September 11, 2025 • 19 mins
EPISODE 24
A special “best of” highlight reel from our first 20 episodes: we reframe luck as reps and compounding, warn founders about burnout’s hidden interest, and show how AI can supercharge solo building without replacing your voice. We dig into making clear asks and saying “no” to protect focus, why empathy is the real product moat, and how durable edges come from hardware + proprietary data and real-world feedback. You’ll hear scrappy GTM plays (like sub-50 sq ft kiosks and Kickstarter’s customer collision), safety-first execution from nuclear to drones, pre-LLM agents, the 80/20 (AI/human) rule for content, and why creators should monetize ownership—not just virality.
CHAPTERS
00:00 – Manufactured luck & the burnout tax05:04 – Empathy is the moat (AI changes jobs)09:49 – Safety & simplicity: nuclear-ready drones13:52 – Fail-forward roadmaps (hypotheses over hype)16:18 – AI 80/20 and the creator long tail
LINKS
Stay Connected with Founder Mode
Subscribe to our newsletter: foundermode.kit.com
Connect with Kevin
Connect with Jason
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Welcome back to Founder Mode special episode.
This week we're going to do something kind of new and maybe
kind of cool. Yeah, we're going to share our
favorite clips from the 1st 20 episodes of Founder Mode.
Let's get into it. You can't wait.
(00:26):
We open this whole series by redefining luck.
Luck's not a lottery ticket. It's reps, compounding and
surface area. Luck is a skill.
You can manufacture it, Which iskind of counterintuitive, right?
Are people who are lucky actually lucky or did they
create their own luck? People don't see the reps you
put in to go in to get the finaloutcome.
(00:47):
How does luck turn into that compounding sort of effort that
allows you to say here, now, I created the situation where I
was lucky? Founders think they can Sprint
forever. But burnout always collects
interest. I think there's this like
fallacy that I can just run really fast at everything all at
once. Like founders are like, I am
(01:08):
superhuman and I, you know, the laws of physics don't apply to
me. I don't need sleep.
I don't need to stop and eat that meal.
My family will be there wheneverI stop working.
Like it's all going to be fine. I'm just going to run really
fast and, you know, break thingsin the tech startup sense.
But like, the reality is that you're actually breaking stuff.
(01:29):
And like between breaking your own body, your own mind, your
kind of support system, like there's a bunch of casualties.
Alex built A6 figure sass solo. But .01% of the code was
actually typed by him. AI did the rest.
I built creator buddy over the last six months, launched it
about a month ago. I basically took my content
(01:51):
creation system I used over the last three years while I was
working a nine to five, right. I was trying to come up with
content as I, you know, had thisvery strenuous job.
I took that system that allowed me to come up with endless
content ideas, push out tons of content and grow to hundreds of
thousands of followers. I took that system, I turned it
into software, right. And so first thing I was
(02:13):
thinking about when do this, I just want to build an app for my
that makes my life a lot easier.And as I started building it,
I'm like, OK, other people coulduse this and this would bring
tons of value to them. Anxiety isn't a bug, it's fuel.
Bill Russell threw up before every game.
Founders feel the same. Anxiety just has to be there,
and the sooner you accept the fact that it, it will always be
(02:36):
there no matter what, it's thereto help you and not harm you.
If you know how to ride that Bill Russell, one of the most
legendary NBA players ever, he threw up for every single game
he played. That's 1200 something games.
So that anxiety was there when it mattered and when it didn't
matter. So why would a founder in 2025
(02:59):
accept there to be no anxiety ifBill Russell is playing a last
place team like the best player of the era and he's throwing up
because of her nervous he is. The biggest unlock?
Just ask but. Ask with clarity, fairness, and
conviction. People are much more likely to
like fund your company or hire you if you show that you're
(03:19):
trustworthy. And one of the best ways to do
that is through references. I started looking up people at
Shopify that I was interviewing with and I realized that I had a
mutual connection with with the president of the company.
And so I reached out and I askedhim.
I said, hey, if you truly, I believe I'd be a great fit, I'd
(03:42):
appreciate an intro or some sortof like endorsement.
And it was never like, hey, I need a favor, please hype me.
I was really authentic with him.And I said no pressure
whatsoever. I think that's also super
important where it's like, yeah,you want to be grounded.
