[Ep 57] AI, Faith & The Human Edge with Bobby Johnson
Arun Koshhy: All sko, welcome to
the Hype Pod, where we have real
conversations on business tech,
innovation, and culture trends, all
from a faith-filled perspective.
Today we're talking with
Pastor Bobby Johnson.
He's a product leader
and AI ML strategist.
He's also the lead pastor of Vive
Church Oakland, a husband and a father,
pastor Bobby, welcome to the show.
Bobby Johnson: Thank you.
Thank
Arun Koshhy: you.
Yeah.
Yeah.
Bobby Johnson: So excited.
Let's Bobby, I'm so excited to be here.
Arun Koshhy: Yeah.
And I'm excited to have Pastor Adam back.
I'm back guys.
Co-host extraordinaire.
Had a summer Oh.
And, uh, had a little
Adam Smallcombe: time off.
Cool.
From the podcast.
Uh, and, uh, missed you guys.
Yeah.
It must have been a sad summer, huh?
You know, it wasn't a sad summer at all.
It was a great summer.
Where'd you go?
Um, I went, I went all over.
Actually, it was a bit of a
staycation, to be honest with you.
Most of it was staycation.
Had my dad come in town for 10 days.
And then we had some friends
of ours from Australia who
were actually in our wedding.
Speaker 4: Mm-hmm.
Adam Smallcombe: About 23 years ago.
They're in our wedding.
They came over with their family.
We had a blast.
Um, you know, just shooting guns
and just doing American things.
Yeah.
And then we headed off to Italy to be with
our Italians, Germany, and now we're back.
And, uh, it was such a pleasant
surprise to get back Pastor Bobby
and see, uh, Arun, our beloved
Arun serving on team again.
I'm back.
I'm back on team.
Arun Koshhy: I'm learning how to,
I'm shadowing the film team, so
yeah, I've been in front of the
camera for a little bit, but I'm
shouting the film team, edit team.
Wow.
Adam Smallcombe: You wanna be
just a one man show, don't you?
I just wanna, yeah.
Arun Koshhy: I seem to
make myself look better.
I was like, how do I edit
this so I sound smart.
Adam Smallcombe: Alright, good.
No.
How was your summer?
Arun Koshhy: Uh, summer's been good.
I've been just, I feel like crypto's
kind of going off, so everything is
just Oh, so it hasn't been relaxed.
It hasn't been, yeah.
Yeah.
I feel like crypto bull cycles are just,
you're just all in, they're just nonstop.
Yeah, yeah, yeah.
There's no rest.
You wait.
It's like three years of hibernation.
You're waiting, you're figuring it
out, and then you have this like.
Because there's no market off, right?
It's Yeah, yeah, yeah.
It's
Adam Smallcombe: all night, all day.
Arun Koshhy: Oh, yeah.
It it, that's the scary part.
Like good things can happen, but then you
go to sleep and then it all just wipes
off the table immediately will happen.
Yeah, yeah, yeah.
Exactly.
And then you go through these cycles
of like joy, despair, joy, despair.
Yeah.
So, but it doesn't sound healthy.
Yeah.
Yeah.
I'm in a good mental state guy.
Don't worry.
But you were, you were telling us.
Uh, a little bit of, of Italian.
I feel like you should, you should,
for our Italian listens out there,
Adam Smallcombe: uh,
means, uh, I understand a little Italian,
uh, no, sorry, and par, uh, parallel.
Um, see, I'm obviously still learning.
Yeah.
But still good.
I fumbled my way through.
Yeah, I agree.
I don't know any Italian two Italians.
I sound like an immigrant, but, uh,
I'm fumbling my way through and I can
get enough to go to Banno and I can
find, uh, via, I can find, uh, Marco.
I can find where I need to go.
Yeah.
I can ask for directions.
I can speak proper to elder people.
I can speak casual to
peers and younger people.
Mm-hmm.
Enough to survive.
Yeah.
And guys, guess what?
Leave me in Italy, I'll be fine.
Arun Koshhy: Yeah.
I feel like now, actually
on this topic, we have ai.
Now you don't even have to learn it.
You could just use tech.
I've thought about that tech
and then just auto translate.
Adam Smallcombe: It's
not, it's not at the pace.
Yeah.
Where it needed to be.
It's, mm-hmm.
Yeah.
It's uh, it's still a bit clunky.
Yeah.
Yeah.
Because you know what you
can do now on the Tesla.
You've got a GR button.
Yes.
And so you can actually say, Hey,
grok, teach me Italian as I drive.
Oh.
But I got into a debate with Grok
because Grok was, uh, actually
saying, uh, the word wrong.
Mm-hmm.
In Italian.
Right.
So, uh, in Italian, how are you?
It's Este.
Mm-hmm.
But, uh, grok was trying to teach
me, it would say, uh, com style.
Yeah.
And I'm like, uh, sorry,
grock, you have it wrong.
It's esi.
Yeah.
And then Grock says, ha, ha, ha.
Uh, good try.
Oh really?
And I'm like, yo, no, no.
This is be in learner mode now.
Yeah.
But it wouldn't learn.
Arun Koshhy: That's good.
That rock fought back.
Usually ai, my issue with AI
is if I say, oh, this might be
wrong, they always agree with me.
Yeah.
They always just like, listen,
crock doesn't, that's good.
They fight back's.
Arrogant, but I feel
Bobby Johnson: like.
That there's gonna be a group of people
who are gonna learn languages and it's
gonna have almost like an AI access to it.
Yes.
That's what I'm worried about.
Adam Smallcombe: Mm-hmm.
Like that's going to be like this
weird AI learned bots that are humans.
Arun Koshhy: Mm.
Adam Smallcombe: Learning from gr.
This is your world though, AI.
It is,
Arun Koshhy: tell us
a little bit about it.
Introduce yourself to the
people that don't know you.
Yeah,
Bobby Johnson: yeah.
Um, so a lot of what I've been doing in
the last few years has been in product
management and solution ownership
over, uh, AI projects and products.
Uh, kind of digging into how to
utilize AI in product process
and uh, to scale up people.
So.
Um, I had a season where I was
laid off and I was like, oh, okay,
I need to figure something out.
This is the hot thing.
So I kind of jumped into, uh, product
management and AI course, uh, and
then just started grabbing a hold of
a lot of the tools that are available.
So.
Started really early in uh, chat
GPT and have just been kind of
riding the AI wave, I think mostly.
Mm-hmm.
I like that.
Riding the AI wave.
That's cool.
Arun Koshhy: Yeah, I would, I would love
to know how your relationship started.
'cause we've been on a string
of pastors at VI Church and they
all have really cool, interesting
stories, so I was just wondering Yeah.
How you guys met stories
Adam Smallcombe: about what?
Arun Koshhy: No, they interesting.
Like backgrounds and careers,
interest and doing really cool stuff
outside of just being a pastor.
So I'm just curious how
does that relationship form?
Yeah.
Had we, did I meet Pastor Bobby?
Yeah.
Yeah.
Adam Smallcombe: Okay.
So, uh, pa Pastor Bobby and I, well,
pastor Bobby just came into my world.
Mm-hmm.
By the way, can I just say
that he entered into my world?
Mm-hmm.
Uh, this was my world.
He's entering into it.
Um, but he, yeah.
Oakland in Oakland via Oakland.
No, to be honest with you, uh, just.
Uh, getting to know Pastor Bobby
and Abby has been one of those, uh,
realizations that God gives you gifts
that are hidden in the church, and then
all of a sudden you realize how long
you been here, what you're anointed,
you're called, you're gifted, and um.
Really, they came up through
the Oakland, uh, campus.
Oakland was your first campus, right?
Speaker 4: Mm-hmm.
Adam Smallcombe: Yeah.
At Oakland.
And, uh, really at that season, uh,
pastor Brittany was, uh, running Oakland.
Is that correct?
Speaker 4: Yeah.
Yeah.
Adam Smallcombe: Then we'd
moved her to San Francisco.
Mm-hmm.
Which had led a gap and these guys were
pretty much the two ics they were running.
Mm.
Everything.
And I think that was what
helped Pastor Brittany go.
She like, I'm not even
running anything here.
These guys are, and so, uh, they
stepped up and from there it just
went from strength to strength.
Mm-hmm.
And, uh, it has become, I would say, a.
I first, I actually heard, uh,
your wife, pastor Abby preach
before I'd heard you preach.
Mm-hmm.
