July 21, 2025 • 38 mins
Ryan and Will talk about the growing role of AI in startups. They explore how AI is automating traditional founder tasks like writing, coding, and analyzing data. The conversation delves into the existential threat AI poses to the founder's role and how founders can adapt to this new landscape. They argue that founders should embrace AI to focus on creativity, curation, and leadership rather than being bogged down by routine tasks. The episode underscores the importance of evolving with technology to enhance productivity and innovation, ultimately redefining the founder's job description for the future.
Resources:
Startup Therapy Podcast
https://www.startups.com/community/startup-therapy
Website
https://www.startups.com/begin
LinkedIn
https://www.linkedin.com/company/startups-co/
Join our Network of Top Founders
Wil Schroter
https://www.linkedin.com/in/wilschroter/
Ryan Rutan
https://www.linkedin.com/in/ryan-rutan/
What to listen for:
00:34 Can AI Replace Founders?
01:16 The Evolution of AI Capabilities
02:03 AI's Impact on Copywriting and Creativity
03:28 AI in Advertising and Idea Generation
04:44 Embracing AI for Efficiency
06:31 AI as a Force Multiplier
09:44 The Future of AI and Founders
14:36 AI in Coding and Development
17:08 Leveraging AI for Business Growth
19:13 Simulating Business Scenarios with AI
19:38 The Role of AI in Business Decision Making
19:53 The Timeless Nature of Business Challenges
20:48 Embracing AI for Efficiency
21:12 The Unique Human Element in Business
23:40 The Future of Founders in an AI-Driven World
25:50 The Evolution of Founders' Roles
28:05 Leadership and Determination in Business
30:47 The Importance of Vision and Adaptability
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Welcome back to the episode ofthe Startup Therapy Podcast.
This is Ryan Rutanjoint, as always by my
friend, the founder,and CEO of startups.com.
Will Schroeder, will AIis now writing our copy,
debugging our code, AB testingour pricing while we sleep,
to the extent that we do.
So the existential questionsgetting pretty loud and clear.
(00:21):
If the smartest person inthe room now costs 49 bucks a
month, what's left for founders?
So I think, man, it'd be funtoday to break down like what
parts of us get automatedand which parts become
more valuable than ever.
The
question a lot of peopleare saying is, can
AI replace a founder?
Yeah.
And, and I, I'm gonna beheretical when I say this.
Yes.
(00:41):
Now I say that as a founder,as somebody who does not
want to be replaced, right?
Yeah.
So I wanna be clearly, I, Ihave a strong bias to say,
no, this is where we fightback against Skynet will.
It kind of is, weknow how this goes.
It doesn't end well for us.
Although we do get aa, a cool Austrian sent
back to it to help us.
You sure do.
So here, here'swhat I would say.
When I say yes, I don't mean,uh, universally all of us,
(01:03):
I'm saying Yeah, yeah, yeah.
A lot of, I think whathistorically I considered
my founder traits.
Um, my ability to workreally hard on all these
different things, those thingsjust aren't hard anymore.
Right.
Like, I'll give you an example.
I've prided myself fora very long time on
being a strong writer.
You've seen me write unlimitedamounts of copy, right?
Yep.
And it was like we could justtake any topic, uh, whether
(01:25):
it's landing page copy, oreducational copy, or your
newsletters, and you'relike, we'll, write something.
And I would just, right?
Yep.
And, and like I said in,and my writing was decent
and it was passable.
Yeah.
But now.
AI can write as well, ifnot better than I can.
And it's almost like that skill.
Okay.
This is just one examplekind of went away.
(01:47):
Now the, the importanceof having something to say
has not gone away here.
Right, right.
What you're trying, tryingto say part very much
I.
But your ability to conveythat in words is no longer a
unique skill that some peoplehave and some people don't.
Some people, of course,are gonna be better at it,
but no, if, if chat GBTcan crank out 50 landing
page variants before lunch,then we can't treat the
(02:09):
copywriting piece as thefounder superpower, right?
Knowing what to say, important.
Saying it less.
So I wanna dig into thatjust a little bit more.
'cause I, this is just sofascinating to me that this
whole, this whole era rightnow is so fascinating to me.
Right now we're at a pointwhere AI, and, and I, I would
(02:29):
argue, I always use the,the corollary to, uh, gaming
machines from Atari 2,600 Yeah.
To like, you know, uh,current Xbox, so to speak.
And I, and I would say AI isstill in roughly somewhere
between the Atari 2,600 andthe in television stage, right.
Just like the Right,the very early stages.
And at the time peoplenever see it that way.
Like, like when you had yourAtari 2,600, you thought it
(02:50):
was the most amazing gamingsystem that it e ever existed.
Well, it was becauseit was right.
'cause right before that Ihad pong, and right before
that I had just tic-tactoe on a piece of paper.
You bet, you bet.
So I say this to saythere are many, many, um,
iterations that are coming,but even in the current
iteration, at the most basic.
What AI can do now, it makesa lot of the stuff we were
(03:12):
spending so much time doing.
Kind of irrelevant or at worse.
Yeah, man.
Duplicative.
Yeah.
Yeah.
Anything that's repeatable,data-driven, pattern based.
To me, man, that'salready AI turf.
Right.
And as you said, we're justat the beginning stages.
And
here's anotherway to look at it.
I used to run a largead agency, right?
Yeah.
And we had hundreds and hundredsof people, uh, copywriters
(03:33):
and creative directors.
Yeah.
You know, and allthese folks, right.
And the copywriters, forexample, of which I had
a, a strong affinity for,would be able to sit in a
room and we'd be working ona campaign, and they'd come
up with dozens and dozens ofdifferent ideas and approaches.
Right now, within thatcontingent of copywriters.
Like anything else,like athletes or
anything else like that?
There were a few thatwere exceptional.
(03:54):
Let's say the A players.
Yep.
And there were a handful thatwere B players and a lot of
people that were C players.
Okay.
Yeah.
Now what I would argue ismost AI can perform about
as well as the B players andmost certainly the C players.
Yes.
Yeah, I agree.
Are they replicatingthe A players?
No.
And they, they may never.
(04:14):
May never.
Okay.
However, that's here nor there.
The point is, if I can nowhave copywriting and and idea
generation done at a B level.
Yeah, for $0 for$49 a month, right?
Yep.
Why wouldn't I
be doing that?
Exactly.
No.
The accelerative power,that is insanity.
Right?
And this is, this is whatwe should all want, right?
(04:34):
This is what we all want.
Correct.
We've always wanted this.
We're like, how could I justget them to produce way more?
Right?
I can only afford one person.
How can I turn them into 10?
Here's your answer.
And again,
we're focusing this episodeinward to say, and what if I
wanna be the force multiplier?
Here's a, a great way for meto, to, to show the difference.
Uh, a few years ago when I waswriting copy for something,
(04:55):
I kind of knew what my rateof output would be, and I
knew that, let's say I'mjust making this up on, in
any one day, I could create,uh, you know, a five page
document, just, you know,using it as a unit of measure.
Okay, well now in a day I cancreate 10, five page documents.
Yes.
Right.
So because the AI can basicallytake the ideas from my head
(05:15):
and just put all the wordsthat was like me manually lay
laying bricks before, right?
Yeah, exactly.
And I just don'thave to do that.
No.
So part of me would say, man,that felt very special to me.
You know, I felt very uniquethat I had that capability.
Uh, and, and, and that's beenquote taken away from me,
but there's another side tosay, but was that really was
(05:38):
my ability to, to, to writewords and compose the value?
Or was it the ideabehind those words?
Yeah, I mean thatwas the thought.
The always was the idea.
We just didn't haveanother way of manifesting
the idea in the past.
Right.
There was have idea.
Now put it to paper.
Right.
We just don't have to do allof the manual work for that.
It's funny, man.
I had this realization theother day that there are tasks
(05:59):
that I think once were likethe signals of my genius, that
now would just be a signal thatI lost my open AI password.
Like nothing else.
Like if you see me doingthem, you're like, wait,
where I did the power off?
Like what are you doing?
And again, like to your point,like things that used to seem
really, really important to me,like digging through analytics
and pattern spotting, right?
Right.
Seeing things otherpeople couldn't see.
(06:21):
I'm good at that.
Yeah.
I'm nowhere near as good atthat as a machine that can
crunch through millions ofrecords of data in minutes.
Um, just not, so
I think this is thefirst building block
of this conversation.
I think let's first explorewhat should be replaceable.
Right.
So I know everyone has this,this I immediate reaction
and it's visceral to say,I can't be replaced by
(06:43):
machine, blah, blah, blah.
I'm like, dude, you can, right?
Yeah.
And you will be.
So let's talk aboutwhy you should be and
what you do about it.
Right.
Sure.
Not just like this, thisnatural tendency to just push
back and just be like angryabout it that buys you nothing.
Yep.
And there is a very, very,very long list of tombstones
of people that had exactlythat reaction and didn't
(07:05):
work out well, particularlyin the business world.
So again, Ryan, when youthink about things that are
replaceable, that shouldbe replaceable, that like,
while, like you just mentionedanalytics, that a moment
ago you were like, Hey,this is what I got paid to
do, and now it's like it'sa $49 a month, uh, problem.
Yep.
What else comes to mind?
It's the, the
time consuming part of it.
You can just look at itfrom a, a pure time calculus
(07:26):
perspective, like, and replacethe things that are gonna
take you the most time andstill, like, they're valuable.
They, they need to bedone, but they don't
need to be done by you.
You know, as we were cominginto this episode today, will,
I was, I was thinking aboutthe process you've gone through
with building the house.
Mm-hmm.
Right.
And it occurred to me like, youknow, you at the beginning, like
you could have treated yourselfas the architect, the GC.
(07:49):
The plumber, the electrician,the framing guys.
Right.
All of it.
Right.
You could've just done allof it and you've done, you've
done a lot of stuff, but atsome point, like it makes
a lot more sense for you tohire a GC and a plumber and
electrician and all of that.
Right?
Right.
Well, I think that, you know,we were once as founders,
we were once the architect,GC plumber, all of it.
Like we just hadto do everything.
We got so used to, especiallyat the time you and I
(08:11):
came up through this.
Yeah.
We had to do everything.
There wasn't anybodyelse to do it.
You really did.
Yep.