You want to be clear, you want to make sure your ask is
(04:02):
reasonable, but also give peoplean out.
The word that saves founders No.Boundaries build trust, not
enemies. An absence of answer is in a
sense leaving that bridge open. But if you do like, I mean,
yeah, playing strategically leaving like that open and the
response, right, I think that isthat is the key because it
(04:24):
depends like who's from, right, if it's somebody.
Strategic that I actually. Do want to work with like then?
Yeah, there is definitely a scenario where I'm like, sorry,
like we, we don't have the bandwidth right now.
Like that's that's usually the most honest answer I think.
It's like. If you're running a startup of
any size, I'm sure like regardless of how much resource
you have. You always.
Have a confine in how much output, like it's always a
(04:46):
resource game, right? Then it's just like it's hard to
allocate that unless there's like a budget allocated to it.
Like unless there's revenue on the table, like it's hard to
allocate that right. So I think that is usually the
play, like it's, yeah, I'm like interested, but like down the
road, not right now. From flurry to accompli, Peter
(05:06):
cracked Go to market. It all comes down to empathy.
The biggest single word that I think is the most important
word, which is also probably going to be a relief to people
dealing with the question of will AAI replace my job?
AI won't replace your job, it'llchange your job.
A big difference between a good market and a good product person
is an empathy. And so, and what I mean is like,
(05:29):
can you put yourself in the shoes of that customer and like,
what are their, what's their pain?
What's their aspiration? You know, what are the dynamics
like at work for them to make a decision?
How easy are you making it for them to try you and make a
decision? So like that momentum piece is
really, I think it comes from somehow understanding the
(05:51):
customer really well. If it's.
Just an AI wrapper, it won't last.
The Moat is solving what used tobe impossible.
People think that the bottleneckis actually the hardware when in
reality, it's not the hardware, it's the real like time data
where there's just simply not enough models to kind of train
it on. With software, it's a lot
(06:12):
different because you can just train it off of X or Reddit or
like you know, with Grok, for example, just pulling from all
the X posts. But I think to get real life
footage is, you know, that's whyTesla was ahead of the curve
when they were doing FSD. And you know, then you look at
Waymo or you look at, you know, in the past Cruise for Zooks
(06:34):
where they're having to basically send vehicles as test
vehicles to collect it. I mean, the reason Elon was
ahead of the curve with Tesla isbecause he was able to just have
drivers like us going around. Data for Elon.
For we're the Tesla training team right here at The Trio.
Exactly right. And so I think that was a
brilliant move on his part because everyone else's is
(06:56):
behind. Megan hacked building permits to
launch retail. Under 8 feet and under 50 square
feet. No permit required.
Spy reveal was a pop up company.We basically did small format
kiosk that got deployed pretty much anywhere consumers were.
So sidewalks, hotel lobbies, corporate lobbies, malls, public
plazas, parks, pretty much anywhere.
(07:19):
And it was. This small format boutique that
could get deployed in under an hour where consumers could come
touch and feel products and thentransact directly on site.
And so then I started looking atthe permit and I realized that
if something was under 8 feet and under 50 square feet,
technically it was not a building.
Therefore, it did not need to meet building codes or building
(07:41):
permits. Nice.
It just, I just needed a sales permit and all I need to do is
like the business that it was infront of, they needed to approve
it. And then I needed like a
seller's license and I could pretty much go anywhere I
wanted. So I was just trying to hack the
building permits and that was how I designed the kiosk to be
able to fit within those parameters.
From Outlook Mobile to AI Co pilots.
(08:03):
JJ says every founder now has tothink fast software.
We are. In this weird moment when?
AI is like, really, really useful.
Extremely useful. Almost too useful to be
dangerous, right? But.
At the same time, not very. Reliable, so it cannot really do
things fully autonomously. So the paradigm between sort.
(08:23):
Of like a. Copilot Paradigm and the the.
Independent to another dev on the team paradigm we are.
Quickly moving toward that, but not there yet.