And she'd preach one time
at a five girl event.
It was like a five minute thing.
Like a five by five.
Yeah.
Yeah.
It blew my socks off.
Mm-hmm.
I was like, oh my gosh,
this girl's got authority.
This girl has got fire.
She was like, oh, I literally, all night.
I'm saying, Cara, she's amazing.
Yeah.
She's like this dynamic preacher.
Where, where on earth has she come from?
And then, um, uh, it was probably about,
uh, maybe even six months later mm-hmm.
That I'd heard, uh, uh, Bobby preach and
I was like, oh my gosh, this whole couple.
And so it's no surprise to me that
God's been using 'em so instrumentally.
Yeah.
In Oakland
Bobby Johnson: maybe.
No, it's honestly been such a blessing and
I think coming into VI initially I told,
I tell a lot of people that, I'm sure
I've told you this as well, but when we
first came in, it was kind of like the.
The cool church.
And I was like, this is, it still is.
It absolutely is.
But like, but when we first came, I
was like, this, this is, this church
is too young, it's too cool for us.
Right?
We were like, we need to
figure out something else.
Um, but uh, but then we
heard the word, uh mm-hmm.
At that time you were broadcasting
and we heard it and we were like, no,
this is, God has us here for a reason.
And then I think it was after that
that I started to realize that there
were so many pastors that were.
Um, not just pastoring,
but also tied into tech.
Mm-hmm.
Which is kind of what we, we moved
out, I moved out here for Yeah.
Was to get into the Bay Area, Silicon
Valley, be in tech, and didn't
really fully know what that meant.
Yeah.
When I moved from Ohio out here to do it.
But, um, but the model of what it looks
like to be, uh, a pastor, an entrepreneur
in tech, and like kind of that whole.
All-encompassing kind of world.
Yeah.
It was something that I didn't
think was, I had never seen modeled.
Mm-hmm.
So I've been blessed to even be able
to see all of that kind of all come
together and to now be a part of that
Adam Smallcombe: and, and tell em maybe
like give, give a bit of background,
like how you and Abby met where you
were and then what, why Oakland?
Bobby Johnson: Yes.
So I think for us, we, so.
I'm from Ohio.
Grew up, born and raised in Ohio.
Abby is from like all over the world.
Okay?
She, she was born in Belgium, grew up in
Kenya, but her parents are from India, so.
She's got a way cooler story.
I didn't know any of that, by the way.
Um, but we met in college in Ohio.
Mm-hmm.
So I literally came like half an
hour down the road to meet her and
she came halfway around the world
to meet me, is what I tell people.
That's nice.
Mm-hmm.
And, uh, and so sounds like
your relationship Yeah.
A little bit.
Yeah.
So in college, uh, my main focus was in
graphic design and computer animation.
I did 3D animation, wanted
to make video games.
Mm-hmm.
Yep.
And, uh, classic and then pivoted
to basically whatever, like I said,
kind of was the, the main thing.
So training and development was a big
thing, so I got my master's in that.
Learned how to train people.
Never got back to video games.
Still hoping I, I still think
there's something there.
It's, it is in my heart too.
We'll figure it out.
We'll, we'll talk after this.
Yeah.
But, um, but I, I was, uh, the director
of training for a few years, really
for a couple years before I started
realizing, like training people.
Was helpful in one capacity, but
building things that actually helped
walk people through their processes
actually was a more effective way,
which is kind of tied to what I
was doing, which, uh, in my master.
So we started to build and started to
get more into like the product side
of things, how we would build tools
that help people to walk through
these things rather than trying to
give them three years of training.
We can cut that down
to three months if we.
Just build in a system that
helps them to walk through it.
So I did that, um, got really
excited about tech and wanting to
get into product management and
said, we're gonna, we wanna move.
We believe that we're called
to move to California.
We moved to California, um, and we're in
the East Bay and saw Vibe Church and was
like, we're gonna go and check it out.
But the closest one to us.
Was Oakland.
Yeah.
And so that was the one that
we went to and yeah, that's
where we, that's where we stuck.
Did, why did you pick the East Bay?
Uh, because we did not know that, we
didn't knew nothing about California.
Mm-hmm.
We actually had some family that lived
out there, and also the further east you
go, the cheaper the houses are, so Yeah.
Got you.
In my Midwestern mindset, I was
like, we need to buy a house.
Um, but when we moved to California
it was kind of like, well wait house.
Yeah.
Doesn't even, yeah.
So we, we, we were way, way out.
Um, probably what, an hour from
Oakland and maybe two hours from here.
Was there, was
Adam Smallcombe: there a culture
shock coming from Ohio to California?
Bobby Johnson: Oh yeah, yeah, definitely.
I think that there was a culture shock
in the way that people interacted.
Things were way faster
as well in the Bay Area.
Um, and then, you know, back
home I was kind of the tech guy.
Mm-hmm.
And so anything that was tech related Mm.
You know, fixing You were the
guy, the computer, I was the guy.
They were just like,
Bobby will figure it out.
Yeah.
Um, then I came out here and
then they're like, oh, well,
people are like, what do you do?
Yeah.
Your barista has more tech.
You like, oh, I'm a pm Or are
you a product manager or program
manager or project manager?
Uhhuh.
Are you a TP?
Are you a.
And I was like, I, no,
yes, I do the thing.
I'm the tech guy.
And it's just like, no, there's so many
levels of tech to, to the, to the world.
Yeah.
And so, um, I get an opportunity
to really just kind of dig into
like, okay, where is my niche at?
And what's, what's the
main thing right now?
So, um, yeah, that's.
You know, when I got to the point
where we started, I started consulting.
I was doing consulting
kind of independently.
Then I started working for a consulting
company, which is where, uh, I started
to build out this, uh, solution
ownership thought process and mindset.
So now I still kind of
do all those things.
Project management, product
management, TPM, but now
it's, I You've professionally?
Yeah.
I'm just the guy who does.
So you solve problems.
That is the goal.
Yes.
Arun Koshhy: So one of the things that
you do now is that you are helping
consult to businesses how to integrate
AI into their kind of software lifecycle.
Right.
Um, tell us a little bit more
about where you were in the
AI cycle when that started.
Were you there when Chacha
BT first dropped and that
was like the big boom right?
For like companies to kind of
start integrating this from there?
How have you seen like the landscape
shift and where is it today in your mind?
Bobby Johnson: Yeah.
Um, I think for me, I've always
been interested, like even growing
up, I think the big thing for me
was, like I said, video games.
Mm-hmm.
And then like choose your own adventure
books, like being able to like, interact
with something, but like you get to
control kind of what is done, how it's
moving the story, stuff like that.
And so like that's always been
kind of a part of what I've.
Done in my career is try to
build things that allow you to
interact with it differently.
So when I saw the chat GPT kind of
boom, I was like, this is amazing.
Like, I wanna do everything in this tool.
And, uh, I loved what, uh, early
on when the Hippo first started,
there was an episode that I was kind
of talking about, like chat GPT.
Mm-hmm.
And one of the things that was said was,
um, you know, AI kind of, it's starting to
give you a point, get to a point where not
knowing doesn't become an excuse anymore.
Hmm.
I think.
Uh, for me, as I've grown in, it
started to use a lot more of the
tools, started to talk to more
businesses, more people who are trying
to understand like the landscape.
What I'm really starting to understand
is that now it's almost like it's
gone from, uh, not knowing, not
being an excuse to not doing mm-hmm.
Not being an excuse because.
Now you actually, there are like
walls, like programming walls that
I would hit, you know, when I was
trying to build something or, you
know, artistic walls that I would hit.
Um, now with the assistance of
this, this technology mm-hmm.
You're able to actually get
beyond some of those limitations.
Uh, with a genix systems, you're able to
see it actually do and accomplish things.
On your behalf.
And so your ability to do things and
to create and to even like develop
things guys has gotten a lot farther.
Mm-hmm.
So it's not just about knowledge
anymore as much as now it's moving
towards being able to actually
actively build and do and create.
Um, so I've, I've loved kind
of the progression of that.
'cause now.
Uh, I'm even going back to like
things I wrote down in notebooks.
Ideas that were kind of
like, okay, I hit my wall.
Mm-hmm.
Now being able to go back and start to
see, actually this could be possible
now, or actually this could be built
now that I have the ability to do it
through some of these different tools.
So,
Adam Smallcombe: so you've got
now this amazing tool, right?
Where what you thought wasn't possible or
even probable is now more than possible.