There weren't tools to do stuff.
We did a littleof everything now.
AI steps in as the GC andAI just hired the plumber.
Yep.
Now you can just stay inthe architect's chair.
Right.
That's the way I see it.
So what, what I'm looking atis like, how can I just stay
(08:31):
a above that kind of messywork layer in a way that
doesn't keep me uninformed,but doesn't have me overworked
in the sense that like I couldbe doing more valuable stuff.
So instead of havingto crunch all the
analytics like I used to.
I can get the insight fromthat and I can spend more
time deciding what to dobased on that insight.
Right, right, right.
So the, the, the base levelanalysis, all of those kind
(08:52):
of things just going away,anything that requires like
mass creation or mass ingestionand kind of normalizing of
information just gonna go away.
Right.
So at the marketing level,looking at across, like, you
know, if we're honest withourselves, we talk about AB
testing and, and you know,optimization and all this
stuff that we, that we do.
(09:13):
Yet, it's sort of done on kindof like a rolling window basis.
Mm-hmm.
Do I look all the way backto the very first experiments
I ran on Facebook ever.
When I'm, yeah.
When I'm running, runningthe ones now, yeah.
I can't, I can'tlook at 6,000 ads.
AI can, so now it allows meto do these things better.
And it was never aboutthe analysis before.
(09:34):
It was about gettingto the answer and then
acting on the answer.
Right.
And I still think thatthe acting on the answer
piece is still very much.
In founder camp.
We'll see for how long?
I
think a big part of where thepushback is, is there's an
overwhelming focus, justifiablyso right now around what
AI is taking away from us.
There's not enough of theconversation, and I think the
(09:56):
smart founders are alreadypointing this direction around
what it is enabling us to do.
Not that nobody's evermentioned what, what AI can do.
Yeah.
What I'm saying is asfounders, as founders,
there has to be a versionof us that says, dude, like.
I'm about to be Superman.
Right.
Yeah.
Like for example, when Iwas first starting in my
career, I didn't understandanything about finance.
(10:17):
I mean, not even a little bit.
And as you know, I've been astartup CFO for 25 years in
addition to my, my day job.
And it's not because I'ma financial wizard, it's
because as soon as you putdollar signs in front of
those numbers and you said,Hey, that's your money.
Yeah.
All of a sudden I was like the,the rainman of finance, right?
Like I was just like,I, I understood it.
All right?
The matrix, it all made sense.
(10:38):
But think of how under.
Armed.
I was at that time, right?
He here, here's how Ifigured out finance.
I drove to Barnes and Noble.
I found a book that was probablypublished 15 years prior
on small business finance.
And I read it cover to cover,which took days, right?
Yes.
And it was the driest,most boring book ever.
(11:00):
However, changed my life.
But think of all the thingsthat that had to happen in
order for me to be able todo this one simple thing,
which was put together anincome statement, right?
Yeah.
Yeah.
Now I can do that like aseasoned pro without even
knowing how to do it.
I just, I have to go back tothis anecdote where, where
you talked about, where youwere asked for the income
(11:20):
statement and you, you pulledout the, uh, you pulled the
piece of paper outta yourpocket that just had like the
monthly, like the things youhad spent money on that month
and like showed the investor.
I love that man.
This, that was, that was,this is the moment where
that education becameobvious and necessary
to you.
That would've been reallyuseful at that time.
Hey, what's your proforma income statement?
My, my what?
My what?
(11:40):
My, um, here's a listof things I'm gonna
buy at the store later.
That the team, yes.
I literally gave a, a worddocument of, of how much I
owed my roommate that month.
Yeah.
But again, you know,plan it out a little bit.
So we keep talkingabout the things that's
taking away from us.
But when I say, you know, someof these traits, you know,
are replaceable, I'm sayingnot only are they replaceable,
(12:02):
we should lean into that.
Yeah, right.
A hundred percent.
My first reaction being AIis taking away my uniqueness
of being a copywriter.
Okay, got it.
Okay.
But how can AI also makeme 10 x the copywriter?
Right.
Yeah.
If, if, if I have so muchto say, how can I say it
exponentially faster, moreefficiently, et cetera.
(12:23):
Like, I've got an AI that'strained on my voice so I can
tell it anything I want andit will respond in my voice.
Now I don't love it.
Right.
I don't love it.
So, so like, uh, when I useda brainstorm stuff, it'll come
back in my voice, but it's, asyou know, AI has a smell to it.
Yeah.
Yeah.
So,
so it's always likethe, the kind of.
AI version of my voice.
(12:44):
Yeah,
it's 80%, but it ain'ta hundred percent.
You mean auditory voice?
You're not saying writesomething in your voice
and tone, you mean?
Oh yeah.
Your actual voice point.
Good point.
Actually, the auditoryvoice is spot on.
No, in this case, I mean the,uh, the, the, the writing
and, and I gotta say, man,uh, what I get back if I
published it as is, no onewould know the difference.
I'm the only person that cantell the difference in this
case, but I can tell anyway.
(13:05):
Maybe you just don't likeyour writing will maybe that
I, I I always used to make thejoke that if we ever come up
with an AI that can replicateyourself, that will be the
most hated person of all time.
Oh, yeah.
Or, or AI hang out withyourself all day long.
Oh, how about this?
God.
An AI that is purely basedon your search history.
Yes.
That person, the worsthuman of all time.
(13:27):
Oh God, I hate that person.
Anyway, I think that again,we're having that first initial
reaction that says, you know,AI pad, blah, blah, blah.
Um, we might wannalook at it going.
Should you have had to knowall of finance in order
to advance your business.
Yep.
And there's gonna be someargument where someone says Yes,
but that knowledge and what youwent through Uhhuh, you know,
(13:47):
taught you to do X, Y, Z, and,and there's some truth to that.
Sure.
But dude, that is a reallypainful, hard way to me,
like 10 years to do it.
Right.
I needed that informationday one, not 10
years from now.
This is here, here's,this is a big question.
I don't think anybody cananswer this just yet, but
one of the questions is.
How will this be different forpeople who didn't come up, who
(14:08):
didn't have those hard lessons?
So if you're just born now Yeah.
And you, you justgrow up an AI native.
Yeah.
You just have alwayshad this stuff at hand.
You're always using it.
How does that change yourability to use it to see
the difference between the,the, the good, the bad.
To know that difference?
Like you talked before aboutit A level, B level, C level.
Yep.
If you've never worked witha bunch of copywriters and
seen a, B and C level copy.
(14:30):
How will you know when AI'sactually doing the right thing?
Right.
You won't.
I, I, I actually justwent through this, right.
You know, as I'm designingmy house, I use a tool called
SketchUp, you know, which Iput in the 3D model Within
that I have to, I built all thecabinetry in the house and all
the, the closets and everything.
Yeah.
And I had to builddrawers, you know, for,
for every one of those.
Well, in order to go intoSketchUp and create what, what
(14:50):
amounted to 88 drawers to sizethem, create the 3D geometry
and whatever, and get 'em allset up, would've taken forever.
Okay, so this is a perfectexample of me coming from
a space that I'm not in andbeing able to do something.
There's no way I would'veever been able to do.
I go to GPT and I said, here'sa list of all the drawers that
I need, just like the, thewidths and the, uh, depths.
(15:11):
I said, make meruby code, right?
That I can import into SketchUp.
That will automaticallycreate all of the drawers
that would, that wouldfit for these dimensions.
Give them a, a notch and adata and you know, all this
stuff, and create all thegeometry, the full model.
Right.
Five seconds now.
Okay.
Nuts, dude.
Unbelievable, right?
(15:32):
Yeah.
Would've taken me a week.
Okay.
Yeah, and I've got, Iwould've gotten it wrong.
Now, a couple things.
Number one, 99% chance, there'ssomething in there that's wrong.
There's a drawer witheight sides to it or
something like that.
Yeah.
GT's.
Just not very accurate.
So it helps to knowwhat you're looking for.
But here's my point.
This is the samewith with finance.
(15:52):
What it would've taken me.
To perform the same operation.
Yeah.
To learn Ruby code, right?
Yeah.
To recheck all of my math,to be able to sit and see, I
mean, like it would've takenme, uh, forget the Ruby Code.
It would take me aweek to do it manually.
Right?
Yeah.
And longer to learn Ruby Code.
And someone might say, yeah,but then you'd know Ruby Code
and here's what I would say.
(16:12):
Cool.
I may never need that again.
Right.
I just needed todo that one thing.
All my action script andcold fusion knowledge
is getting used.
Not at all at this point, right.
I'm just say love that stuff.
Like
I don't necessarily, like, Iget learning things and again,
I, I love learning things.
So they we're in a platformwhere we teach people
stuff, but not everythinga founder needs to do.
(16:35):
They need to be theeducated expert in all the
eccentricities of that.
It doesn't matter ifyou know a, B or C grade
copy, you just need copy.
Yeah.
And if it doesn't soundlike total shit, it's gonna
work.
It's gonna be agood starting point.
At least you canalways adjust it.
Yeah.
'cause the reality is we'renot gonna know whether it's
good copy or not until itfaces some other humans.
Mm-hmm.
Or maybe some AI andgets picked up and search
(16:55):
results and whatnot.
Right, right.
Yeah.
So, you know, from mystandpoint, I look at all of
these things that I can now do.
That I couldn't dofive minutes ago.
Okay.
Yeah.
So for example, the big thingnow is people can write code.
Yeah.
Right?
Which is, so, I mean, Ryan, youand I have haven't written a
line of code in 20 years, right?
Our knowledge is so far fromuseful at this point, but now I
(17:16):
can write code in any languagefor any purpose I can get
MVPs put together, et cetera.
How is that not
a force multiplier?
The time it takes me tocreate a landing page now.
Right now, not aread to chip, right?
It's gotta go over to Dev andthey gotta do their stuff track.
But the time it takes me tocreate a landing page now is
less than the time it would'vetaken me to go through the
(17:37):
drawer to find the pencilto do the sketch on paper.
Right?
It's bananas.
How much faster we can moveand how we can get onto like
the important parts, right?
Because the important partwasn't drafting the wire frame.
The important part isgetting that page launched.
Getting traffic point ofthat page and starting to
then have AI analyze theresults and tell me what
(17:58):
still needs to be changed.
Fantastic.