This is why Devon may not work the best today yet for a lot of
tasks, but I'm sure that in a period of time everyone will be
(08:45):
on the other side. Design love means nothing if the
box never ships. Sam built GIR from spatula to
pattern brands. I have this phrase that I that I
didn't coin but that I come backto all the time, which is the
collision with the customer and like no plan survives the
collision with the customer. Launching our Kickstarter was a
pre market collision with the customer because I used to, I
(09:07):
mean, see pre selling spatulas, but I had effectively customers
that I owed, you know, time and attention and ultimately
products too. And and so I think that
crowdfunding is such a fascinating tool for that
reason. Like you get to learn from your
customers, but you also have to support them much earlier on
that you might have yes, my job was product and yes, my job was
(09:29):
building our website and like, and also, yes, my job was
talking to factories, but my jobwas talking to customers and
making them happy, you know, before they had anything.
But they had, you know, they hadn't even given me dollars,
actually pledge to give me theirdollars.
I'm like, I have nothing, but I owe you a lot.
From Hollywood robots to nuclearpower plants, ASA shows why
(09:53):
safety and simplicity win. This stunt double reactor
practice run blew my mind. Yeah, as you can imagine,
they're, they're very, very careful people, those who run
nuclear reactors. Makes sense.
They're they're very careful kind of ramp up to get to the
point where you can, you can go in, fly during an active, active
mission, you know, to the point where where you have to go and
(10:15):
weigh every single part of your drone.
It gets weighed before it goes in and it gets weighed when it
comes out. There had better be the exact
same mask. It has to be able to be used
with glove or essentially sort of hardening it to the point
where you're sending it into outer space.
Almost kind of just hammered on this problem for maybe a year
and a half, something like that to go through all of the
different checks proving that wecould do it inside at our at our
(10:38):
place. And then we go to their training
reactor or they have a full training mock up, which is the
full size thing sitting, you know.
Yeah. And then you know.
Exact same size. Just 100% yeah.
Do that enough times. You know we had a the.
Movie analogy. The film analogy is like the
stunt double. This is the stunt double nuclear
reactor. Yep, I understand this is very
(10:58):
different from from software. You just can't.
You know you can't just. Throw it out there, yeah.
I'm like shipping apps as we were like logging onto the
podcast. Yeah, not quite.
Dennis built AI scheduling agents years before there were
LLMS. The brutal lesson 32,000,000
hand labeled emails. Any startup is probably one of
(11:21):
two things is 1000 things, but it's certainly a startup where
you're trying to solve either some market challenge and or a
science challenge. Try not to solve both at the
same time. It's going to be twice as hard.
So Airbnb is not a technical challenge.
That is you trying to figure outwhether can I have murderers
sleep in my apartment while I'm here and I'm OK with that or
(11:45):
whatever the business plan mightlook like.
Or is it a science challenge like X dot AI for where we at
that moment in 2014 thought we could carve out just a corner of
the language universe, build a set of models where we could
navigate that pre LLM for where it kind of looks like we solved
language, certainly to a large screen now, but certainly not
(12:07):
back then. So to answer your question,
we're in that bucket where we were brave enough or foolish
enough to believe we carve out that particular corner for where
we could create this agent that could converse on scheduling
only. And we did it on a labeled data
set which we assembled, so 32,000,000 hand labeled items.
(12:33):
Founders don't need perfect days.
It's just a daily 15 minutes that compounds.
Small wins everyday push ups count.
Think of like. Your your body and.
When you wake up in the morning is like, what's the first thing
I need to do for my? Body in a healthy weight.
Mentally and physically, you know, and that would be probably
(12:54):
still, you know. Whether?
Whether? It's just getting on the.
Floor doing some abs. Or you know, doing a set of push
ups, it's it starts there, the small things will create.
You don't need a gym, you know. It really does, you know, And it
goes across the board with the eating, too, yeah.
Suja says stop selling pieces, ship outcomes.
Build applications, not tool kits.
(13:16):
People want an. Uber.
And right now they're being soldcar parts.
They're like, here you go, there's a steering wheel,
there's a here's some engine, gobuild it yourself.
Yeah, bro, I just. Want to?
Go from point A to point B. Like just give me.
An Uber. And get me there.
Yeah. And I think.
That's the opportunity right now, right that.
You don't. Need to sell the tools.
(13:38):
You want to be able to think through.
The AI. Applications from like first.
Principles and be able to deliver.