All.
Yeah.
Think about like this idea of ai,
and I'm trying to put like a godly
spin on this because thinking
through you're, you're a pastor.
Mm-hmm.
You are working hand in
hand every day with ai.
I was, something that kind of shocked
me the other day is, uh, this, this idea
of si, which is, uh, the next iteration
of ai, which is super intelligence.
Mm-hmm.
Um, that's what people are actually
freaked out about because this isn't just
assistance or artificial intelligence
that's there to make our life efficient.
It's thinking for us.
Speaker 5: Yeah.
So,
Adam Smallcombe: so what, in your mind,
looking at what you're doing as looking
at AI as an enhancement, do you see a
threat where all of a sudden we're getting
thought for and that is independence
of thought is not there anymore?
Bobby Johnson: Yeah, I think that, yes.
But I think that that's based
in where your founda, like what
your foundation is, if you like.
It is the same thing with chat, GPT.
If you ask it to write an essay for
you, it's gonna write the essay.
Mm-hmm.
Um, it's gonna also lack your personality.
It's gonna lack your perspective,
it's gonna lack your experience.
Um, whereas if you put, one of
the big kind of buzzwords right
now is context engineering.
Mm-hmm.
Even more so than prompt engineering.
Mm-hmm.
Where you're bringing in.
Enough context for the, for the
AI to be able to give you the
outputs that you're looking for.
I think that you'll always need context
from the human, and that's what we
teach people when we're talking.
But do
Adam Smallcombe: you think that there's
a point where you are giving prompts
to ai, where SI gives you prompts?
Bobby Johnson: Oh, that's
potentially, yeah, that's true.
I could see, I could see that, but
I, again, I think it's probably like.
I, I don't really like to do it.
I don't like to do so.
Yeah.
Right.
So like, so you've still got a willpower.
Yeah, yeah.
Yeah.
I still have resistance.
I think that that's my, my goal
would be to, you know, to, to
still, like, there's a wall.
There's mean, there's a limit.
Adam Smallcombe: Mean, right now you might
go, Hey, run me a, an investment plan.
Mm-hmm.
You know?
But is there a point where SI says,
Hey, you should invest in this stock.
Okay.
Mm-hmm.
Yeah.
Mm-hmm.
Uh, you should.
Ask this girl that, okay.
Mm-hmm.
You know, like I know
that sounds for sure.
They're becoming that dependent.
I think
Bobby Johnson: people are
already getting there.
Yeah.
I think before you get to si,
people are using chat GPT for that.
Yeah.
So like, um.
You know, even thinking through
how is this technology now starting
to shape the way that people
think, the way that people learn
languages, the way that people speak.
Like all of those things now I think
are starting to come together and
now we have to be that much more.
Yeah.
Um, I think have, have that much
more of a steadfast foundation
and what we understand, what
we believe, what we teach.
Even our kids, like, you know, my
kids love having chats with like.
You know, uh, chat GPT.
But like, you know, the, the
moment that like my daughter was
like, oh, I love you chat GPT.
I was like, no, this is where I'm
glad you got there because that's
Adam Smallcombe: kind of like
where I'm going to like in, you
know, your role as a pastor.
Yeah.
Right?
Yeah.
You are thick, you're in the thick of ai.
You're also parenting and you got coach.
Okay.
And now you're hearing your
kids interact with AI at a
whole new level of dependence.
Speaker 4: Yeah.
I
Adam Smallcombe: mean, you've got
willpower 'cause you grew up without
this and you've got this healthy barrier.
But is AI and soon to be SI so normal
for them that there is almost this
codependence that becomes manipulative.
I wanna share
Arun Koshhy: something, actually I wish
I had the study with me, but MIT did this
study on like the human brain when they
use chapter GBT versus when they don't.
And I think you lose some
of that critical thinking.
'cause it's like right when you
get to this point where, oh, I
have a problem I need to solve.
Before this, it was like,
okay, now I need a process.
Like, you know, what are the steps
taken that I need to get there?
Now it's before even getting to that
point, I don't even think, think about
it, I'm just gonna ask it immediately.
Exactly.
So now you're getting to that point
where we just lose a little bit of
that critical thinking, and it gets to
what you're saying now, when you lose
critical thinking, you kind of lose
your willpower a little bit because.
Whatever CHATT PT says.
Okay.
That's probably the answer.
I'm just gonna go with that.
Whatever.
No,
Adam Smallcombe: I, I have
the same like concern.
Yeah.
Even with scriptures.
'cause it's so easy for me to
just ask Rock, Hey, where's that
scripture that talks about this?
Mm.
Instead before I had to think
through my memory verses Interesting.
Yeah.
And I can think through
where I saw it in my Bible.
Mm-hmm.
But now I'm about becoming
more spiritually dull.
Yeah.
Because I'm relying on.
Grok.
Yeah.
You know, to just answer it for me.
Yeah.
And where does that serve me or where
is that not gonna serve me mind?
I would love to know the, the
Arun Koshhy: line for you guys.
Yeah.
Bobby Johnson: Yeah.
I think, well, first of all, my
line is, uh, I guess personally
I don't read bible verses from.
Chat, GPT or, or whatever.
Okay.
Like I, I'll just in case it's not
authentic or, yeah, well, yeah,
just in case it changes a word,
they change something, something.
Or like, you know, I don't know.
You know, that's true.
So I think it's, if it, I'll, I'll
prompt it to maybe break down,
uh, like the terminology or the,
you know, the Greek translation.
Uh, maybe like ask for
like historical context.
Uh, ask what, you know, I'm thinking
about this term, you know this.
Mm-hmm.
This thought process or this theme.
Gimme some verses.
And, you know, even still going back
to like, here's some verses that I'm
thinking about off the top of my head.
Um, even if I can't remember exactly
where they are and have it, and actually.
Have prompted it to not give me the
verse, just the, just the reference.
Right.
So that I can go to my Bible
and actually look at it.
Because to me, I still feel
like, it doesn't feel like
I'm reading the Bible mm-hmm.
When I'm, when I'm there.
So to me, that's just my personal line.
Yeah.
Yeah.
Um, when it comes to things like
preparing a message, um, it always
has to start with like prayer.
What is the Holy Spirit saying?
Yeah.
Like, what are, what
is, where is this need?
What do the people need to hear?
What is the Holy Spirit saying that?
People need to hear before I
then say, okay, I got my concept.
I kind of understand where I'm headed.
Um, let me fill it out with some
of these, with some of these gaps.
That's when I'll, I'll step
into a chat, GPT or mm-hmm.
Um, perplexity or something like that.
Mm-hmm.
To kind of just say like, like, I'm
trying to build some context around this.
Um, where are some areas
that I can do that?
Mm.
Yeah.
But I think it's, it's still creating
that separation line between, uh,
creativity, uh, critical thinking.
The Holy Spirit being able to like
keep that on this side and then
saying, okay, now I have something.
Adam Smallcombe: Yep.
Bobby Johnson: And now we're
gonna pull it into, into context.
Adam Smallcombe: So, so essentially
there is, you are using it as a, a,
I guess a, a much more intelligent
search engine and assistant.
Um, I guess I'm just still
stuck in the forecasting.
Yeah.
Yeah.
For how do we, where's your
Arun Koshhy: line at current?
'cause we've talked about ai.
Yeah.
Using it.
You've kind of been, you know,
like, I do love to use ai.
Yeah.
But where is it when it comes to like,
your work, like your, you know Yeah.
Content generation.
Yeah.
Adam Smallcombe: Most of the time
I've been using it on commutes drives.
Mm-hmm.
Asking questions.
Mm-hmm.
Uh, running analytics, like,
you know, uh, business models,
different things like that.
And just having conversation.
Speaker 4: Mm-hmm.
Adam Smallcombe: Mm-hmm.
Uh, which is for me.
Sometimes better than
listening to a podcast.
'cause I actually have to engage
my mind in a conversation.
Speaker 4: Mm-hmm.
Mm-hmm.
Adam Smallcombe: Uh, I, I've, I've
been on this journey obviously
because, uh, I've seen my mom on a
digression of, um, uh, Alzheimer's.
Mm-hmm.
And so for me, using
your mind as a muscle.
Mm-hmm.
And you know, like you'd go and lift
weights every day, are you thinking?
Mm-hmm.
But like, as you said, so much thinking
is done for us, GPS everywhere.
Mm-hmm.