Right?
This is what you want.
It's short circuits, somany tedious processes.
The number of hard choicesI have to make now.
Are so significantly lowerbecause it's like, well, I
can do this, or I could dothat, or I could do that,
or I could do that one.
What are we gonnaget to this week?
All we'll just do,just yes is the answer.
(18:20):
Now, I love being able to justsay, yes, come on, why not?
Right?
Like, oh my God, in, in mygetting customers workshops
on Mondays and Fridays.
The number of requests I can nowfield people that need help with
really heavy lifting stuff werehonestly, most of the lifting
was on their part, but I stillhad to show 'em how to do it.
Yep.
This is part of where I'mjust like automating and,
and AIing like crazy.
(18:40):
Like what am I asking foundersto do over and over and
over again That not onlyare they resistant to, but I
know they desperately need.
Cool.
Let's AI the shit outta thatbecause having something done is
better than having nothing done.
Right.
And in a lot of cases it'sjust motivating them to do
more of sometimes the manualwork, but it shows them the
power of what that thing is.
If, if anything, it just serveslike a little simulation.
(19:01):
IP, or like, you know, I'm ahuge fan of the Mom test, right?
Yeah.
Love that book.
Please go buy the book.
Rob Fitzpatrick dida great job with it.
But the process of goingthrough that, for a lot of
founders, they're like, I'mjust not gonna go do that,
or I'm not gonna go do it.
Right?
It's me going to BarnesNoble and buying that book.
Right?
It them, yeah, yeah, yeah.
I simulate it forthem in 15 minutes.
Then they're like, oh,that's what this tells me.
(19:23):
Yeah, that's what thistells you now, right?
You can trust this to adegree 'cause it's based on
the average of the internet.
It'll get most of it right?
But you should also pick upthe phone and talk to a couple
of your, your, your peopleand get the actual answers.
And it's been a greatmotivational factor,
which is awesome.
So if.
Like, and, and here's thething, every time I do
that, there isn't a foundergoing, but Ryan, aren't
you replacing a part of me?
(19:44):
Yeah, a part of me thatyou didn't want to exist
in the first place.
Right?
Alright.
Right.
Nobody's upset about it.
Nobody.
They're all like, great.
I can get onto thepart where we turn that
into useful informationto sell people stuff.
You know, something that's
really funny about everythingwe talk about here is
that none of it is new.
Everything you're dealingwith right now has been done a
thousand times before you, whichmeans the answer already exists.
(20:06):
You may just not know it.
But that's okay.
That's kind of whatwe're here to do.
We talk about this stuff onthe show, but we actually
solve these problems alldayLong@groups.startups.com.
So if any of this soundsfamiliar, stop guessing
about what to do, let us justgive you the answers to the
test and be done with it.
As a carpenter myself, uh,whenever I'm, you know, building
(20:28):
a house, I think to myself,there's no carpenter ever
that's thinking to themselves.
I'm so pissed off that you gaveme a, a powered circular saw.
My hand saw was so much better.
Yeah.
Dick, come on.
Yeah.
Just rubbing my head withoutblisters on my hands just
doesn't feel the same.
Yeah.
You know, like,need those calluses.
All
right.
So we've definitely said yes.
(20:48):
There's tons of stuff thatcan be replaced, right?
Yeah.
And, and we're almostsaying, and should be.
Right.
Like, like do not resist.
In this case, a lot of thestuff, not everything, and
we're about to go the otherdirection, a lot of this
stuff should be replaced.
We should embrace thereplacing like we're shedding
skin because the worldjust got way easier for us.
Yeah.
But let's talk aboutthe other side.
(21:09):
You know, we said the,the topic of this is
can AI replace founders?
But let's talk aboutthe other side.
What makes us unique?
What is AI notlikely to replace?
And I wanna caveat this bysaying there is a sci-fi
version of everything.
Where you could say and makean argument, and I'm one
of the people that can makethis argument that would
say, no matter how hard youtry to say that humans are
(21:29):
always going to be unique.
We are an algorithm, andalgorithms can be replicated.
Right?
So I, I wanna be clear when Isay this, we're not, Ryan and
I aren't sitting here beinglike humanity is, you know,
irreplaceable in any way.
That like, we, we recognize thatat some extreme level there's
probably an AI that, that can dosome things, but at a practical
level, at a level that actuallymakes sense for, for either now
(21:50):
or in, in the very near future.
I don't believe that AI cando some really human things.
Sure.
And I don't believe it should.
I think, you know, if, ifwe're saying AI can create
music, but should it, right?
Yeah.
Yes.
It can create something Right.
But the point of musicwasn't to make the music.
It was a form of expression,
(22:11):
which usually comes from aplace of need to express Right.
AI will never needto express itself.
Correct.
It doesn't takeanything away from that.
Correct.
Right.
And I think now there'ssomething really fascinating
about how we get touse our uniqueness.
Let's say again, I'm gonnago back to me as a writer,
imagine for a second.
There are two componentsof me being a writer.
The part where I can generatea unique thought that people
(22:33):
might give a shit about,and the part where I can
translate that to words thatpeople might wanna read.
Yeah.
Okay.
The latter has beenmostly replaced, right?
Yeah.
Just about any thought Ihave can be translated to
words that are 90% of whatI would've used anyway.
Okay.
As weird as thatis, it's It's true.
But the unique thought.
This, this unique idea,this, this, this unique
(22:54):
opinion, et cetera.
AI can come up with variants.
They can't come up withmy variant necessarily.
Right?
So what happens is my voice now,I mean my voice vis-a-vis the
work that I do, the things thatI produce in life, et cetera,
becomes force multiplied.
Yeah.
If before I could, my opinioncould generate one article, now
(23:15):
my opinions can generate 50.
Right now they can generatea goddamn movie in the
not to distant future.
That is amazing.
But in order for all that tohappen, the uniqueness has to
be me having something to say.
Yeah.
It starts with that, right?
Like the, the things thatit's, you know, the, the key to
differentiators at this pointare, you know, like vision,
narrative, human urgency,moral compass, like these
(23:36):
things can't be replaced yet.
Humor still notparticularly good at that.
Right.
Um, but I wonder to some degree,like you said, that like, it
can't come with my thought, but.
Part of that, how, howmuch of that is, is the way
that we use it right now?
Mm-hmm.
Right
Now, I have testedthis a little bit.
I played around with like,instead of saying, here's
(23:58):
what I think, right?
The rest of this, it'ssort of like, here's
the thing that happened.
What do you think myreaction to this is?
Right.
It's interesting.
It's really interestingand, and basically like,
what do I think about this?
What do you think Ithink about this based on
what you know about me?
What do you think?
I think because we've bothgiven AI plenty of information
about ourselves, and certainlyit learns all the time as
(24:18):
we're, as we're utilizing it,but to your point, it comes up
with some, some possibilities.
I, I would say it's still.
Generally not close.
But then I wonder like howmuch of that is my own bias
saying that's not exactly whatI would've thought about that.
Or, well, when I thoughtabout it already, that's
not what I thought about it.
But how about this?
What if we were takingthis a step further and
we're to say, what mattersnow is what you want.
(24:41):
Not what the AI mightrespond to mm-hmm.
But what you specifically want.
In other words, I can gointo an AI right now and I
can say, give me ideas forwhat startups might want.
Okay.
And it'll, it'll comeup with tons of ideas.
And, and I would say it'llcome up with probably
75% of the ideas that Iwould've come up with.
Right.
Just enough darts at a wall.
Right.
Like, uh, you know, whatever.
(25:01):
But the, and, and, and oneof them might've matched
what I was gonna say.
That's perfectly fine.
Right.
But the point is,I'm the curator.
I'm the one thatsays, no, not that.
Yes, that.
No, not that.
Yes, that.
My vision, what Iwant is what matters.
That goes back to music.
Yes, AI can create someversion of music, but the
composer is what matters.
(25:21):
The composer decides right.
Is that what they want?
Right.
It's not like, uh,I'm a composer.
The AI creates a song andI'm like, I hate it, but
I have to have it anyway.
Right.
The whole point is, yeah, weare, we are the final decision.
We, you know, we are, to yourpoint earlier, the architect
of all of this we're the onethat our curatorial ability
to be able to say, yes,this, yes, this, yes, this,
(25:44):
whether it's people, product,marketing, copy, et cetera,
is what makes us unique.
Yeah, I mean at this point, likego back to some of our other
analogies, like AI can draftall the blueprints, but like
it can't convince the city tobuild the skyline, can't it?
Right.
Like
Right.
Or just have the vision to whyyou'd wanna do it to begin with.
(26:05):
Right.
What I'm saying iswhat makes us unique is
this curatorial vision.
What doesn't make usunique is laying all the
bricks to do it anymore.
No, I know.
So again, going back to to,to my, my place in the world
as a writer, which I'vegiven a lot of thought to.
What makes me uniqueis the fact that I have
a lot of unique ideas.
Right, or just, or unique to me,whether anybody else has thought
(26:26):
'em before, who cares, right?
Unique to me.
But now I no longer havethe time element, the
writer's block element,et cetera, that prevents
me from expressing them.
Right.
So whereas before, you know, Iread a, a weekly newsletter that
you, that we published, that we,that we do this podcast based
on before, I'm like, okay, ittakes me x amount of time to
take whatever unique idea I had.
(26:48):
Like, can I replace foundersand convert that into, uh, copy.
Now it still takes meroughly the same amount of
time, uh, because I liketo play with the idea.
I like to wrestle with the idea.
It's not like I'm just tryingto write a, a newsletter.
Like I try to, like,I, it's very cathartic
for me, but ultimately.
It's my idea, right?
Uh, you know, that I'm breakingAI can gimme 50 variants of,
(27:10):
of how AI would talk aboutit, but I'm still the curator.
I'm still the one to say,yes, this, no that, yes,
this, no, that it's my job.
And this is, this is a metaphorfor building a startup.
It's still my job toorchestrate all of this.
Well, it always was, right?
It's, it's funny to me that wedon't think about it in these
same ways because in a lot ofcases, at the early stages, as
the founder, we, we start bydoing everything, and then the
(27:31):
second we can afford it, westart to hire out specialists.
We start to hire out otherpeople, sometimes not even
specialists, just peoplecan do the work, right?