The transportation is what the clients want, not the tools that
they need. To go and do that with Angus
carries a road map of hypothesis.
Fail forward fast and travel 0 drag.
I think it's there's really liketwo different situations here.
(14:01):
Internally, it needs to be aboutyour road map of hypothesis
instead of like every single thing we build is going to is
going to crush. I actually think like I like to
pull the failure forward. Every PM should be failing first
thing in the morning. Like you should be hitting the
model, like seeing whether it can do the thing you want it to
do. And so like, I think bring those
failures forward and then learn whether you're going to make it
(14:23):
do what you want it to do. The Moat isn't hype, it's
hardware and proprietary data. Hybrid, old, smart, fast,
fearless. Let's.
Start from the software like. Not only do I have to build just
the best software that people are that want to use, they like
to use as an API everything thata software company has to do,
but we also have to like make something that physically works
(14:45):
in the water that you open up that has, you know, for us a
cartridge and a battery, it degrades over time.
It's sitting in water, it gets waterlogged.
Like no software company has those issues, right?
So if you get the hardware right, you now have a Moat is is
kind of thing. Number one thing too is that you
actually get a proprietary data mechanism, right?
(15:06):
It's probably one way to think about it.
Like The thing is inside the water and it's collecting data
that otherwise a SAS, like just a pure chemistry water SAS
company would never have access to unless they asked Jason, hey,
can you go in your pool, measureyour pool water and give us the
data? But now I'm making you the human
do something. And so I think that's kind of
the real benefit of having both the hardware, software and a
(15:28):
service all kind of combined into one.
Ride with the people who'd follow you anywhere.
Boomerangs are rare. Build with your Rider dice.
It's really hard to make the return trip work when it's the
same company. And so the difference between
these two concepts are bringing people back in that have left.
They left for a reason. Whether you let them go or they
(15:49):
quit, there's a reason they left.
And maybe they're at a differentplace in their journey.
But I think it's kind of like putting people on a PIP like,
yeah, a lot of them never make it out.
And so it's, it's a very rare case that the boomerang thing
works well from my personal experience than others I've
seen. And then the boomer ride thing
ultimately, like, you know, the people that are good, you've
worked them a long time and those are the ones you bring
(16:10):
along with you and, and that's your crew.
And sometimes people kind of fall off the boat and maybe they
get to get back on, but maybe not.
AI gets you to 80%. Your voice is the final 20.
Don't forget distribution and your rolodex.
I think the, the people that useit, the best that I've seen are
(16:31):
they follow my 8020 rule. So I don't think that any AI is
that where the point where you can just be 100% like it can
talk for you, do everything for you.
So for me and the people that I work with, they I just want to
get it to 80%. So if I'm right at using it to
write a post or write a VSL or write a, you know, a YouTube
(16:52):
script, whatever it is, I want to get it to 80% percent, you
know, 80% AI, 20% Chris. And the people that I see that
are doing that, like there's some you're still putting some
emphasis on the real part of, you know, you.
I think we're getting really close to major major major major
(17:14):
revolt against all just AI generated slot.
Most creators live in the long tail.
Own your business, not just yourcontent.
Monetization beats going viral every time.
Like everybody sees. Mr. Beast and is like I want to
be Mr. Beast but. Mr. Beast, by definition is a
.0001% of the greater economy, right?
(17:35):
The. Vast majority of people are on
the long tail, right? And the long tail is you.
Have these viral? Moments where you have millions
of views, but the payouts you get from these platforms are
like. Dollars, maybe 10s.
Of dollars. For like.
Millions of views. So the platform, the social
media platforms. Are able to generate.
All sorts of viewership serve all sorts of ads.
And give you like a DE. Minimis payment off the back of
(17:56):
that, even though your product or what you put out there had
just gone viral, right? Our thesis around Kajabi is
like, those moments are great. And we're not saying don't have
those moments because distribution is amazing.
But the question is what do you put behind that that makes it
sustainable and helps you? Capture that value.
And turn into a business. Kevin, I can't believe we've
done 20 episodes. I mean more than that by now,
(18:18):
actually. It's pretty cool.
I'm, I'm stoked man. What do you think the next 20
are going to be like? Pretty good.
I think they're going to be better.