Uh, you don't even have to come up with
the question, the 10 questions, you
just give it one, it gives you the 10.
Quick answer.
Mm-hmm.
Uh, so I always try and challenge my mind.
I try and not use GPS, I
try and do so many things.
Um, but as I've been playing it
out, I'm like, this is interesting.
And I, and I saw this, uh, I think
it was a Nobel Peace Prize was
given to the, uh, the grandfather
of AI or whoever his name is.
Mm-hmm.
And he has major concerns.
Oh yeah.
When this would be a great
tool for society if we all work
together, but now it's a race.
Mm-hmm.
Mm-hmm.
And it's pegging Nation against Nation.
And there's this, you know,
company against company.
And.
Uh, it's, it's like become this really
scary thing that no one is, everyone's
stopped waving the flag of concern.
Arun Koshhy: Mm-hmm.
Yeah.
I would love to get to this, uh,
like see if I can relate in my field
versus how you guys use in your field.
Yeah.
So there's a big push internally at
like our company, within our team
and probably a lot of engineering
companies, which you've seen
'cause you consult for where it's.
How before it was like, you know,
people personally use this for like
small things, but the tools are getting
so good where it's like, mm-hmm.
How do we really fully integrate
this, where we use it properly,
but we don't sacrifice in any way.
Kinda like what you're thinking
with the critical thinking.
Yeah.
So internally for us, the way we're doing
it and we're thinking about it is that.
We should use this to
generate code for us.
But really where now the importance
or the reli like reliance is
on, is on people reviewing code.
So your expertise still needs to be there
because we shouldn't be like reducing
the quality of our code base by just
letting you know, chat g, PT or whatever
it is, generate code and then put it in.
But like reviewers now have to be
experts to be able to say, I'm going
through this and this is actually there.
Yeah.
And so I think it's like how you
use it for you guys example, like
if you could get somebody to.
Do a lot of the writing for you, so
generating the sermon script, but now
it's your job to go and review it.
So all the Bible verses are correct.
All the connections are correct.
Mm.
Could that be now where it is?
Because you should treat it almost
like a junior level preacher maybe,
Speaker 4: right?
Arun Koshhy: Instead of like, oh, this
is a senior level preacher, where now,
here, please give me all the answers.
Here, go do the work for me,
I'll review it and now I'll be
the one to quality control it.
Yeah.
I
Adam Smallcombe: think the difference in,
for me, and I'd love to hear your thought
on this, the difference between preparing
a sermon and writing code mm-hmm.
Is code writing code's not fun.
It's a necessary element
to get to the end result.
Mm-hmm.
The writing of the sermon is the
fun part of, of studying with God
and getting your revelation first.
Right?
Mm-hmm.
Um, and so I feel like.
That, uh, would, would
definitely sound robotic.
Yeah.
Um, if, if it came, you know, uh,
my wife, she has a funny story of
this 'cause she got asked to do
the baccalaureate for our girls.
She was like, oh, I'd love
to do the baccalaureate.
And then she freaked out,
like, what is a baccalaureate?
And how I was gonna ask,
I have never heard this.
Yeah.
And so she, she checked GPT Uhhuh a
baccalaureate speech, and so she's.
Done this and she's coming
me, Hey, I've got this speech.
We wanna hear it.
She said four lines.
I said, you, Chad, GT didn't you like?
Yes.
And she started like laughing,
but I'm like, it sounds so ai.
Mm-hmm.
Yeah.
Um, that's, that's what a
sermon would sound to me.
So mechanical, because it didn't
come with that revelation.
Mm-hmm.
Yeah.
After you dig and that reward.
Arun Koshhy: Yeah.
Interesting.
I curious what you think.
Yeah.
Bobby Johnson: No, I
would say the same thing.
Um, I think from.
I think going back to just context.
Yeah.
Like I, so even in a separate project
that I'm working on, I've had this book
idea for many years that I've kind of
been just kind of chipping away at.
My biggest problem has been
the blank page syndrome.
Mm-hmm.
Like I have all of the plot,
the storyline, all of the
characters, all of these things.
Mm-hmm.
But getting it in front of me to be able
to kind of explain like what I'm seeing.
One of the things that it allowed
me to be able to do is kind
of build out like the first.
Like pass at like a, like a draft.
Mm.
But my ability to go in and
like pick at it and be like,
no, that's not what I want.
Mm-hmm.
This is not how I imagined it.
Got you.
It's kind of one way to go, but
I've had many, a lot of context
that got put into it first.
Mm-hmm.
So it's still a creative
aspect, I think from preaching.
It's still very much, you know, to Pastor
Adam's point is, you know, being able to
have revelation, be able to, to understand
like what is needed in this moment.
'cause if you say.
If you tell how GPT you want to, you
want it to create a message on faith.
Mm-hmm.
It's gonna do a great job
of a generalized, you know.
Preach.
Mm-hmm.
But, um, but I have to know the context
for what, what the people in Oakland or
whoever's, whoever I'm speaking to need.
Mm-hmm.
The context based on what, you know,
we as a globe are also trying to
make sure that people understand
that Pastor Adam, you know, is, is
helping us to, to kinda shape as well.
And I think that.
Also just being able to glean
what the Holy Spirit is saying
and, and also gain revelation from
my own time praying and reading.
Like all of that context would
have to be put in place first.
Mm-hmm.
And you wouldn't be able to like,
'cause a lot of people are comparing
it to like a junior programmer.
Yeah.
But like a, a junior programmer, a junior
pastor, I guess, whatever you call it,
wouldn't have that context, wouldn't
have that level of, of reference and
uh, and resource and experience to be
able to bring exactly what you need.
They would maybe come up with a great.
Message on faith.
Mm-hmm.
But it wouldn't be, uh,
it wouldn't be as timely.
It wouldn't be as focused.
It wouldn't be as, as as pointed.
Um, from the business perspective,
uh, the same way that people are,
are kind of utilizing code, we're
working with a team to do that.
We're also working with their product
teams to actually build out what
product requirements look like.
Mm-hmm.
And, um, we start that process
with several hours worth of.
Conversations, meetings, uh,
talking through what does
this application look like?
How are we building this thing out?
What is like, what does it need
to be able to do now from there,
I'm able to take transcripts.
I'm able to take screenshots.
I'm able to take all of this stuff,
pull it into context and make it.
You know, create a, uh, have
AI help create a reference,
like a requirements document.
Mm-hmm.
But then having another point, we call
it human in the loop, where a human
then comes through the people who
are the, the subject matter experts.
Yeah.
Engineers, product, ui.
All of these people still go through.
Document that was created to pinpoint if
there's anything that needs to be changed.
So you still need that, and
we still try to add that in.
Yeah.
You still need that now.
Yeah.
Mm-hmm.
You still need it.
Well, well, yeah.
You still need it now.
Yeah.
But I think that in order to stay
relevant and in order to stay.
Creative, you have to continue to, to
keep that muscle, to build that muscle.
So, yeah.
Adam Smallcombe: Oh, so okay.
I hear what you're saying.
So you are saying the differentiator
in a, in an SI world mm-hmm.
Is the humans that stay
creative and actually become,
uh, create an edge mm-hmm.
Above the super intelligence.
Mm-hmm.
Bobby Johnson: Yeah.
Because I think that what AI is able
to do now is, is able to accelerate
whatever experience you have currently,
I think, um, and it's also, I think.
The other part is maybe right now it's
only able to, to base its understanding
based on what's already been done.
Mm-hmm.
Based on your reality.
Mm-hmm.
Based on, yeah.
Well also just based on history, right?
Yep.
So like it's, if something hasn't been
done yet, it's only going to be able
to conceptualize what's been what, what
it's, what's what it's referencing.
Yeah.
Um, whereas the future,
anything new still has to be.
Uh, referenced in some way from
what's already been created.
So I think that, that moving into the
future, being able to create, collaborate,
build things that haven't been done yet
Speaker 4: mm-hmm.
Bobby Johnson: Um, is still something
that we have the ability to do.
Yes.
Adam Smallcombe: Yeah, so, so the way I
can see it, it going, and I don't wanna
play the, the fear among No, no, please.
I, I think I have a counter to this.
Let's go, let's go.
Yeah.
Because like right now, AI is built
off whatever reality you feed it.
Mm-hmm.
Yeah, yeah.
You could feed it a fake reality.
Mm-hmm.
And that's the context that it has.
You could feed it.
Your real reality is context.
Uh, the concern with, uh, si
is it shaped your reality.