If it's just rote work, youhire people capable of rote work
and they do the work, right?
You don't hire a bricklayer.
If you're building a 6,000square foot high house,
you hire a bunch of them.
Right?
Right.
So we've always had thisconcept that like, you know,
we'll, we'll hire it out.
We'll hire out the work.
(27:52):
We'll, we'll divide the labor.
We just have a differentability to do that.
Right.
GPT just became the assemblyline to our Henry Ford.
Right.
We got the same 24hours of work in.
Yep.
But we've got a hundredx the output now.
You bet.
You bet.
Now I also think anotheraspect which, which I, I don't
wanna overlook or not getinto, is just the fundamental
concept of leadership.
Be it a mixture of pureanxiety, like pure,
(28:16):
unadulterated, anxiety,excitement in fervor, right?
Yeah.
Vision, which I think is aterm that's overused, but it's,
it's certainly a part of this.
Right.
And just fucking determination.
Determination.
Just be able tosay determination,
deciding that we are, we'regoing to keep doing this
until we get it right.
Correct.
AI's not
gonna care.
Yeah.
Right.
And, and nor should it.
(28:36):
Right.
That's the point though.
Like if you look behind anysuccessful company and frankly
many unsuccessful ones, whatyou have behind this company is
a freakishly determined person.
Yeah.
To see this through,despite all logic.
Right.
That's what makes foundersthe most unique, is that
we are able to run in, runnaked, into the abyss despite.
(28:59):
Everything that wouldtell us otherwise.
Yes.
And figure out how to somehowcome out on the other side.
I'll give you an example.
Uh, you know, one of myfavorite would be both
Airbnb and Uber, right?
Two famously successfulcompanies with the
dumbest ideas ever.
Okay?
When I say dumb, I, I, I,obviously, I, I'm kidding when
I say this, but when those poorbastards had to go out there
into the market and pitch thisidea, they both got eviscerated.
(29:23):
Rent your couch
to strangers, right?
Everyone who owns a carcan be a taxi driver.
Now, who would fund that?
Who would fund that?
Right.
That's the dumbest idea.
Like do you have anyunderstanding the first
time someone doesn't comehome from an Airbnb because
yeah, they got taken.
Or somebody that doesn'tever get out of a Uber cab
is the end of that business.
(29:43):
Like no one willever use it again.
And you know, I,I go back further.
I go back to eBay.
Wait, someone's gonna putpictures on the internet
Uhhuh and say that theyown something, right?
Yep.
And someone's just gonnasend them money and hope
they, they actually had it.
I'm gonna the money and hope
it gets here.
Yeah.
Right.
I'm like, that'll never work.
Yeah.
Somehow it did.
Right?
But
my point is, doyou know how many
kidneys I've boughtfor the Prince of of
Kenya?
(30:03):
So, but here's the point, man.
In order for that towork, you have to have an
individual behind it thathas an optimism, that looks
beyond analysis, that looksbeyond, you know, rationality.
Yep.
And says, yeah,but what if, right?
But what if Right.
Like, and gets is the core of us
agree, right?
(30:24):
And gets other peopleto agree, right?
From Absolutely.
From partners to, to, tofunders, to customers, to staff.
And I get, that's what I wasgetting at before, right?
Like again, AI cancraft the blueprint.
AI could probably even contractout the house, but like.
You gotta live in it, yougotta convince somebody else
to buy it, to live in it.
That's not yet, we'renowhere near that yet.
(30:45):
Uh, so where we have thatlevel of confidence in this
thing.
So let me take this tothe next level, right?
If, if, if all this is true,if, if our uniqueness is our
ability to create, innovate,lead, et cetera, then if
we were to think aboutwhat's the job description?
Of the nextgeneration of founder.
Here's how I've beenthinking about it.
You know, again, as afounder myself, I've been
thinking about it as well.
Number one, my jobdescription is to stop
(31:05):
doing things that somethingelse can do for me, right?
Yep.
Not of laziness, out ofstraight up efficiency.
The second thing, and I thinkthis is the hardest bit for,
for me, Ryan, has been stopthinking in, in what your
previous output terms were.
It goes back what Isaid a moment ago.
In one day I couldcreate five pages.
Yep.
You're no longer the guythat can create five pages.
(31:26):
That's taken me a long time to,to get outta that mentality.
I still ha like, I thinkabout my output still
being a, a version oneof will, so to speak.
That's been hard.
How about
you?
I think for me it wasn't justabout recalibrating the output,
but even reconsidering the KPI.
Right?
So it wasn't like, okay, so notfive pages of content, a hundred
(31:48):
pages of content, neither.
Right?
So it used to be fivepages of content.
Now it's.
How much impact did it have?
Right?
How many likes did it get?
How, what?
What was the intent of the car?
Right?
It depends on what theintent of the content was.
Yeah, yeah, yeah.
But rather like gettingto that final outcome.
Because writing the writing thenewsletter wasn't the point.
It's now let's measure it in howmany pieces of positive feedback
(32:08):
did we get on that thing, right?
Because now we can focuson something entirely
different, and by the way,feed that back into the
machine and have it help usto achieve even more of that.
So it's a shift from doing thework to creating the impact.
Correct.
So I think that's really whereit goes from because we used
to just have to do the workand then measure the impact.
Now I think we can juststay at more like the impact
(32:31):
layer, at least theoretically,because we truly no longer
have to do the work.
At least three parts
and, and also likeagain, how we think about
building our organizations.
It used to be add lots ofheadcount, like that was the
seminal like bragging point.
Right.
Yeah.
You know, we've done wholeepisodes about this where it's
like a CEO would say, well, Ihave this many people, and it
(32:52):
would imply that that's like anarmy that they have the ability
to, to go to war with, right?
And, and more people.
When are we gonna
hear this?
Brag Will, we'regonna hear this brag.
We were gonna dropthe headcount thing.
Nobody's gonna say like, yeah,we just hit a thousand people.
Like, yeah, we just hit,uh, 1500 a day on our,
uh, on our API costs.
Yeah, yeah, yeah.
At some point thatgoes there, right?
Like that becomes the new brag.
(33:12):
It's like we are running so muchAI within that, you're gonna
have a lot of peoplewho have to kind of
recalibrate what that means.
Sure.
Particularly founders, becauseagain, our old job description
was get lots of resources,ergo, capital, humans, et
cetera, to get big things done.
And all of a suddenthose same pieces don't
mean what they used to.
And we're like, wait a minute.
(33:33):
My job isn't to gohoard all that stuff.
My job is to figure out howto be wildly efficient to, to
be able to do 10 x with, uh,10 x more, with 10 x less.
Which by the way, for,for early founders is the
best news in the world.
Exactly.
You don't have any resources.
Right.
But I think this is particularlydifficult for, call it
more seasoned founders likeourselves, is that we're
(33:53):
calling ourselves these days.
I'm trying to think areally kind word, but
I think for, for us.
You know, we have to, tozoom out and say, dude, the
way I was built in, in theworld I came from before,
thankfully does not exist.
And I've gotta evolve.
Yeah.
I have to evolve tothis next version of me.
Or like everything else,someone else evolve for me.
(34:16):
Yeah, man, I'm, I'm hopingwe see more and more like.
Comfort around that evolution.
'cause I'm still seeinga lot of people see that.
Oh, you never do though,
man.
This is, this is thepart, you and I have
been through this before.
I know we went the earliest and,and we were still kids, so to
speak, was the PC revolution.
Remember how many, oh my God.
How many people werelike, I don't want one
of those on my desk.
Right?
(34:36):
Yeah, yeah, yeah.
Even my secretaryand a typewriter.
That went well.
And then, you know, the erathat we grew up, smack dab in
the middle of the internet erain the nineties, every single
person was like, you know, Idon't want that damn internet.
Right?
Like, oh, that guy ain't
around anymore.
Right.
I will never forget it wasn'tthe sales call, it was a
consult meeting, like was,we used to know, walk into
(34:57):
people's offices and satdown and started explaining.
And you, you've talked aboutthis too, like explaining to
people what the internet was.
And this guy totally smug islike, oh, I have that right
here in my desk drawer and pullsout the aol an a OL three and a
half floppy, or maybe not aol.
May have been, mayhave been copy.
It was somebody rightpoint being like.
I had no idea.
(35:17):
And it was one of thosepeople probably not around.
Right.
And all of those are dinosaurs.
Right.
The key for us as foundersis not to be that dinosaur.
Yeah.
In order to not to bethe dinosaur, you gotta
embrace the change.
Yeah.
Even if you don't likeit, it it, no one's saying
you have to like it.
No one's saying youhave to be pro ai.
You can hate ai.
Yeah.
But you have to evolve with it.
It's part of theworld right now.
(35:37):
And yes, it's got a crazynumber of drawbacks.
Right.
Yeah.
That's okay.
That's okay.
Uh, some of thosewill get worked out.
Some of those will getworse if we're being honest.
Yeah.
But our job isn't to figureout why it's wrong per se.
Yeah.
Our job is to be able to say,okay, now I get to do this.
Now I get to do this.
Like now I get to writeRuby Code to make.
Uh, drawer boxes, right?
(35:58):
Yep.
In, in a way that I couldnever do before, right?
Yeah.
Now I get to be able to answerquestions in five seconds.
That would've taken me a week.
Right?
Now I get to doall these things.
What does it mean for me?
What is v two of me knowingthat all, like everything
has changed, right?
And that I can kind of belike, uh, forever extended.
Ryan, do you remember,uh, in the Matrix when,
(36:19):
uh, he downloads Kung fuwhen Kea Reeves, uh, do
downloads kung fu, right?
That classic, oh,I know Kung fu.
I tried for weeksto download Kung fu.
It turns out I, it doesn't work.
Well, so, but my point is like,we now have that ability, right.
So to speak.
Yeah.
Right.
Like we now have thatability, if we don't evolve
and leverage that ability,we will get replaced.
(36:42):
Yeah, of course.
Not by ai, but by theother founders who, who are
actually using ai using it.
Right, right.
And, and using it andembracing it in the way
that it's supposed to be.
And so, so here's whatI would say overall.
I think, you know, whenwe look at the question,
can AI replace founders?
Yes.
If we let it, yes.
(37:03):
If we let it, if we become nomore than just those automatons
that we're doing, you know,rote work, then, then yes,
it can and, and frankly,right, it probably should.