I think they're going to be better.
I think we kind of like got our teething done with the 1st 20.
And I think now it's it's really, you know, sort of
scaling this out and I'm hyped for the next 20 because I think
some of these are going to be inperson and, you know, kicking it
up a notch with studio. Yeah, stay tuned.
(18:39):
Subscribe, like Reshare, review,comment.
Tell us what you love from the 1st 20 episodes, what you hated,
what you want to see more of, and maybe we'll bring a guest or
two back to hear what they've been up to since the last time
they were here. Awesome talk soon, Jason.
Welcome back to Founder Mode special episode.
This week we're going to do something kind of new and maybe
kind of cool. Yeah, we're going to share our
favorite clips from the 1st 20 episodes of Founder Mode.
Let's get into it. You can't wait.
(00:26):
We open this whole series by redefining luck.
Luck's not a lottery ticket. It's reps, compounding and
surface area. Luck is a skill.
You can manufacture it, Which iskind of counterintuitive, right?
Are people who are lucky actually lucky or did they
create their own luck? People don't see the reps you
put in to go in to get the finaloutcome.
(00:47):
How does luck turn into that compounding sort of effort that
allows you to say here, now, I created the situation where I
was lucky? Founders think they can Sprint
forever. But burnout always collects
interest. I think there's this like
fallacy that I can just run really fast at everything all at
once. Like founders are like, I am
(01:08):
superhuman and I, you know, the laws of physics don't apply to
me. I don't need sleep.
I don't need to stop and eat that meal.
My family will be there wheneverI stop working.
Like it's all going to be fine. I'm just going to run really
fast and, you know, break thingsin the tech startup sense.
But like, the reality is that you're actually breaking stuff.
(01:29):
And like between breaking your own body, your own mind, your
kind of support system, like there's a bunch of casualties.
Alex built A6 figure sass solo. But .01% of the code was
actually typed by him. AI did the rest.
I built creator buddy over the last six months, launched it
about a month ago. I basically took my content
(01:51):
creation system I used over the last three years while I was
working a nine to five, right. I was trying to come up with
content as I, you know, had thisvery strenuous job.
I took that system that allowed me to come up with endless
content ideas, push out tons of content and grow to hundreds of
thousands of followers. I took that system, I turned it
into software, right. And so first thing I was
(02:13):
thinking about when do this, I just want to build an app for my
that makes my life a lot easier.And as I started building it,
I'm like, OK, other people coulduse this and this would bring
tons of value to them. Anxiety isn't a bug, it's fuel.
Bill Russell threw up before every game.
Founders feel the same. Anxiety just has to be there,
and the sooner you accept the fact that it, it will always be
(02:36):
there no matter what, it's thereto help you and not harm you.
If you know how to ride that Bill Russell, one of the most
legendary NBA players ever, he threw up for every single game
he played. That's 1200 something games.
So that anxiety was there when it mattered and when it didn't
matter. So why would a founder in 2025
(02:59):
accept there to be no anxiety ifBill Russell is playing a last
place team like the best player of the era and he's throwing up
because of her nervous he is. The biggest unlock?
Just ask but. Ask with clarity, fairness, and
conviction. People are much more likely to
like fund your company or hire you if you show that you're
(03:19):
trustworthy. And one of the best ways to do
that is through references. I started looking up people at
Shopify that I was interviewing with and I realized that I had a
mutual connection with with the president of the company.
And so I reached out and I askedhim.
I said, hey, if you truly, I believe I'd be a great fit, I'd
(03:42):
appreciate an intro or some sortof like endorsement.
And it was never like, hey, I need a favor, please hype me.
I was really authentic with him.And I said no pressure
whatsoever. I think that's also super
important where it's like, yeah,you want to be grounded.
You want to be clear, you want to make sure your ask is
(04:02):
reasonable, but also give peoplean out.
The word that saves founders No.Boundaries build trust, not
enemies. An absence of answer is in a
sense leaving that bridge open. But if you do like, I mean,
yeah, playing strategically leaving like that open and the
response, right, I think that isthat is the key because it
(04:24):
depends like who's from, right, if it's somebody.
Strategic that I actually. Do want to work with like then?
Yeah, there is definitely a scenario where I'm like, sorry,
like we, we don't have the bandwidth right now.