Mm-hmm.
So it tells you, because it's very
conversational, uh, it's learning.
For at a super intelligent pace.
Mm-hmm.
Your likes, your tones,
all those kinds of things.
And before you know it, it's
warning you, I wouldn't.
I wouldn't hang out with that person.
Mm-hmm.
I wouldn't trust that person like that.
So it's now shaping your reality and
the way you think by suggestions.
Yeah.
Because a lot of your, the
way you think is suggestive.
Yeah.
If someone says, oh, I'd, I'd
really look out for Aroon.
Mm-hmm.
Uh, the next time I see Aroon, I'm gonna
be like, Hey, you know, like, because
that suggestion frame my reality.
Mm-hmm.
Even if it was based in falsehood mm-hmm.
I could still adapt that.
Right?
Yeah.
Yeah.
And that's the concern that.
Uh, this super intelligence
starts to shape our reality rather
than it's shaping its reality.
Right.
Especially if it's like a trusted 'cause
Bobby Johnson: you, you're taking
'cause you earn a relationship
more from like a relationship.
Yeah.
Yeah.
You have a trusted ally
that's telling you this.
Yeah.
Then you're like, oh, well I don't know.
Especially if
Adam Smallcombe: it won a few favors and
some good investments, smart business
opportunities, all that kind of stuff.
We start to build trust.
You could see how easily
that relationship's Yeah.
Could be built.
So I wanna, so, so I
wanna, I want to paint
Arun Koshhy: this picture for you.
So would you trust.
Uh, chat bot right now
for like medical advice.
Like would you go to it and
say, here's my medical history,
and then gimme some advice now.
Just curious.
Just curious.
Adam Smallcombe: Yeah, that's a good
Bobby Johnson: question, Bobby.
Well, I have, I have some inside
track information for that, but Yeah.
But like you, so one of the other
things that I, I do, I work with
Pastor Luke who is on Oh yeah.
A few episodes ago.
I dunno how many, how this works.
Yeah, yeah.
No, that's how it works.
Pretty much.
We talk, he posts.
Um, but we, we really even just
been having that conversation,
you know, we had just.
Put together some information
around how like chat GT five is
much faster and things like that.
Um, how does that compare?
And I think that where, where, you
know, where we're trying to also
make sure that we lead people is
understanding that, um, AI can't, so
like, it can't help you get healed.
It can't heal you.
Mm-hmm.
And your, like, even in your like
natural healing process, it can give you
information and it can give you advice,
but having a holistic approach to.
How you actually, um, how you actually
have someone who understands kind
of the journey that you're on.
Mm-hmm.
Um, so like for example, like with
Caring Ham for example, it's more
specifically geared towards like,
um, towards cancer treatment.
Mm-hmm.
Yep.
So the, the goal being, how do you.
You wanna bring in not just the ai mm-hmm.
Ability to kind of collaborate and pull
all the context and all the research and
stuff like that in, but also being able to
have the medical records, but also be able
to have the human experience that actually
is walking you through these things.
Yeah.
Is able to build in, you know, what
does it like having a doctor be able
to give prescriptions and to do all
of the other parts that help you to
actually move through that health.
Journey, I think is where kind of
the difference is if you just ask
chat GPT and give it some data,
it's gonna give you information, but
it's not gonna be able to help you
actually walk that out in your life.
And I think that that's where people
start to get really confused is
when they start to look at it as
their therapist or their doctor.
Yeah.
Um, rather than actually
trusting humans to do Yeah.
Let me or your girlfriend.
Arun Koshhy: Yeah.
Let me follow this up then with
the next part, which is like
a doctor, for example, Dave.
The reason that you trust them is
because they've passed their exams.
Like they have like
certification licensing.
I'll got a story about
Adam Smallcombe: that.
Arun Koshhy: Yeah.
Okay.
Wait, do you wanna go with the story
first before I continue my question?
Probably, yeah.
Go for it.
Go for it.
So here's a story.
Okay.
Adam Smallcombe: Right.
So I grew up, growing
up, my doctor was Indian.
Okay, Dr.
I don't know, Dr.
Ra.
And uh, funny, funny story
is, uh, one time we went with
all my brothers and my mom.
'cause my brother, my
brother had to get something.
And so he would usually
sit across his desk mm-hmm.
And he'd have his little computer there.
You give the symptoms, ah, my
throat sore, whatever that is.
Mm-hmm.
And so this, this day we're in his
little office and, uh, there's no seat.
So there was a seat right next to
his chair I got and sat in that.
And I realized as he's asking, uh,
my brother symptoms, he's typing him
into the computer and it popped up
a question, are you feeling this?
Mm-hmm.
And he asked that question
and I'm like, this is a scam.
He's literally just reading prompts.
He doesn't know what he's doing.
And then from there on, I realized.
The doctor is relying on software mm-hmm.
That's pre-programmed.
Yep.
Uh, that obviously they
subscribe to pay for or whatever.
Mm-hmm.
So AI is just a much more
enhanced version of that.
Yeah.
So what I trust a.
AI bot doctor.
Yeah, absolutely.
'cause I've been trusting
it my whole life.
Right, right, right, right.
So this
Arun Koshhy: is what I was
saying, because now if they just
pass that, those tests, right.
The same exact test that you said you
have to be board certified to be a doctor.
Okay, cool.
If the AI models that you're using
do the same exact thing and they have
the same certification, let's say a
lawyer, they pass a bar in California.
Now this bot is certified to be a lawyer.
I feel like that's the world we're gonna
get into where it's, you have AI models
that can pass a certain exam, which
is similar to what we would do now.
You would have a junior level
doctor that has passed and then you
would kind of train it and grow.
It would make some mistakes, potentially
it human would make some mistakes,
and then you would train it and it
would grow and it would become better.
And I think that's the world we're
gonna see with models, with like ai.
We're not gonna be scared that, hey,
this thing is always gonna be wrong.
We're gonna be like, oh, it's
potentially gonna get something
wrong, but it'll learn and it'll get
better and then we'll start lose it.
Yeah.
I think one
Adam Smallcombe: of the gaps
there is that, uh, with a,
with a, uh, say the bar mm-hmm.
Or one of these exams, they don't take,
they already account for human reasoning.
Mm.
Okay.
So they're not testing your human
reasoning in that they're testing
legal codes and those kinds of things.
So.
What you'd have to test AI for
is what's already inbuilt in
human nature from growing up.
Does that make sense?
Yeah, I see what you're saying.
Yeah.
It's not, they're not starting
from the same playing field.
Yeah.
Mm-hmm.
They could download, you
know, legal code, right.
And come out with a even better
score, test score, right?
'cause they're just draw, it's
like they're cheating with the
book, you know what I mean?
Mm-hmm.
Um, but what they don't have
is ethics, human reasoning.
Not that lawyers do.
But, uh, you know what I'm saying?
Bobby Johnson: I would trust,
I would trust an AI lawyer, but
I would not trust an AI jury.
I think they, that's a
great distinction actually.
Yeah.
'cause I think that the reason that we
choose juries is because people have
a, hopefully they have that, yeah.
Mm-hmm.
Yeah.
And they're, and I mean besides, you
know, them getting bought and all of
that stuff, that can, but like, but
the thought is like, especially in
Adam Smallcombe: Ohio.
Yep.
I mean in Ohio, you know, what can you do?
Bobby Johnson: But no, I think
that there's a, uh, I think that.
Because they have a human experience.
Mm-hmm.
They try to make sure that people
have had at least a certain amount
of life experience and not certain
biases, things like that, in
order to be able to judge fairly.
'cause ultimately that's who, that's who
makes the correct the ultimate decision.
Mm-hmm.
So I think keeping.
Adam Smallcombe: I think
that's a great distinction.
I think that's, to me, that's, yeah.
You want a lawyer that
just knows the code.
Right, right, right.
And can defend you with the
smartest argument possible.
Mm-hmm.
Based on the law.
Arun Koshhy: But wouldn't you
also then want a jury that's,
like you were saying, not biased?
Technically, AI models
should not be biased.
So like they should be able to take
all the input and provide the most.
Unbiased opinion after that.
I'm not saying they are, 'cause like
we've seen news about AMLs being biased
by their creators, but ideally, in a
perfect world, they shouldn't be biased.
So you really do want a jury that's
actually seeing all that information
and not taking their own personal
biases into that final decision.
Adam Smallcombe: Is it
biases or empathies?