But if we zoom out and say,okay, uh, you know, some
of that's scary, et cetera.
But if I use this, if Iexist as the architect,
the curator, the creativemind, right, the mastermind
(37:23):
behind all these tools, nowI can create 10 x what I
would've ever created before.
Now I can become a founderand, and have the, the outcome
and the output that I'venever even dreamed of before.
At which point weevolve to that person.
Literally anything is possible.
Overthinking your startupbecause you're going it alone.
(37:44):
You don't have to, and honestly,you shouldn't because instead,
you can learn directly frompeers who've been in your shoes.
Connect with bootstrapfounders and the advisors
helping them win in thestartups.com community.
Check out the startups.comcommunity@www.startups.com
to see if it's for you.
Could be just thething you need.
I hope to see you inside.
Welcome back to the episode ofthe Startup Therapy Podcast.
This is Ryan Rutanjoint, as always by my
friend, the founder,and CEO of startups.com.
Will Schroeder, will AIis now writing our copy,
debugging our code, AB testingour pricing while we sleep,
to the extent that we do.
So the existential questionsgetting pretty loud and clear.
(00:21):
If the smartest person inthe room now costs 49 bucks a
month, what's left for founders?
So I think, man, it'd be funtoday to break down like what
parts of us get automatedand which parts become
more valuable than ever.
The
question a lot of peopleare saying is, can
AI replace a founder?
Yeah.
And, and I, I'm gonna beheretical when I say this.
Yes.
(00:41):
Now I say that as a founder,as somebody who does not
want to be replaced, right?
Yeah.
So I wanna be clearly, I, Ihave a strong bias to say,
no, this is where we fightback against Skynet will.
It kind of is, weknow how this goes.
It doesn't end well for us.
Although we do get aa, a cool Austrian sent
back to it to help us.
You sure do.
So here, here'swhat I would say.
When I say yes, I don't mean,uh, universally all of us,
(01:03):
I'm saying Yeah, yeah, yeah.
A lot of, I think whathistorically I considered
my founder traits.
Um, my ability to workreally hard on all these
different things, those thingsjust aren't hard anymore.
Right.
Like, I'll give you an example.
I've prided myself fora very long time on
being a strong writer.
You've seen me write unlimitedamounts of copy, right?
Yep.
And it was like we could justtake any topic, uh, whether
(01:25):
it's landing page copy, oreducational copy, or your
newsletters, and you'relike, we'll, write something.
And I would just, right?
Yep.
And, and like I said in,and my writing was decent
and it was passable.
Yeah.
But now.
AI can write as well, ifnot better than I can.
And it's almost like that skill.
Okay.
This is just one examplekind of went away.
(01:47):
Now the, the importanceof having something to say
has not gone away here.
Right, right.
What you're trying, tryingto say part very much
I.
But your ability to conveythat in words is no longer a
unique skill that some peoplehave and some people don't.
Some people, of course,are gonna be better at it,
but no, if, if chat GBTcan crank out 50 landing
page variants before lunch,then we can't treat the
(02:09):
copywriting piece as thefounder superpower, right?
Knowing what to say, important.
Saying it less.
So I wanna dig into thatjust a little bit more.
'cause I, this is just sofascinating to me that this
whole, this whole era rightnow is so fascinating to me.
Right now we're at a pointwhere AI, and, and I, I would
(02:29):
argue, I always use the,the corollary to, uh, gaming
machines from Atari 2,600 Yeah.
To like, you know, uh,current Xbox, so to speak.
And I, and I would say AI isstill in roughly somewhere
between the Atari 2,600 andthe in television stage, right.
Just like the Right,the very early stages.
And at the time peoplenever see it that way.
Like, like when you had yourAtari 2,600, you thought it
(02:50):
was the most amazing gamingsystem that it e ever existed.
Well, it was becauseit was right.
'cause right before that Ihad pong, and right before
that I had just tic-tactoe on a piece of paper.
You bet, you bet.
So I say this to saythere are many, many, um,
iterations that are coming,but even in the current
iteration, at the most basic.
What AI can do now, it makesa lot of the stuff we were
(03:12):
spending so much time doing.
Kind of irrelevant or at worse.
Yeah, man.
Duplicative.
Yeah.
Yeah.
Anything that's repeatable,data-driven, pattern based.
To me, man, that'salready AI turf.
Right.
And as you said, we're justat the beginning stages.
And
here's anotherway to look at it.
I used to run a largead agency, right?
Yeah.
And we had hundreds and hundredsof people, uh, copywriters
(03:33):
and creative directors.
Yeah.
You know, and allthese folks, right.
And the copywriters, forexample, of which I had
a, a strong affinity for,would be able to sit in a
room and we'd be working ona campaign, and they'd come
up with dozens and dozens ofdifferent ideas and approaches.
Right now, within thatcontingent of copywriters.
Like anything else,like athletes or
anything else like that?
There were a few thatwere exceptional.
(03:54):
Let's say the A players.
Yep.
And there were a handful thatwere B players and a lot of
people that were C players.
Okay.
Yeah.
Now what I would argue ismost AI can perform about
as well as the B players andmost certainly the C players.
Yes.
Yeah, I agree.
Are they replicatingthe A players?
No.
And they, they may never.
(04:14):
May never.
Okay.
However, that's here nor there.
The point is, if I can nowhave copywriting and and idea
generation done at a B level.
Yeah, for $0 for$49 a month, right?
Yep.
Why wouldn't I
be doing that?
Exactly.
No.
The accelerative power,that is insanity.
Right?
And this is, this is whatwe should all want, right?
(04:34):
This is what we all want.
Correct.
We've always wanted this.
We're like, how could I justget them to produce way more?
Right?
I can only afford one person.
How can I turn them into 10?
Here's your answer.
And again,
we're focusing this episodeinward to say, and what if I
wanna be the force multiplier?
Here's a, a great way for meto, to, to show the difference.
Uh, a few years ago when I waswriting copy for something,
(04:55):
I kind of knew what my rateof output would be, and I
knew that, let's say I'mjust making this up on, in
any one day, I could create,uh, you know, a five page
document, just, you know,using it as a unit of measure.
Okay, well now in a day I cancreate 10, five page documents.
Yes.
Right.
So because the AI can basicallytake the ideas from my head
(05:15):
and just put all the wordsthat was like me manually lay
laying bricks before, right?
Yeah, exactly.
And I just don'thave to do that.
No.
So part of me would say, man,that felt very special to me.
You know, I felt very uniquethat I had that capability.
Uh, and, and, and that's beenquote taken away from me,
but there's another side tosay, but was that really was
(05:38):
my ability to, to, to writewords and compose the value?
Or was it the ideabehind those words?
Yeah, I mean thatwas the thought.
The always was the idea.
We just didn't haveanother way of manifesting
the idea in the past.
Right.
There was have idea.
Now put it to paper.
Right.
We just don't have to do allof the manual work for that.
It's funny, man.
I had this realization theother day that there are tasks
(05:59):
that I think once were likethe signals of my genius, that
now would just be a signal thatI lost my open AI password.
Like nothing else.
Like if you see me doingthem, you're like, wait,
where I did the power off?
Like what are you doing?
And again, like to your point,like things that used to seem
really, really important to me,like digging through analytics
and pattern spotting, right?
Right.
Seeing things otherpeople couldn't see.
(06:21):
I'm good at that.
Yeah.
I'm nowhere near as good atthat as a machine that can
crunch through millions ofrecords of data in minutes.
Um, just not, so
I think this is thefirst building block
of this conversation.
I think let's first explorewhat should be replaceable.
Right.
So I know everyone has this,this I immediate reaction
and it's visceral to say,I can't be replaced by
(06:43):
machine, blah, blah, blah.
I'm like, dude, you can, right?
Yeah.
And you will be.
So let's talk aboutwhy you should be and
what you do about it.
Right.
Sure.
Not just like this, thisnatural tendency to just push
back and just be like angryabout it that buys you nothing.
Yep.
And there is a very, very,very long list of tombstones
of people that had exactlythat reaction and didn't
(07:05):
work out well, particularlyin the business world.
So again, Ryan, when youthink about things that are
replaceable, that shouldbe replaceable, that like,
while, like you just mentionedanalytics, that a moment
ago you were like, Hey,this is what I got paid to
do, and now it's like it'sa $49 a month, uh, problem.
Yep.
What else comes to mind?
It's the, the
time consuming part of it.
You can just look at itfrom a, a pure time calculus
(07:26):
perspective, like, and replacethe things that are gonna
take you the most time andstill, like, they're valuable.
They, they need to bedone, but they don't
need to be done by you.
You know, as we were cominginto this episode today, will,
I was, I was thinking aboutthe process you've gone through
with building the house.
Mm-hmm.
Right.
And it occurred to me like, youknow, you at the beginning, like
you could have treated yourselfas the architect, the GC.
(07:49):
The plumber, the electrician,the framing guys.
Right.
All of it.
Right.
You could've just done allof it and you've done, you've
done a lot of stuff, but atsome point, like it makes
a lot more sense for you tohire a GC and a plumber and
electrician and all of that.
Right?
Right.
Well, I think that, you know,we were once as founders,
we were once the architect,GC plumber, all of it.
Like we just hadto do everything.
We got so used to, especiallyat the time you and I
(08:11):
came up through this.
Yeah.
We had to do everything.
There wasn't anybodyelse to do it.
You really did.
Yep.
There weren't tools to do stuff.
We did a littleof everything now.
AI steps in as the GC andAI just hired the plumber.
Yep.
Now you can just stay inthe architect's chair.
Right.
That's the way I see it.
So what, what I'm looking atis like, how can I just stay
(08:31):
a above that kind of messywork layer in a way that
doesn't keep me uninformed,but doesn't have me overworked
in the sense that like I couldbe doing more valuable stuff.
So instead of havingto crunch all the
analytics like I used to.
I can get the insight fromthat and I can spend more
time deciding what to dobased on that insight.
Right, right, right.
So the, the, the base levelanalysis, all of those kind
(08:52):
of things just going away,anything that requires like
mass creation or mass ingestionand kind of normalizing of
information just gonna go away.
Right.
So at the marketing level,looking at across, like, you
know, if we're honest withourselves, we talk about AB
testing and, and you know,optimization and all this
stuff that we, that we do.