Like that's that's usually the most honest answer I think.
It's like. If you're running a startup of
any size, I'm sure like regardless of how much resource
you have. You always.
Have a confine in how much output, like it's always a
(04:46):
resource game, right? Then it's just like it's hard to
allocate that unless there's like a budget allocated to it.
Like unless there's revenue on the table, like it's hard to
allocate that right. So I think that is usually the
play, like it's, yeah, I'm like interested, but like down the
road, not right now. From flurry to accompli, Peter
(05:06):
cracked Go to market. It all comes down to empathy.
The biggest single word that I think is the most important
word, which is also probably going to be a relief to people
dealing with the question of will AAI replace my job?
AI won't replace your job, it'llchange your job.
A big difference between a good market and a good product person
is an empathy. And so, and what I mean is like,
(05:29):
can you put yourself in the shoes of that customer and like,
what are their, what's their pain?
What's their aspiration? You know, what are the dynamics
like at work for them to make a decision?
How easy are you making it for them to try you and make a
decision? So like that momentum piece is
really, I think it comes from somehow understanding the
(05:51):
customer really well. If it's.
Just an AI wrapper, it won't last.
The Moat is solving what used tobe impossible.
People think that the bottleneckis actually the hardware when in
reality, it's not the hardware, it's the real like time data
where there's just simply not enough models to kind of train
it on. With software, it's a lot
(06:12):
different because you can just train it off of X or Reddit or
like you know, with Grok, for example, just pulling from all
the X posts. But I think to get real life
footage is, you know, that's whyTesla was ahead of the curve
when they were doing FSD. And you know, then you look at
Waymo or you look at, you know, in the past Cruise for Zooks
(06:34):
where they're having to basically send vehicles as test
vehicles to collect it. I mean, the reason Elon was
ahead of the curve with Tesla isbecause he was able to just have
drivers like us going around. Data for Elon.
For we're the Tesla training team right here at The Trio.
Exactly right. And so I think that was a
brilliant move on his part because everyone else's is
(06:56):
behind. Megan hacked building permits to
launch retail. Under 8 feet and under 50 square
feet. No permit required.
Spy reveal was a pop up company.We basically did small format
kiosk that got deployed pretty much anywhere consumers were.
So sidewalks, hotel lobbies, corporate lobbies, malls, public
plazas, parks, pretty much anywhere.
(07:19):
And it was. This small format boutique that
could get deployed in under an hour where consumers could come
touch and feel products and thentransact directly on site.
And so then I started looking atthe permit and I realized that
if something was under 8 feet and under 50 square feet,
technically it was not a building.
Therefore, it did not need to meet building codes or building
(07:41):
permits. Nice.
It just, I just needed a sales permit and all I need to do is
like the business that it was infront of, they needed to approve
it. And then I needed like a
seller's license and I could pretty much go anywhere I
wanted. So I was just trying to hack the
building permits and that was how I designed the kiosk to be
able to fit within those parameters.
From Outlook Mobile to AI Co pilots.
(08:03):
JJ says every founder now has tothink fast software.
We are. In this weird moment when?
AI is like, really, really useful.
Extremely useful. Almost too useful to be
dangerous, right? But.
At the same time, not very. Reliable, so it cannot really do
things fully autonomously. So the paradigm between sort.
(08:23):
Of like a. Copilot Paradigm and the the.
Independent to another dev on the team paradigm we are.
Quickly moving toward that, but not there yet.
This is why Devon may not work the best today yet for a lot of
tasks, but I'm sure that in a period of time everyone will be
(08:45):
on the other side. Design love means nothing if the
box never ships. Sam built GIR from spatula to
pattern brands. I have this phrase that I that I
didn't coin but that I come backto all the time, which is the
collision with the customer and like no plan survives the
collision with the customer. Launching our Kickstarter was a
pre market collision with the customer because I used to, I
(09:07):
mean, see pre selling spatulas, but I had effectively customers
that I owed, you know, time and attention and ultimately
products too. And and so I think that
crowdfunding is such a fascinating tool for that
reason. Like you get to learn from your
customers, but you also have to support them much earlier on
that you might have yes, my job was product and yes, my job was
(09:29):
building our website and like, and also, yes, my job was
talking to factories, but my jobwas talking to customers and
making them happy, you know, before they had anything.