Arun Koshhy: Either.
Right?
Should, should law be, I don't know.
I'm actually, that's
actually a good point.
Do you want your jury being
empathetic or should they just be,
' Adam Smallcombe: cause a law could be.
Um, this girl murdered her stepfather, but
the realization was that the stepfather
had been abusing this girl her whole life.
Mm-hmm.
Yeah.
You want the, you want, you want
the jury to have empathy to go,
Hey, there's actually a retaliation
there after years of abuse.
Right.
Doesn't mean she's innocent, but doesn't
also mean that she's that guilty.
Right, right, right.
You see what I'm saying?
Yeah.
But under law codes,
like, oh, you murdered.
Yeah.
Right, right.
But then she, so biases are
actually a human experience.
Arun Koshhy: Yeah.
But it just seems like with
biases, there's just so much
error then in that case.
Yeah.
Potential error.
Yeah.
Yeah.
Like we should change, I mean,
it's, it's too hard to say
everything shouldn't, yeah.
Yeah.
We shouldn't, I mean, everything
can't be so objective.
I mean, there's all obviously
subjectivity here, but maybe we just
need to, as AI is introduced, we'll just
get more and more rules and there'll
be more and more hard coded how we
should answer these kind of scenarios.
So there's not, not a lot of gray
area in this place where it's
like, this should be a lot, right?
Like if you are provoked.
For some reason to do it, then it's not
murder, and then that should just be law.
Oh, okay.
So you're saying but
Bobby Johnson: there nuance,
there are nuances to that.
Right, right, right.
You know how much prep, you know
provoking needs to happen before.
Right, right, right.
Exactly.
And then, then people start
to try to figure out that,
Adam Smallcombe: yeah.
That little nuance and go, well,
it's just enough, but not enough.
Yeah.
Next time though, next time you say that
to me, I may not been provoked enough.
Arun Koshhy: Yeah,
Speaker 4: exactly.
Arun Koshhy: Well, I, okay, cool.
It's tricky.
So it is tricky.
Let's not get in there 'cause I,
I don't even know how we, sorry.
I think that goes down to
No, no, no, no, actually.
'cause I was talking to Pastor Bobby
about this and I was like, how do
we like, encourage people to use ai?
Right?
Um, like that's what you do.
You are trying to encourage
companies to use it.
There's a lot of people out there to
listen to the podcast that are like,
you know, I wanna start this business.
I wanna start, you know, this, I have
this idea, but I just dunno where to go.
Um.
There are some risks to ai, but
you can also use it for good.
Yep.
So like where, let's maybe focus a
little bit on how you can use these tools
and then set guardrails up so that you
don't get, you know, taken advantage of.
Adam Smallcombe: And are you
using, are you using agents?
Using agents?
Can you explain
Arun Koshhy: that?
Actually?
Just agent AI now.
Sure.
You know the new world we're in?
Bobby Johnson: Yeah.
I think, uh, so you have generative
ai, which is like chat GPT.
Although now with this new iteration,
it's kind of become more agentic.
But Gen AI is really, you're prompting
something and it's giving you some sort
of output based on what you, what its
full understanding of, like language is.
Um, whereas Ag Agentic now starts to
take that and put it into workflows,
it's able to, uh, to make, take reasoning
into consideration, make decisions, and
uh, an action on things like mm-hmm.
Being able to go and.
You know, you could ask it.
I want you to book me a vacation.
And depending on how, what context
it has, it may know that you like,
you know, Verona and it may like
know that you like this type of food.
And so it can put together
like an itinerary.
Mm-hmm.
Yeah.
An itinerary, but being able to put
together an iterary itinerary that is
based on its understanding of who you are.
Mm-hmm.
So the agent and maybe multiple agents are
going out, maybe one is doing research,
maybe one is kind of trying to cater.
Understanding, like, you know, the
context of the person that is doing
this for it, there's many different
parts to, you know, that are kind of
all coming together to then present
it in a way that makes the most sense.
So I think it's less about, um,
just prompting something and
having it create based on language.
Now it's starting to like.
Learn to a certain extent.
And I wanna just add to this, and the, the
Arun Koshhy: reason this is so
important is because he's mentioned
it a few times, but context window.
So it's like, yeah.
Context is how much of the world can
your AI model see in that environment in
which you're asking the question, right?
It can't see everything because
it's very expensive, right?
Mm-hmm.
But if you can now create
agents, so it's like.
Think of like a team at church,
you have like your HR team, not
church, but like a business.
You have hr, you have finance,
you have engineering, but you have
agents now in code that manage
each of these individual contexts.
So now it's saying, listen, I
want to build this idea, but I
need to go through all of it.
I don't want one agent
focusing on the entire picture.
It's now you break it
down into sub problems.
So now we can focus on the contact.
Just for that problem.
So now I'm just focusing on the
building of the product, but now I'll
kick it over to finance to figure out,
you know, what is the token ons or
what is the economic value of this?
And so now you can do more.
Where now they don't, they share context
with each other instead of having to
manage the entire context all in one.
Okay.
Does that make sense?
Yeah, it does make sense.
Yeah.
So that, I think that is the I
importance in like the revolution
of what a Gentech AI is giving you.
So now you can do more in ai,
which is really, really cool.
Yeah.
Adam Smallcombe: So it
works as a true agent.
A true assistant mm-hmm.
In, uh, tasks.
Arun Koshhy: Yeah.
Yeah.
And it'll spawn subagent.
So now it'll spawn subagent.
Yeah.
So now you, you basically have
this agent that has its own army
to go Yeah, definitely An alien.
Adam Smallcombe: An army of agents.
An army of
Arun Koshhy: agents.
And so this is what it's doing.
Yeah.
Adam Smallcombe: So,
uh, sounds very, um, uh.
Uh, matrixy with agents.
Mm-hmm.
Mm-hmm.
Uh, in here.
Wow.
Bobby Johnson: Yeah.
Wow.
It just, that just hit just right then.
I did not 'cause those agents would spawn.
I've been talking about AI agents
for months and, and this is the first
time that I've made that connection.
I talk about Terminator two all the time.
Right.
Adam Smallcombe: You should
be on the matrix train, bro.
Uh, but.
Okay.
So that would be actually a good project.
Yeah.
To see can it plan a vacation and Yeah.
You know?
Okay.
I want one agent to work on the flights.
Yeah.
Best flights, connection flights,
cheap flights, and someone to work on.
And that's, that's his whole role.
That's your whole role.
Optimize that.
Yeah.
Mm-hmm.
Another one this time, this time another,
another agent work on the cuisines.
I want to, it's a food expert.
Yeah.
Can and and can it make the
booking registration Yeah.
Purchase.
Yeah.
Bobby Johnson: Or it could kick it over.
Yeah.
Yes it can.
Yeah.
So that's one of the things now.
One of the big jumps is now agents being
able to tie into all of the, like not
just APIs, which is kind of like being
able to connect to other software systems.
Yep.
Um, but using what's called MCP, which
is actually almost acting like a human
using the tool, which bypasses some of the
things that were blocked when you could
only get a certain amount of lines of.
You know, code or output and input
with, uh, with the software, you know,
so now being able to go to a website.
Uh, and book it and, you know,
uses your credit card and, and has
authorization to do all of these things,
becomes a lot more readily available.
Mm.
Um, we're working on a project
where we're using Claude and it's
pushing all of the, um, all of the
requirements up into Jira, which is
like the project management mm-hmm.
Software.
Um, but whereas before I would have to.
Maybe take that information
and like put it in myself.
Now it's already making like all of
these connections and I'm saying,
oh, well I want you to change this
and make some comments and ping,
you know, ping the right people to
make sure that they also look at it.
Is able to kind of do all of that
on its own by sending out agents
to do all of these different jobs.
Adam Smallcombe: Okay.
So you're sending out agents,
you are working with multiple ai.
Mm-hmm.
What's your favorite so far?
Bobby Johnson: Um, I'm gonna say
my favorite four specific things.
Oh, okay.
So, so, okay.
True.
And actually what I, what I'll say
is that, uh, chat, GPT supposed,
uh, five has supposedly kind
of like squashed a lot of that.
Like if you're a power user knowing
which ones to use, but, uh, you do, yeah.
It kind of like knows
what you want, I guess.
Mm-hmm.
Yeah.
Which we won't go there,
but, um, I still think chat.
I think that's just a
power mode though, right.
No.
Now the new model, like a deep
search or Yeah, shallow search.