(09:13):
Yet, it's sort of done on kindof like a rolling window basis.
Mm-hmm.
Do I look all the way backto the very first experiments
I ran on Facebook ever.
When I'm, yeah.
When I'm running, runningthe ones now, yeah.
I can't, I can'tlook at 6,000 ads.
AI can, so now it allows meto do these things better.
And it was never aboutthe analysis before.
(09:34):
It was about gettingto the answer and then
acting on the answer.
Right.
And I still think thatthe acting on the answer
piece is still very much.
In founder camp.
We'll see for how long?
I
think a big part of where thepushback is, is there's an
overwhelming focus, justifiablyso right now around what
AI is taking away from us.
There's not enough of theconversation, and I think the
(09:56):
smart founders are alreadypointing this direction around
what it is enabling us to do.
Not that nobody's evermentioned what, what AI can do.
Yeah.
What I'm saying is asfounders, as founders,
there has to be a versionof us that says, dude, like.
I'm about to be Superman.
Right.
Yeah.
Like for example, when Iwas first starting in my
career, I didn't understandanything about finance.
(10:17):
I mean, not even a little bit.
And as you know, I've been astartup CFO for 25 years in
addition to my, my day job.
And it's not because I'ma financial wizard, it's
because as soon as you putdollar signs in front of
those numbers and you said,Hey, that's your money.
Yeah.
All of a sudden I was like the,the rainman of finance, right?
Like I was just like,I, I understood it.
All right?
The matrix, it all made sense.
(10:38):
But think of how under.
Armed.
I was at that time, right?
He here, here's how Ifigured out finance.
I drove to Barnes and Noble.
I found a book that was probablypublished 15 years prior
on small business finance.
And I read it cover to cover,which took days, right?
Yes.
And it was the driest,most boring book ever.
(11:00):
However, changed my life.
But think of all the thingsthat that had to happen in
order for me to be able todo this one simple thing,
which was put together anincome statement, right?
Yeah.
Yeah.
Now I can do that like aseasoned pro without even
knowing how to do it.
I just, I have to go back tothis anecdote where, where
you talked about, where youwere asked for the income
(11:20):
statement and you, you pulledout the, uh, you pulled the
piece of paper outta yourpocket that just had like the
monthly, like the things youhad spent money on that month
and like showed the investor.
I love that man.
This, that was, that was,this is the moment where
that education becameobvious and necessary
to you.
That would've been reallyuseful at that time.
Hey, what's your proforma income statement?
My, my what?
My what?
(11:40):
My, um, here's a listof things I'm gonna
buy at the store later.
That the team, yes.
I literally gave a, a worddocument of, of how much I
owed my roommate that month.
Yeah.
But again, you know,plan it out a little bit.
So we keep talkingabout the things that's
taking away from us.
But when I say, you know, someof these traits, you know,
are replaceable, I'm sayingnot only are they replaceable,
(12:02):
we should lean into that.
Yeah, right.
A hundred percent.
My first reaction being AIis taking away my uniqueness
of being a copywriter.
Okay, got it.
Okay.
But how can AI also makeme 10 x the copywriter?
Right.
Yeah.
If, if, if I have so muchto say, how can I say it
exponentially faster, moreefficiently, et cetera.
(12:23):
Like, I've got an AI that'strained on my voice so I can
tell it anything I want andit will respond in my voice.
Now I don't love it.
Right.
I don't love it.
So, so like, uh, when I useda brainstorm stuff, it'll come
back in my voice, but it's, asyou know, AI has a smell to it.
Yeah.
Yeah.
So,
so it's always likethe, the kind of.
AI version of my voice.
(12:44):
Yeah,
it's 80%, but it ain'ta hundred percent.
You mean auditory voice?
You're not saying writesomething in your voice
and tone, you mean?
Oh yeah.
Your actual voice point.
Good point.
Actually, the auditoryvoice is spot on.
No, in this case, I mean the,uh, the, the, the writing
and, and I gotta say, man,uh, what I get back if I
published it as is, no onewould know the difference.
I'm the only person that cantell the difference in this
case, but I can tell anyway.
(13:05):
Maybe you just don't likeyour writing will maybe that
I, I I always used to make thejoke that if we ever come up
with an AI that can replicateyourself, that will be the
most hated person of all time.
Oh, yeah.
Or, or AI hang out withyourself all day long.
Oh, how about this?
God.
An AI that is purely basedon your search history.
Yes.
That person, the worsthuman of all time.
(13:27):
Oh God, I hate that person.
Anyway, I think that again,we're having that first initial
reaction that says, you know,AI pad, blah, blah, blah.
Um, we might wannalook at it going.
Should you have had to knowall of finance in order
to advance your business.
Yep.
And there's gonna be someargument where someone says Yes,
but that knowledge and what youwent through Uhhuh, you know,
(13:47):
taught you to do X, Y, Z, and,and there's some truth to that.
Sure.
But dude, that is a reallypainful, hard way to me,
like 10 years to do it.
Right.
I needed that informationday one, not 10
years from now.
This is here, here's,this is a big question.
I don't think anybody cananswer this just yet, but
one of the questions is.
How will this be different forpeople who didn't come up, who
(14:08):
didn't have those hard lessons?
So if you're just born now Yeah.
And you, you justgrow up an AI native.
Yeah.
You just have alwayshad this stuff at hand.
You're always using it.
How does that change yourability to use it to see
the difference between the,the, the good, the bad.
To know that difference?
Like you talked before aboutit A level, B level, C level.
Yep.
If you've never worked witha bunch of copywriters and
seen a, B and C level copy.
(14:30):
How will you know when AI'sactually doing the right thing?
Right.
You won't.
I, I, I actually justwent through this, right.
You know, as I'm designingmy house, I use a tool called
SketchUp, you know, which Iput in the 3D model Within
that I have to, I built all thecabinetry in the house and all
the, the closets and everything.
Yeah.
And I had to builddrawers, you know, for,
for every one of those.
Well, in order to go intoSketchUp and create what, what
(14:50):
amounted to 88 drawers to sizethem, create the 3D geometry
and whatever, and get 'em allset up, would've taken forever.
Okay, so this is a perfectexample of me coming from
a space that I'm not in andbeing able to do something.
There's no way I would'veever been able to do.
I go to GPT and I said, here'sa list of all the drawers that
I need, just like the, thewidths and the, uh, depths.
(15:11):
I said, make meruby code, right?
That I can import into SketchUp.
That will automaticallycreate all of the drawers
that would, that wouldfit for these dimensions.
Give them a, a notch and adata and you know, all this
stuff, and create all thegeometry, the full model.
Right.
Five seconds now.
Okay.
Nuts, dude.
Unbelievable, right?
(15:32):
Yeah.
Would've taken me a week.
Okay.
Yeah, and I've got, Iwould've gotten it wrong.
Now, a couple things.
Number one, 99% chance, there'ssomething in there that's wrong.
There's a drawer witheight sides to it or
something like that.
Yeah.
GT's.
Just not very accurate.
So it helps to knowwhat you're looking for.
But here's my point.
This is the samewith with finance.
(15:52):
What it would've taken me.
To perform the same operation.
Yeah.
To learn Ruby code, right?
Yeah.
To recheck all of my math,to be able to sit and see, I
mean, like it would've takenme, uh, forget the Ruby Code.
It would take me aweek to do it manually.
Right?
Yeah.
And longer to learn Ruby Code.
And someone might say, yeah,but then you'd know Ruby Code
and here's what I would say.
(16:12):
Cool.
I may never need that again.
Right.
I just needed todo that one thing.
All my action script andcold fusion knowledge
is getting used.
Not at all at this point, right.
I'm just say love that stuff.
Like
I don't necessarily, like, Iget learning things and again,
I, I love learning things.
So they we're in a platformwhere we teach people
stuff, but not everythinga founder needs to do.
(16:35):
They need to be theeducated expert in all the
eccentricities of that.
It doesn't matter ifyou know a, B or C grade
copy, you just need copy.
Yeah.
And if it doesn't soundlike total shit, it's gonna
work.
It's gonna be agood starting point.
At least you canalways adjust it.
Yeah.
'cause the reality is we'renot gonna know whether it's
good copy or not until itfaces some other humans.
Mm-hmm.
Or maybe some AI andgets picked up and search
(16:55):
results and whatnot.
Right, right.
Yeah.
So, you know, from mystandpoint, I look at all of
these things that I can now do.
That I couldn't dofive minutes ago.
Okay.
Yeah.
So for example, the big thingnow is people can write code.
Yeah.
Right?
Which is, so, I mean, Ryan, youand I have haven't written a
line of code in 20 years, right?
Our knowledge is so far fromuseful at this point, but now I
(17:16):
can write code in any languagefor any purpose I can get
MVPs put together, et cetera.
How is that not
a force multiplier?
The time it takes me tocreate a landing page now.
Right now, not aread to chip, right?
It's gotta go over to Dev andthey gotta do their stuff track.
But the time it takes me tocreate a landing page now is
less than the time it would'vetaken me to go through the
(17:37):
drawer to find the pencilto do the sketch on paper.
Right?
It's bananas.
How much faster we can moveand how we can get onto like
the important parts, right?
Because the important partwasn't drafting the wire frame.
The important part isgetting that page launched.
Getting traffic point ofthat page and starting to
then have AI analyze theresults and tell me what
(17:58):
still needs to be changed.
Fantastic.
Right?
This is what you want.
It's short circuits, somany tedious processes.
The number of hard choicesI have to make now.
Are so significantly lowerbecause it's like, well, I
can do this, or I could dothat, or I could do that,
or I could do that one.
What are we gonnaget to this week?
All we'll just do,just yes is the answer.
(18:20):
Now, I love being able to justsay, yes, come on, why not?
Right?
Like, oh my God, in, in mygetting customers workshops
on Mondays and Fridays.
The number of requests I can nowfield people that need help with
really heavy lifting stuff werehonestly, most of the lifting
was on their part, but I stillhad to show 'em how to do it.
Yep.
This is part of where I'mjust like automating and,
and AIing like crazy.
(18:40):
Like what am I asking foundersto do over and over and
over again That not onlyare they resistant to, but I
know they desperately need.
Cool.