But they had, you know, they hadn't even given me dollars,
actually pledge to give me theirdollars.
I'm like, I have nothing, but I owe you a lot.
From Hollywood robots to nuclearpower plants, ASA shows why
(09:53):
safety and simplicity win. This stunt double reactor
practice run blew my mind. Yeah, as you can imagine,
they're, they're very, very careful people, those who run
nuclear reactors. Makes sense.
They're they're very careful kind of ramp up to get to the
point where you can, you can go in, fly during an active, active
mission, you know, to the point where where you have to go and
(10:15):
weigh every single part of your drone.
It gets weighed before it goes in and it gets weighed when it
comes out. There had better be the exact
same mask. It has to be able to be used
with glove or essentially sort of hardening it to the point
where you're sending it into outer space.
Almost kind of just hammered on this problem for maybe a year
and a half, something like that to go through all of the
different checks proving that wecould do it inside at our at our
(10:38):
place. And then we go to their training
reactor or they have a full training mock up, which is the
full size thing sitting, you know.
Yeah. And then you know.
Exact same size. Just 100% yeah.
Do that enough times. You know we had a the.
Movie analogy. The film analogy is like the
stunt double. This is the stunt double nuclear
reactor. Yep, I understand this is very
(10:58):
different from from software. You just can't.
You know you can't just. Throw it out there, yeah.
I'm like shipping apps as we were like logging onto the
podcast. Yeah, not quite.
Dennis built AI scheduling agents years before there were
LLMS. The brutal lesson 32,000,000
hand labeled emails. Any startup is probably one of
(11:21):
two things is 1000 things, but it's certainly a startup where
you're trying to solve either some market challenge and or a
science challenge. Try not to solve both at the
same time. It's going to be twice as hard.
So Airbnb is not a technical challenge.
That is you trying to figure outwhether can I have murderers
sleep in my apartment while I'm here and I'm OK with that or
(11:45):
whatever the business plan mightlook like.
Or is it a science challenge like X dot AI for where we at
that moment in 2014 thought we could carve out just a corner of
the language universe, build a set of models where we could
navigate that pre LLM for where it kind of looks like we solved
language, certainly to a large screen now, but certainly not
(12:07):
back then. So to answer your question,
we're in that bucket where we were brave enough or foolish
enough to believe we carve out that particular corner for where
we could create this agent that could converse on scheduling
only. And we did it on a labeled data
set which we assembled, so 32,000,000 hand labeled items.
(12:33):
Founders don't need perfect days.
It's just a daily 15 minutes that compounds.
Small wins everyday push ups count.
Think of like. Your your body and.
When you wake up in the morning is like, what's the first thing
I need to do for my? Body in a healthy weight.
Mentally and physically, you know, and that would be probably
(12:54):
still, you know. Whether?
Whether? It's just getting on the.
Floor doing some abs. Or you know, doing a set of push
ups, it's it starts there, the small things will create.
You don't need a gym, you know. It really does, you know, And it
goes across the board with the eating, too, yeah.
Suja says stop selling pieces, ship outcomes.
Build applications, not tool kits.
(13:16):
People want an. Uber.
And right now they're being soldcar parts.
They're like, here you go, there's a steering wheel,
there's a here's some engine, gobuild it yourself.
Yeah, bro, I just. Want to?
Go from point A to point B. Like just give me.
An Uber. And get me there.
Yeah. And I think.
That's the opportunity right now, right that.
You don't. Need to sell the tools.
(13:38):
You want to be able to think through.
The AI. Applications from like first.
Principles and be able to deliver.
The transportation is what the clients want, not the tools that
they need. To go and do that with Angus
carries a road map of hypothesis.
Fail forward fast and travel 0 drag.
I think it's there's really liketwo different situations here.
(14:01):
Internally, it needs to be aboutyour road map of hypothesis
instead of like every single thing we build is going to is
going to crush. I actually think like I like to
pull the failure forward. Every PM should be failing first
thing in the morning. Like you should be hitting the
model, like seeing whether it can do the thing you want it to
do. And so like, I think bring those
failures forward and then learn whether you're going to make it
(14:23):
do what you want it to do. The Moat isn't hype, it's
hardware and proprietary data. Hybrid, old, smart, fast,
fearless. Let's.