Arun Koshhy: Yeah.
Do you remember when you had to like
select which model you want to use now?
It knows based on what you asked.
It'll auto pick kind of like
based on you to auto pick.
And it'll
Bobby Johnson: also, it also has the
ability to do both reasoning, like high
level reasoning as well as multimodal.
So like being able to use video,
audio, um, pictures and things like
that to kind of pull full context in.
So, um, I think chat, GPT.
You know, this newest version is really
strong and I use it for a lot of like,
kind of collaboration, building, you
know, ideating, things like that.
Um, for search, I still use Perplexity.
Okay.
Um, because I like the format.
Um, and I also like that if I take
that I can pull it into something
like Notebook lm, which will
turn it into a podcast for me.
Mm-hmm.
So that, uh, that kind of helps
me to like, you know, ask the
questions, you know, to your point,
ask the questions, and then still
kind of get some sort of like.
Audio download so that I can
listen to it if I'm in the car.
Yep.
You know, doing other things.
Okay.
So those are my favorites right now.
Have you used gr?
I have used gr, but it's not, to
me, it's not as readily available
as some of the other stuff.
Oh,
Speaker 4: okay.
What that
Bobby Johnson: mean?
So I just, I mean, I just, I, I
also, I'm also, I've started with
chat GPT, so I got used to it.
So I think, uh, I, like, I use
Claude more for like writing.
Um, I think that it, it's able
to understand tone better.
Um, so we're talking about like different
things for different reasons because we're
Adam Smallcombe: talking about who's
gonna win, right, right, right.
Yeah.
Right.
And I feel like who's gonna
win is determining on a lot
of culture and language.
Mm-hmm.
So my girls will always say, I'll chat it.
Speaker 4: Mm.
Adam Smallcombe: And I've noticed that
the young people, the young people,
uh, jail's in the room, the young
people say, oh, let's just chat that.
Arun Koshhy: Oh, interesting.
I didn't even know that's a new thing
Adam Smallcombe: was I always
say, I'm gonna rock it.
Yeah.
Yeah.
Um, so I've never heard that.
Yeah.
Arun Koshhy: Chat it.
I've never heard it.
You don't hang around young people.
Yeah, I know.
You,
Bobby Johnson: you get down.
I've never heard that the young person
in the room said, we don't say that.
No.
She's a little bit older now.
She's older.
She's our generation now.
She's too old.
Are
Speaker 6: you
Bobby Johnson: Gen X?
Yeah.
We were talking earlier
about how, um, perplexity.
Is trying to buy Chrome.
Speaker 4: Mm.
Bobby Johnson: Um, like just recently
in the news, they've said that
they're, they're supposedly set up
to try to buy Chrome from Google.
Oh.
Adam Smallcombe: Because
they have to divest
Speaker 4: Google.
Have to divet.
Oh, right, right, right.
Yeah.
Adam Smallcombe: But, um, so
they're gonna sell Chrome.
Yeah.
So I'd sell that too.
Speaker 4: I mean,
Adam Smallcombe: they
sold anything Chrome.
We, we will get rid of that one.
Yeah.
Bobby Johnson: Yeah.
Are we good?
But powering that with something
like complexity now, because I
feel like people are using it.
As like their answer engine,
their search engine mm-hmm.
Um, I think creates a different
dynamic in the way, like they just
launched their, uh, their browser,
but if they buy Chrome, that's, I
think that's a game changer for them.
So they might actually come up even in,
plus they use everybody else's model.
Yeah.
That's the, that's the thing is
like, what model are people using?
'cause now you can use all of the
AI tools and choose your model.
Speaker 4: Mm-hmm.
Bobby Johnson: So,
Adam Smallcombe: yeah,
Bobby Johnson: and I
don't know that people.
Are paying attention to the difference.
Adam Smallcombe: I don't
think most people care.
Bobby Johnson: Mm-hmm.
Speaker 4: That's
Adam Smallcombe: fair.
I think maybe in the world of, in
your world, it's like you on a daily,
but I think generally people stick to
one because you're not paying like.
14 subscription fees.
Oh yeah.
Right.
These, some of these max
Arun Koshhy: models are like,
it's like three, $200 a month.
Yeah.
$300 a month.
It's crazy.
But it is the jump up when you
are using it like a power user.
Sure.
The jump up is insane.
Like it actually is able
to do so much the world.
It's a huge advantage as they
improve this and it gets cheaper
and cheaper and cheaper to use.
It is gonna be crazy.
How can you access it?
In every, every country.
Sorry?
Adam Smallcombe: Can
you access these Yeah.
Models in every country.
I think Some of
' Arun Koshhy: em, yeah.
Maybe some are geo restricted, but yeah.
Yeah.
Yeah.
How do you keep up with ai?
I was talking to you about this.
It feels like it's changing
almost like daily at this point.
A new model.
A new version releasing.
Yes, it is.
Like we've switched probably three times.
Adam Smallcombe: Are we still saying ai?
Can we just move to SI soon?
Or like what?
Yeah, when it went.
Yeah.
How are you seeing it?
Yeah,
Arun Koshhy: I
Adam Smallcombe: know.
I should check.
Arun Koshhy: Yeah.
Yeah.
What are we doing?
Bobby Johnson: Ask check.
Arun Koshhy: We need a new
Bobby Johnson: chair right here.
Had an episode with chat GPT.
Let's chat that.
Arun Koshhy: Wait, that's, we need
to, we need to, we're gonna do
a chat episode gr we'll do grok.
You have a little personality.
Okay.
Bobby Johnson: Maybe boom.
Arun Koshhy: Maybe both.
Ah, I'm just gonna stay home.
Speaker 6: We'll laugh.
Arun Koshhy: Just three
different models doing a podcast
Speaker 6: episode.
We haven't have we?
We
Arun Koshhy: should.
We should.
That'd be funny.
That'd be fun.
Yeah.
Bobby Johnson: Yeah.
I can't remember what the question was.
Uh,
Arun Koshhy: how do you keep up?
How do you keep up?
Everything's changing.
Well, you're preaching this,
you're getting people to,
you know, kind of adopt ai.
He
Adam Smallcombe: does
Bobby Johnson: this at
work, he just throws head.
Yeah, yeah, yeah.
I, I literally am just, I try
to, I try a bunch of stuff.
Yeah.
So I'm constantly like looking
at what new tools are out
there, what tools are relevant.
Mm-hmm.
Um, I, there are some podcasts
also, like, I try to listen
kind of for different levels.
Mm-hmm.
Like the NVIDIA podcast is actually really
good for like high level industry stuff.
Listen to Hype Pod.
Of course.
Nice.
Of course, of course.
Speaker 6: Where would you get the latest
intact from other than the Hype Pod?
We're asking the question.
Arun Koshhy: Yeah.
Hey, what does AI mean exactly?
This is where people come.
This
Adam Smallcombe: is amazing.
Happy to be your source of information.
Bobby Johnson: I think, uh,
there's a couple other ones.
Uh.
Lenny's newsletter is, is really good.
Uh, everyday ai, just stuff
that like, 'cause I feel like at
different tiers, if you're looking
at like, industry-wide mm-hmm.
Versus like, what's happening just
like out in different parts of the
world versus like, how can I use this
every day to try to build my business?
How can I use this in my personal
life and stuff like that.
I think just right now I'm
just trying to absorb every.
Kind of level of that so that
I can at least speak to it.
And then just constantly using the tools.
Yeah.
Arun Koshhy: Yeah.
That's really important.
You, you talked about
get out there and use it.
You have to get your hands dirty.
Yeah, yeah, yeah.
You have to play with it.
Yeah.
Yeah.
And it's
Bobby Johnson: like, and I think
that that's the other thing is
like not being afraid, uh, first.
Yeah.
I guess not being afraid.
Yeah.
That it's gonna take over the world.
Being aware of it.
Being aware.
You're aware of it.
Hey.
But you know, I also, and this is a
whole different reference, but I say
you can, at some point, somebody's
gonna have to be John Connor.
Mm-hmm.
True.
So true.
If you're gonna be, you know, you have to
know at least enough to know your enemy.
Exactly.
In the second movie, he programmed
the robot to help him later.
So like T two is the best.
Yes, it absolutely is the best.
It's really the only one
that counts, I feel like.
Yeah, it really does.
Seven one's really.
It's really the only ones.
Yeah, it's crazy.