Let's AI the shit outta thatbecause having something done is
better than having nothing done.
Right.
And in a lot of cases it'sjust motivating them to do
more of sometimes the manualwork, but it shows them the
power of what that thing is.
If, if anything, it just serveslike a little simulation.
(19:01):
IP, or like, you know, I'm ahuge fan of the Mom test, right?
Yeah.
Love that book.
Please go buy the book.
Rob Fitzpatrick dida great job with it.
But the process of goingthrough that, for a lot of
founders, they're like, I'mjust not gonna go do that,
or I'm not gonna go do it.
Right?
It's me going to BarnesNoble and buying that book.
Right?
It them, yeah, yeah, yeah.
I simulate it forthem in 15 minutes.
Then they're like, oh,that's what this tells me.
(19:23):
Yeah, that's what thistells you now, right?
You can trust this to adegree 'cause it's based on
the average of the internet.
It'll get most of it right?
But you should also pick upthe phone and talk to a couple
of your, your, your peopleand get the actual answers.
And it's been a greatmotivational factor,
which is awesome.
So if.
Like, and, and here's thething, every time I do
that, there isn't a foundergoing, but Ryan, aren't
you replacing a part of me?
(19:44):
Yeah, a part of me thatyou didn't want to exist
in the first place.
Right?
Alright.
Right.
Nobody's upset about it.
Nobody.
They're all like, great.
I can get onto thepart where we turn that
into useful informationto sell people stuff.
You know, something that's
really funny about everythingwe talk about here is
that none of it is new.
Everything you're dealingwith right now has been done a
thousand times before you, whichmeans the answer already exists.
(20:06):
You may just not know it.
But that's okay.
That's kind of whatwe're here to do.
We talk about this stuff onthe show, but we actually
solve these problems alldayLong@groups.startups.com.
So if any of this soundsfamiliar, stop guessing
about what to do, let us justgive you the answers to the
test and be done with it.
As a carpenter myself, uh,whenever I'm, you know, building
(20:28):
a house, I think to myself,there's no carpenter ever
that's thinking to themselves.
I'm so pissed off that you gaveme a, a powered circular saw.
My hand saw was so much better.
Yeah.
Dick, come on.
Yeah.
Just rubbing my head withoutblisters on my hands just
doesn't feel the same.
Yeah.
You know, like,need those calluses.
All
right.
So we've definitely said yes.
(20:48):
There's tons of stuff thatcan be replaced, right?
Yeah.
And, and we're almostsaying, and should be.
Right.
Like, like do not resist.
In this case, a lot of thestuff, not everything, and
we're about to go the otherdirection, a lot of this
stuff should be replaced.
We should embrace thereplacing like we're shedding
skin because the worldjust got way easier for us.
Yeah.
But let's talk aboutthe other side.
(21:09):
You know, we said the,the topic of this is
can AI replace founders?
But let's talk aboutthe other side.
What makes us unique?
What is AI notlikely to replace?
And I wanna caveat this bysaying there is a sci-fi
version of everything.
Where you could say and makean argument, and I'm one
of the people that can makethis argument that would
say, no matter how hard youtry to say that humans are
(21:29):
always going to be unique.
We are an algorithm, andalgorithms can be replicated.
Right?
So I, I wanna be clear when Isay this, we're not, Ryan and
I aren't sitting here beinglike humanity is, you know,
irreplaceable in any way.
That like, we, we recognize thatat some extreme level there's
probably an AI that, that can dosome things, but at a practical
level, at a level that actuallymakes sense for, for either now
(21:50):
or in, in the very near future.
I don't believe that AI cando some really human things.
Sure.
And I don't believe it should.
I think, you know, if, ifwe're saying AI can create
music, but should it, right?
Yeah.
Yes.
It can create something Right.
But the point of musicwasn't to make the music.
It was a form of expression,
(22:11):
which usually comes from aplace of need to express Right.
AI will never needto express itself.
Correct.
It doesn't takeanything away from that.
Correct.
Right.
And I think now there'ssomething really fascinating
about how we get touse our uniqueness.
Let's say again, I'm gonnago back to me as a writer,
imagine for a second.
There are two componentsof me being a writer.
The part where I can generatea unique thought that people
(22:33):
might give a shit about,and the part where I can
translate that to words thatpeople might wanna read.
Yeah.
Okay.
The latter has beenmostly replaced, right?
Yeah.
Just about any thought Ihave can be translated to
words that are 90% of whatI would've used anyway.
Okay.
As weird as thatis, it's It's true.
But the unique thought.
This, this unique idea,this, this, this unique
(22:54):
opinion, et cetera.
AI can come up with variants.
They can't come up withmy variant necessarily.
Right?
So what happens is my voice now,I mean my voice vis-a-vis the
work that I do, the things thatI produce in life, et cetera,
becomes force multiplied.
Yeah.
If before I could, my opinioncould generate one article, now
(23:15):
my opinions can generate 50.
Right now they can generatea goddamn movie in the
not to distant future.
That is amazing.
But in order for all that tohappen, the uniqueness has to
be me having something to say.
Yeah.
It starts with that, right?
Like the, the things thatit's, you know, the, the key to
differentiators at this pointare, you know, like vision,
narrative, human urgency,moral compass, like these
(23:36):
things can't be replaced yet.
Humor still notparticularly good at that.
Right.
Um, but I wonder to some degree,like you said, that like, it
can't come with my thought, but.
Part of that, how, howmuch of that is, is the way
that we use it right now?
Mm-hmm.
Right
Now, I have testedthis a little bit.
I played around with like,instead of saying, here's
(23:58):
what I think, right?
The rest of this, it'ssort of like, here's
the thing that happened.
What do you think myreaction to this is?
Right.
It's interesting.
It's really interestingand, and basically like,
what do I think about this?
What do you think Ithink about this based on
what you know about me?
What do you think?
I think because we've bothgiven AI plenty of information
about ourselves, and certainlyit learns all the time as
(24:18):
we're, as we're utilizing it,but to your point, it comes up
with some, some possibilities.
I, I would say it's still.
Generally not close.
But then I wonder like howmuch of that is my own bias
saying that's not exactly whatI would've thought about that.
Or, well, when I thoughtabout it already, that's
not what I thought about it.
But how about this?
What if we were takingthis a step further and
we're to say, what mattersnow is what you want.
(24:41):
Not what the AI mightrespond to mm-hmm.
But what you specifically want.
In other words, I can gointo an AI right now and I
can say, give me ideas forwhat startups might want.
Okay.
And it'll, it'll comeup with tons of ideas.
And, and I would say it'llcome up with probably
75% of the ideas that Iwould've come up with.
Right.
Just enough darts at a wall.
Right.
Like, uh, you know, whatever.
(25:01):
But the, and, and, and oneof them might've matched
what I was gonna say.
That's perfectly fine.
Right.
But the point is,I'm the curator.
I'm the one thatsays, no, not that.
Yes, that.
No, not that.
Yes, that.
My vision, what Iwant is what matters.
That goes back to music.
Yes, AI can create someversion of music, but the
composer is what matters.
(25:21):
The composer decides right.
Is that what they want?
Right.
It's not like, uh,I'm a composer.
The AI creates a song andI'm like, I hate it, but
I have to have it anyway.
Right.
The whole point is, yeah, weare, we are the final decision.
We, you know, we are, to yourpoint earlier, the architect
of all of this we're the onethat our curatorial ability
to be able to say, yes,this, yes, this, yes, this,
(25:44):
whether it's people, product,marketing, copy, et cetera,
is what makes us unique.
Yeah, I mean at this point, likego back to some of our other
analogies, like AI can draftall the blueprints, but like
it can't convince the city tobuild the skyline, can't it?
Right.
Like
Right.
Or just have the vision to whyyou'd wanna do it to begin with.
(26:05):
Right.
What I'm saying iswhat makes us unique is
this curatorial vision.
What doesn't make usunique is laying all the
bricks to do it anymore.
No, I know.
So again, going back to to,to my, my place in the world
as a writer, which I'vegiven a lot of thought to.
What makes me uniqueis the fact that I have
a lot of unique ideas.
Right, or just, or unique to me,whether anybody else has thought
(26:26):
'em before, who cares, right?
Unique to me.
But now I no longer havethe time element, the
writer's block element,et cetera, that prevents
me from expressing them.
Right.
So whereas before, you know, Iread a, a weekly newsletter that
you, that we published, that we,that we do this podcast based
on before, I'm like, okay, ittakes me x amount of time to
take whatever unique idea I had.
(26:48):
Like, can I replace foundersand convert that into, uh, copy.
Now it still takes meroughly the same amount of
time, uh, because I liketo play with the idea.
I like to wrestle with the idea.
It's not like I'm just tryingto write a, a newsletter.
Like I try to, like,I, it's very cathartic
for me, but ultimately.
It's my idea, right?
Uh, you know, that I'm breakingAI can gimme 50 variants of,
(27:10):
of how AI would talk aboutit, but I'm still the curator.
I'm still the one to say,yes, this, no that, yes,
this, no, that it's my job.
And this is, this is a metaphorfor building a startup.
It's still my job toorchestrate all of this.
Well, it always was, right?
It's, it's funny to me that wedon't think about it in these
same ways because in a lot ofcases, at the early stages, as
the founder, we, we start bydoing everything, and then the
(27:31):
second we can afford it, westart to hire out specialists.
We start to hire out otherpeople, sometimes not even
specialists, just peoplecan do the work, right?
If it's just rote work, youhire people capable of rote work
and they do the work, right?
You don't hire a bricklayer.
If you're building a 6,000square foot high house,
you hire a bunch of them.
Right?
Right.
So we've always had thisconcept that like, you know,
we'll, we'll hire it out.
We'll hire out the work.
(27:52):
We'll, we'll divide the labor.
We just have a differentability to do that.
Right.
GPT just became the assemblyline to our Henry Ford.
Right.
We got the same 24hours of work in.
Yep.
But we've got a hundredx the output now.
You bet.
You bet.
Now I also think anotheraspect which, which I, I don't
wanna overlook or not getinto, is just the fundamental
concept of leadership.
Be it a mixture of pureanxiety, like pure,
(28:16):
unadulterated, anxiety,excitement in fervor, right?
Yeah.
Vision, which I think is aterm that's overused, but it's,
it's certainly a part of this.