Start from the software like. Not only do I have to build just
the best software that people are that want to use, they like
to use as an API everything thata software company has to do,
but we also have to like make something that physically works
(14:45):
in the water that you open up that has, you know, for us a
cartridge and a battery, it degrades over time.
It's sitting in water, it gets waterlogged.
Like no software company has those issues, right?
So if you get the hardware right, you now have a Moat is is
kind of thing. Number one thing too is that you
actually get a proprietary data mechanism, right?
(15:06):
It's probably one way to think about it.
Like The thing is inside the water and it's collecting data
that otherwise a SAS, like just a pure chemistry water SAS
company would never have access to unless they asked Jason, hey,
can you go in your pool, measureyour pool water and give us the
data? But now I'm making you the human
do something. And so I think that's kind of
the real benefit of having both the hardware, software and a
(15:28):
service all kind of combined into one.
Ride with the people who'd follow you anywhere.
Boomerangs are rare. Build with your Rider dice.
It's really hard to make the return trip work when it's the
same company. And so the difference between
these two concepts are bringing people back in that have left.
They left for a reason. Whether you let them go or they
(15:49):
quit, there's a reason they left.
And maybe they're at a differentplace in their journey.
But I think it's kind of like putting people on a PIP like,
yeah, a lot of them never make it out.
And so it's, it's a very rare case that the boomerang thing
works well from my personal experience than others I've
seen. And then the boomer ride thing
ultimately, like, you know, the people that are good, you've
worked them a long time and those are the ones you bring
(16:10):
along with you and, and that's your crew.
And sometimes people kind of fall off the boat and maybe they
get to get back on, but maybe not.
AI gets you to 80%. Your voice is the final 20.
Don't forget distribution and your rolodex.
I think the, the people that useit, the best that I've seen are
(16:31):
they follow my 8020 rule. So I don't think that any AI is
that where the point where you can just be 100% like it can
talk for you, do everything for you.
So for me and the people that I work with, they I just want to
get it to 80%. So if I'm right at using it to
write a post or write a VSL or write a, you know, a YouTube
(16:52):
script, whatever it is, I want to get it to 80% percent, you
know, 80% AI, 20% Chris. And the people that I see that
are doing that, like there's some you're still putting some
emphasis on the real part of, you know, you.
I think we're getting really close to major major major major
(17:14):
revolt against all just AI generated slot.
Most creators live in the long tail.
Own your business, not just yourcontent.
Monetization beats going viral every time.
Like everybody sees. Mr. Beast and is like I want to
be Mr. Beast but. Mr. Beast, by definition is a
.0001% of the greater economy, right?
(17:35):
The. Vast majority of people are on
the long tail, right? And the long tail is you.
Have these viral? Moments where you have millions
of views, but the payouts you get from these platforms are
like. Dollars, maybe 10s.
Of dollars. For like.
Millions of views. So the platform, the social
media platforms. Are able to generate.
All sorts of viewership serve all sorts of ads.
And give you like a DE. Minimis payment off the back of
(17:56):
that, even though your product or what you put out there had
just gone viral, right? Our thesis around Kajabi is
like, those moments are great. And we're not saying don't have
those moments because distribution is amazing.
But the question is what do you put behind that that makes it
sustainable and helps you? Capture that value.
And turn into a business. Kevin, I can't believe we've
done 20 episodes. I mean more than that by now,
(18:18):
actually. It's pretty cool.
I'm, I'm stoked man. What do you think the next 20
are going to be like? Pretty good.
I think they're going to be better.
I think they're going to be better.
I think we kind of like got our teething done with the 1st 20.
And I think now it's it's really, you know, sort of
scaling this out and I'm hyped for the next 20 because I think
some of these are going to be inperson and, you know, kicking it
up a notch with studio. Yeah, stay tuned.
(18:39):
Subscribe, like Reshare, review,comment.
Tell us what you love from the 1st 20 episodes, what you hated,
what you want to see more of, and maybe we'll bring a guest or
two back to hear what they've been up to since the last time
they were here. Awesome talk soon, Jason.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com