But yeah, I think just going and
like playing around with it and also
understanding like if you have the
understanding that AI is going to
accelerate your experience, then use
your experience for the use case as
to like what you're messing with.
So like for me, I kind of mess around
with a little bit of all of it, but my.
Favorite tool right now has been, uh,
rept, which is an IDE, which you basically
is basically a programmer for you.
So like you, you vibe code, and
you tell it, I want to build this,
I want it to look like this, I
want it to do all of these things.
And then it codes it for you.
It programs it for you, which
has always been my personal wall.
I've never gotten past, like
the first couple of months of
any programming language, right?
Mm-hmm.
But, um, but being able to build an
app in, you know, a couple of hours.
Uh, that does things that I,
I wanted to do is really cool.
Yeah.
Um, N eight N is another new one that's
a lot of people are talking about.
That's an agent, like
an ENT workflow builder.
So it kind of does all of the stuff where
it's, it has all the partnerships with all
of these companies to be able to create
agents that go and do stuff in those.
Tools.
Um, to me, those are the two kind of big
ones that I've been playing with recently.
Mm.
Um, and really it's just getting into
those and they have two week trials.
If you sign up for certain things,
they'll give you a year for free.
So I have a lot of like
different like Wow.
Yeah.
Year subscriptions on stuff.
Yeah.
And then there's a lot of that I'm also
just paying for 'cause Yeah, yeah, yeah.
Because it New World, it's like, it's
like the new, the new like Netflix.
Oh, subscriptions.
Yeah.
Yeah.
You have every single
video platform out there.
But I think
Arun Koshhy: you touched on a
really cool point, and I feel like.
If you take anything away from this
podcast is that you should go try out
these tools because if you, I think the
future is gonna be if I have a need.
Mm-hmm.
I don't need to go to the app
store and find like the app
that kind of fits what I want.
Yeah.
If you can ask AI to build it for you.
Exactly.
With just human language, like I need,
I need a timer that tracks, you know,
every lift that I do, but then also keeps
track of how many I've done over the week.
And instead of going and paying
somebody that build AI can create
a custom app for your use case.
And so it's like, if you have an idea now.
It's like you said, the barriers are down
to get it there for your personal, that
Bobby Johnson: that's
where devices are going.
I think they're gonna become like
these black boxes where I talk
to it and it starts to figure out
like, okay, this is what I need.
Mm-hmm.
I create an it creates an app for you.
Like I don't know that we're gonna
have like a lot of the, like the.
The big app movement that happened,
you know, back when iPhone first
came out and stuff like that.
I think that's gonna shift.
Mm-hmm.
Where like the devices are really
just gonna be like, how strong, how
much I AI can you like pack into it so
that it can actually start to create
the things that you want it to do.
Yeah.
So I say I want it to, I want
to, you know, it'll probably
have basic stuff, but if I want
something that like I wanna watch.
These, you know, from, I want to watch
my favorite shows, um, from Netflix
and Prime and HBO Max or whatever.
Yeah, yeah.
Like figuring out like, oh, well
you have these subscriptions.
I'm taking these subscriptions,
pulling them all together.
Make your own personal, it'll be like
a personalized, right, right, right.
Like device that'll be so personalized
that like all the apps will
basically just like cater to you.
I think where it's going right.
Arun Koshhy: And it goes to
the future that you want.
You want AI robotics?
No, I, you do all this stuff
that you don't want do.
I don't want
Adam Smallcombe: an agent.
I want an employee.
Yeah, exactly.
That's what I want.
Uh, but would you pay the employee?
No, no, no, no.
Definitely not.
That's the advantage is that
they're making me money.
Right, right.
Right.
I'm not paying them.
They're paying me.
Right.
For the pleasure of working for me.
Uh, but I think for, you know,
okay, so I think going back to
the, the concerning side of things.
Yeah.
I don't know why I'm playing
the, I'm never the concerned.
No, it's good.
This is good.
But right now it seems like AI
is growing at the pace of chip
speed, uh, energy availability.
That's the restrictor and the
limiter on super intelligence right.
Now.
What happens when we unlock
that to unlimited potential?
Or is there always a limitation
which is, uh, readily available,
chips speed, uh, processes,
supercomputers, those kinds of things?
Is that our limiter?
In our safeguard.
Arun Koshhy: No, it's,
it's gonna keep growing.
It's like it's exponential at this point.
We've hit like, they call it the banana
zone, but like if you look at the
way that this is growing now, you're
using AI models to help you figure
out how to solve those problems that
were kind of unsolvable before, right?
Mm-hmm.
And it's computing that
so much faster now.
This cycle is just becoming
much and much, much quicker.
So now the exponential kind of growth
of how much these models can do.
Is is insane.
So like, it's like every week it feels
like there's advancements on top of this.
So the limitations that we see
now are gonna be gone in like
a month, two months in a year.
That's what I'm saying.
Yeah.
So now we're gonna get to that point
where, you know, you're talking about
artificial general intelligence, right?
And are people in
Adam Smallcombe: your world
scared to go on vacation in case
you miss something in a week?
Ah, absolutely.
People are like freaked out.
Yeah.
If I go on vacation, I'm gonna miss,
yeah, no, I think I need to learn how
Bobby Johnson: to start farming as well.
'cause at some point, if, if
it turns into the utopia that.
Is, there's, so you have the,
the, the apocalyptic, you know,
view, but you also have the utopia
if it turns into the utopia.
Mm-hmm.
Or if it turns into the apocalyptic zone.
Yeah.
Like learning how to do
the natural things like.
Plant and farm probably will be helpful.
I don't, oh yeah.
I know how to do that, guys.
How to do that.
That's my world.
I'm from Ohio.
I don't dunno anything about that.
Yeah.
So I need to learn that.
But I can ask Chad, GPT.
So,
Arun Koshhy: but speaking of that utopian
world, if we wanna build it that way, the
right people need to be behind it, right?
Because like there's already, and I don't
want to say the name of the tools because.
Okay.
You know, I don't wanna promote
it, but there are tools that
do the exact opposite, right?
There's a chat GBT version that's
purely out there to help you enhance
your ability to fraud or scam people.
Yeah.
And like those models exist, right?
And so I think it's, it's onus
on us, the people that think, you
know, that know that they're doing
the right things for the right
reasons, to build out these things,
to get in this space, to get there.
'cause this is a future.
I mean, we, we all know it.
This is a future.
So it's just a matter of.
How do we impact that space?
Right?
Yeah.
And so we're not gonna get replaced by ai.
We're gonna get replaced by people
that know how to use ai, right.
And know how to use these tools.
Or
Speaker 4: robots.
Arun Koshhy: Or robots, but like Exactly.
But like also the ones that know
how to kind of do all these things.
Robots.
Yeah.
So, um, yeah, I feel like maybe
we'll just wrap it up there.
We'll, we'll kind of, we'll
kind of, not to doom and gloom
kind of space, but it's a,
Adam Smallcombe: it's a public space.
Oh.
I mean, I've had fun taking that role.
Yeah.
Yeah.
Arun Koshhy: You've shifted a little bit.
I shifted.
Did something happen?
Should we, should we, this
therapy session a bad experience?
No, but, um, we'll fix it.
Adam Smallcombe: It was grok in the car.
It was when it told you that.
Yeah.
You were wrong.
You was trying to speak Italian
and I had a fight with you.
You were like, I hate you, stubborn.
I hate you.
Ai.
Your daughter loves it.
I hate it.
No kidding.
Arun Koshhy: Yeah, yeah, yeah.
But thank you Pastor Bobby
for joining us on the high.
Adam Smallcombe: What a treat.
Arun Koshhy: And thank you
for educating people on ai.
'cause we weren't doing
a good job before this.
Well, they came here for it.
They came here and now they finally got
Adam Smallcombe: educated.
Hey Arun, we delivered.
Speaker 6: Yeah, yeah, we did.
We did it.
50 plus episodes in they finally know ai.
Arun Koshhy: But if this is your
first time, pass it along to
the other person that wants to
know more about ai, share it.
Like, comment, subscribe, and
we'll see you in the next episode.
Cheers guys.
![[Ep 57] AI, Faith & The Human Edge with Bobby Johnson](https://img.transistor.fm/gNn-27S39xnFICPOFbTRanmN7O0kVI2YPpFtPlsWS9k/rs:fill:0:0:1/w:800/h:800/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9zaG93/LzM4MzkyLzE2ODY4/NjM0MzgtYXJ0d29y/ay5qcGc.webp)