Right.
And just fucking determination.
Determination.
Just be able tosay determination,
deciding that we are, we'regoing to keep doing this
until we get it right.
Correct.
AI's not
gonna care.
Yeah.
Right.
And, and nor should it.
(28:36):
Right.
That's the point though.
Like if you look behind anysuccessful company and frankly
many unsuccessful ones, whatyou have behind this company is
a freakishly determined person.
Yeah.
To see this through,despite all logic.
Right.
That's what makes foundersthe most unique, is that
we are able to run in, runnaked, into the abyss despite.
(28:59):
Everything that wouldtell us otherwise.
Yes.
And figure out how to somehowcome out on the other side.
I'll give you an example.
Uh, you know, one of myfavorite would be both
Airbnb and Uber, right?
Two famously successfulcompanies with the
dumbest ideas ever.
Okay?
When I say dumb, I, I, I,obviously, I, I'm kidding when
I say this, but when those poorbastards had to go out there
into the market and pitch thisidea, they both got eviscerated.
(29:23):
Rent your couch
to strangers, right?
Everyone who owns a carcan be a taxi driver.
Now, who would fund that?
Who would fund that?
Right.
That's the dumbest idea.
Like do you have anyunderstanding the first
time someone doesn't comehome from an Airbnb because
yeah, they got taken.
Or somebody that doesn'tever get out of a Uber cab
is the end of that business.
(29:43):
Like no one willever use it again.
And you know, I,I go back further.
I go back to eBay.
Wait, someone's gonna putpictures on the internet
Uhhuh and say that theyown something, right?
Yep.
And someone's just gonnasend them money and hope
they, they actually had it.
I'm gonna the money and hope
it gets here.
Yeah.
Right.
I'm like, that'll never work.
Yeah.
Somehow it did.
Right?
But
my point is, doyou know how many
kidneys I've boughtfor the Prince of of
Kenya?
(30:03):
So, but here's the point, man.
In order for that towork, you have to have an
individual behind it thathas an optimism, that looks
beyond analysis, that looksbeyond, you know, rationality.
Yep.
And says, yeah,but what if, right?
But what if Right.
Like, and gets is the core of us
agree, right?
(30:24):
And gets other peopleto agree, right?
From Absolutely.
From partners to, to, tofunders, to customers, to staff.
And I get, that's what I wasgetting at before, right?
Like again, AI cancraft the blueprint.
AI could probably even contractout the house, but like.
You gotta live in it, yougotta convince somebody else
to buy it, to live in it.
That's not yet, we'renowhere near that yet.
(30:45):
Uh, so where we have thatlevel of confidence in this
thing.
So let me take this tothe next level, right?
If, if, if all this is true,if, if our uniqueness is our
ability to create, innovate,lead, et cetera, then if
we were to think aboutwhat's the job description?
Of the nextgeneration of founder.
Here's how I've beenthinking about it.
You know, again, as afounder myself, I've been
thinking about it as well.
Number one, my jobdescription is to stop
(31:05):
doing things that somethingelse can do for me, right?
Yep.
Not of laziness, out ofstraight up efficiency.
The second thing, and I thinkthis is the hardest bit for,
for me, Ryan, has been stopthinking in, in what your
previous output terms were.
It goes back what Isaid a moment ago.
In one day I couldcreate five pages.
Yep.
You're no longer the guythat can create five pages.
(31:26):
That's taken me a long time to,to get outta that mentality.
I still ha like, I thinkabout my output still
being a, a version oneof will, so to speak.
That's been hard.
How about
you?
I think for me it wasn't justabout recalibrating the output,
but even reconsidering the KPI.
Right?
So it wasn't like, okay, so notfive pages of content, a hundred
(31:48):
pages of content, neither.
Right?
So it used to be fivepages of content.
Now it's.
How much impact did it have?
Right?
How many likes did it get?
How, what?
What was the intent of the car?
Right?
It depends on what theintent of the content was.
Yeah, yeah, yeah.
But rather like gettingto that final outcome.
Because writing the writing thenewsletter wasn't the point.
It's now let's measure it in howmany pieces of positive feedback
(32:08):
did we get on that thing, right?
Because now we can focuson something entirely
different, and by the way,feed that back into the
machine and have it help usto achieve even more of that.
So it's a shift from doing thework to creating the impact.
Correct.
So I think that's really whereit goes from because we used
to just have to do the workand then measure the impact.
Now I think we can juststay at more like the impact
(32:31):
layer, at least theoretically,because we truly no longer
have to do the work.
At least three parts
and, and also likeagain, how we think about
building our organizations.
It used to be add lots ofheadcount, like that was the
seminal like bragging point.
Right.
Yeah.
You know, we've done wholeepisodes about this where it's
like a CEO would say, well, Ihave this many people, and it
(32:52):
would imply that that's like anarmy that they have the ability
to, to go to war with, right?
And, and more people.
When are we gonna
hear this?
Brag Will, we'regonna hear this brag.
We were gonna dropthe headcount thing.
Nobody's gonna say like, yeah,we just hit a thousand people.
Like, yeah, we just hit,uh, 1500 a day on our,
uh, on our API costs.
Yeah, yeah, yeah.
At some point thatgoes there, right?
Like that becomes the new brag.
(33:12):
It's like we are running so muchAI within that, you're gonna
have a lot of peoplewho have to kind of
recalibrate what that means.
Sure.
Particularly founders, becauseagain, our old job description
was get lots of resources,ergo, capital, humans, et
cetera, to get big things done.
And all of a suddenthose same pieces don't
mean what they used to.
And we're like, wait a minute.
(33:33):
My job isn't to gohoard all that stuff.
My job is to figure out howto be wildly efficient to, to
be able to do 10 x with, uh,10 x more, with 10 x less.
Which by the way, for,for early founders is the
best news in the world.
Exactly.
You don't have any resources.
Right.
But I think this is particularlydifficult for, call it
more seasoned founders likeourselves, is that we're
(33:53):
calling ourselves these days.
I'm trying to think areally kind word, but
I think for, for us.
You know, we have to, tozoom out and say, dude, the
way I was built in, in theworld I came from before,
thankfully does not exist.
And I've gotta evolve.
Yeah.
I have to evolve tothis next version of me.
Or like everything else,someone else evolve for me.
(34:16):
Yeah, man, I'm, I'm hopingwe see more and more like.
Comfort around that evolution.
'cause I'm still seeinga lot of people see that.
Oh, you never do though,
man.
This is, this is thepart, you and I have
been through this before.
I know we went the earliest and,and we were still kids, so to
speak, was the PC revolution.
Remember how many, oh my God.
How many people werelike, I don't want one
of those on my desk.
Right?
(34:36):
Yeah, yeah, yeah.
Even my secretaryand a typewriter.
That went well.
And then, you know, the erathat we grew up, smack dab in
the middle of the internet erain the nineties, every single
person was like, you know, Idon't want that damn internet.
Right?
Like, oh, that guy ain't
around anymore.
Right.
I will never forget it wasn'tthe sales call, it was a
consult meeting, like was,we used to know, walk into
(34:57):
people's offices and satdown and started explaining.
And you, you've talked aboutthis too, like explaining to
people what the internet was.
And this guy totally smug islike, oh, I have that right
here in my desk drawer and pullsout the aol an a OL three and a
half floppy, or maybe not aol.
May have been, mayhave been copy.
It was somebody rightpoint being like.
I had no idea.
(35:17):
And it was one of thosepeople probably not around.
Right.
And all of those are dinosaurs.
Right.
The key for us as foundersis not to be that dinosaur.
Yeah.
In order to not to bethe dinosaur, you gotta
embrace the change.
Yeah.
Even if you don't likeit, it it, no one's saying
you have to like it.
No one's saying youhave to be pro ai.
You can hate ai.
Yeah.
But you have to evolve with it.
It's part of theworld right now.
(35:37):
And yes, it's got a crazynumber of drawbacks.
Right.
Yeah.
That's okay.
That's okay.
Uh, some of thosewill get worked out.
Some of those will getworse if we're being honest.
Yeah.
But our job isn't to figureout why it's wrong per se.
Yeah.
Our job is to be able to say,okay, now I get to do this.
Now I get to do this.
Like now I get to writeRuby Code to make.
Uh, drawer boxes, right?
(35:58):
Yep.
In, in a way that I couldnever do before, right?
Yeah.
Now I get to be able to answerquestions in five seconds.
That would've taken me a week.
Right?
Now I get to doall these things.
What does it mean for me?
What is v two of me knowingthat all, like everything
has changed, right?
And that I can kind of belike, uh, forever extended.
Ryan, do you remember,uh, in the Matrix when,
(36:19):
uh, he downloads Kung fuwhen Kea Reeves, uh, do
downloads kung fu, right?
That classic, oh,I know Kung fu.
I tried for weeksto download Kung fu.
It turns out I, it doesn't work.
Well, so, but my point is like,we now have that ability, right.
So to speak.
Yeah.
Right.
Like we now have thatability, if we don't evolve
and leverage that ability,we will get replaced.
(36:42):
Yeah, of course.
Not by ai, but by theother founders who, who are
actually using ai using it.
Right, right.
And, and using it andembracing it in the way
that it's supposed to be.
And so, so here's whatI would say overall.
I think, you know, whenwe look at the question,
can AI replace founders?
Yes.
If we let it, yes.
(37:03):
If we let it, if we become nomore than just those automatons
that we're doing, you know,rote work, then, then yes,
it can and, and frankly,right, it probably should.
But if we zoom out and say,okay, uh, you know, some
of that's scary, et cetera.
But if I use this, if Iexist as the architect,
the curator, the creativemind, right, the mastermind
(37:23):
behind all these tools, nowI can create 10 x what I
would've ever created before.
Now I can become a founderand, and have the, the outcome
and the output that I'venever even dreamed of before.
At which point weevolve to that person.
Literally anything is possible.
Overthinking your startupbecause you're going it alone.
(37:44):
You don't have to, and honestly,you shouldn't because instead,
you can learn directly frompeers who've been in your shoes.
Connect with bootstrapfounders and the advisors
helping them win in thestartups.com community.
Check out the startups.comcommunity@www.startups.com
to see if it's for you.
Could be just thething you need.
I hope to see you inside.
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.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.