There's just fewer people.
Yeah, uh significantly fewer.
And I think that trend willcontinue.
Can we wipe out all departments?
(47:59):
Well, I hope not.
We can't all be sitting aroundletting machines do all our
work.
I at least I hope not.
SPEAKER_01 (48:08):
Or you know, you know, what's what's our value?
SPEAKER_03 (48:10):
You know, and then machines might decide amongst
themselves to get rid of us.
SPEAKER_01 (48:16):
Well, if if it if that's the world uh we had to
live in, then um I'll be thefirst in line to say goodbye.
SPEAKER_03 (48:22):
Hey man, you must have seen the matrix, right?
SPEAKER_01 (48:24):
I know I I still like human beings.
If uh human beings are beingtaken over by AI, then well,
that's my time to you know saygoodbye.
SPEAKER_03 (48:32):
Uh go go watch the matrix again tonight.
SPEAKER_01 (48:35):
No, I I prefer not to.
I won't watch uh maybe uh AnnieGriffith rerun or something.
SPEAKER_03 (48:43):
Before we go, think hard about how you can help
support this podcast so we cando more interviews with more
thoughtful leaders in the creditunion world.
What we're trying to figure outhere in these podcasts is what's
next for credit unit.
What can they do to really,really, really make a difference
in the financial team?
Can't all be megabanced, can it?
(49:05):
It's my hope it won't all bemega banked.
It'll always be a place forcredit unit.
That's what we're discussinghere.
To figure out how you can help,get in touch with me.
This is RJMGGarvey at gmail.com,Robert McGarvey, and that's
rjmegarvey at gmail.com.
We'll figure out a way that youcan help.
We need your support, we wantyour support, we thank you for
(49:27):
your support.
The CU2.0 podcast.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.