Latest / AI for Founders with Ryan Estes / Stop designing screens. Start designing patterns.

Transcript

Stop designing screens. Start designing patterns. · AI for Founders with Ryan Estes

0:00
Most startups don't break because they lack talent.
0:02
They break because the product quietly splits into parallel
0:05
universes. One team ships a feature,
0:08
another ships the same feature differently.
0:10
Both are right. Everyone's moving fast and no
0:13
one is moving together. Chris Stroll saw this up close
0:16
and decided chaos was not a growth strategy.
0:18
Before Co founding Knapsack, he watched high performing teams
0:21
drown and edge cases, forks and silent product drift.
0:25
Same company, same road map, multiple realities.
0:29
The fix wasn't more syncs, it was fewer interpretations.
0:32
Knapsack became infrastructure, a single system where product,
0:36
design and engineering all agree before shipping.
0:39
Enterprise adoption. Multi X growth without burning
0:42
people out. And here's the founder lesson.
0:44
Hiding underneath scaling isn't about better judgement.
0:48
It's about building systems so judgement isn't constantly
0:51
required. So sit with this.
0:53
What part of your business only worked because you're still
0:56
manually resolving contradictions every day?
0:59
Welcome to AI for founders. I'm Ryan Estes.
1:01
Stick around and you'll learn how real teams scale without
1:04
fragmentation, why systems beat heroics, and how founders stop
1:08
being the merge conflict. This is for founders, operators,
1:12
product leaders and anyone building past the early stage
1:15
chaos. And if it's made you slightly
1:16
uncomfortable, perfect. This is the newsletter.
1:19
Join us at AI for founders.co and leave a review of to save
1:23
you from one unnecessary decision this week.
1:25
Now most founders think podcasting is about attention.
1:29
It's actually about trust. Kick caster gets founders book
1:32
where trust already exists. These are shows your customers
1:35
already listened to, conversations that sound like
1:38
thinking, not pitching. Kickcaster handles the research,
1:41
the outreach, the pitching, scheduling and the follow-ups.
1:44
You just show up and talk like a human.
1:46
Clients use podcast guesting to close deals, raise capital,
1:50
recruit talent and become familiar before the 1st sales
1:53
call. If you want distribution without
1:55
becoming a content influencer, go to kickcaster.com.
1:59
Talk Once Compound foreverthatskickcaster.com by MO
2:03
Burst. Bruce, thank you so much for
2:21
joining me. Welcome to the show.
2:22
Hey. Ryan, how you doing?
2:24
Doing great. OK, before we get started here I
2:26
want to start with some important questions.
2:29
I was reading a bit about you and it sounds like you are a
2:32
pretty accomplished DND dungeon master.
2:34
Yeah, yeah, I it's funny, I actually got a 3D printer for
2:38
Christmas and so I'm currently printing a giant piece of
2:42
terrain. It is a 7 foot tall castle that
2:45
is going to be the set piece for our new campaign.
2:50
Tremendous. Now, which poses the question,
2:53
how are we feeling about Stranger Things right now?
2:56
So I haven't watched the newest episodes.
2:57
I was actually on vacation at at a beach house where like we
3:01
didn't spend a lot of time on the television.
3:03
But I I've watched up until the early December episodes and it's
3:08
been pretty fun, though I hear the new stuff like there's all
3:10
this like social media hate going on about it, but I don't
3:13
know. I'll judge for myself in a
3:14
couple days. Totally.
3:16
Well, I won't spoiler it for you, but basically we are
3:19
waiting for tomorrow. Tomorrow the series finale hits,
3:23
you know, which I'm pretty stoked about.
3:25
And you know that there's there's plenty to gripe about
3:29
and there's some great kind of fan theories of where it's
3:33
going. And I won't speculate too much
3:36
on that, but there is a going theory that like the whole thing
3:40
wraps up and they're back at the DND table.
3:43
Yeah, I heard this. Yeah.
3:44
Yeah. Where it's like the IT was all
3:46
just a dream type thing. Yeah, yeah.
3:48
I don't know. I hope they don't do that.
3:49
That's kind of a cop out. Yeah, if they do that, man, I
3:53
mean, there are some kind of like hat tips to it because
3:57
there's always recapitulation, you know what I'm saying?
4:00
So like, they have to explain stuff all the time.
4:02
They'll like stop and be like, wait, I have an idea.
4:04
There's actually a second dimension in a wormhole and
4:07
blah, blah, blah. Like so they're like, and it's
4:09
always Dustin, you know what I mean?
4:10
Like, so maybe that happens. I feel like that would be like
4:14
kind of a major cop. Out I mean I kind of love that
4:17
the show does follow that format though, where you have like, you
4:21
know, you're useful idiot character or you're super genius
4:23
character that's able to kind of like help the audience follow
4:26
along. I like, I don't know modern
4:28
writing is what it is. I, I'm a big stupid TV show
4:31
buff. And I think that, like what the
4:34
plague of modern television is, is that you have to cater to an
4:37
audience that's also on their phone while they're watching TV.
4:40
And so you have to have people like explain exactly what a
4:42
character is doing at every given moment.
4:44
And I get, I get annoyed by that, but at the same time, like
4:47
the writing and Stranger Things, at least to date, is pretty
4:50
good. So you know, right?
4:52
It's great and you know, I like poking fun at it, but also, boy,
4:55
I love it. I'm, I'm gonna really miss it,
4:58
you know, largely because I love that it put D&D culture on a
5:02
massive stage, particularly framed around the 80s when it
5:06
was seen as like some kind of station Satan worshipping cult.
5:09
And they have fun with that. But also, I'm a long time float
5:12
tank guy. So when it came out, I was like,
5:15
Oh my God, they're using the tank to transport dimensions.
5:18
I was in heaven, you know, and Mr. Clark is actually a good
5:22
friend of mine. Shout out to Randy.
5:23
I know he listens to the show. So I was really happy to see
5:26
he's got like an integral role in the show this season.
5:29
And I I'm hoping like he's got a, he's got a big component to
5:34
the finale. Nice, nice.
5:36
Yeah, that's cool. Nice man.
5:40
Well, I'm really excited to talk to you largely based around your
5:43
company knapsack, which has had an absolutely meteoric rise.
5:47
Congratulations. I see you just raised a seed in
5:50
October for another. Round actually, so.
5:54
Yeah, series, series. A yeah, yeah.
5:56
Series A yeah. So amazing.
5:59
No, super excited. I mean, welcoming a new investor
6:01
is always awesome. Getting Mike Rosengarten from
6:05
from builders over. He was formerly Camino and Open.
6:08
Gov Mike's a great partner for us, awesome board member, you
6:12
know, $10 million to to go make the next stage of the evolution
6:15
happen, which we're really excited about.
6:18
So exciting and I I want to know how you're going to spend it.
6:21
You got a big bag of cash. But first fill us in.
6:23
What does Knapsack do? Yeah, so we we kind of started
6:27
in this this idea where Evan, my Co founder and I, we're into big
6:33
digital production like we did lots of like large scale web
6:36
builds for big product companies.
6:37
So, you know, if it was like a Major League Soccer, the NFLNBA,
6:43
we did a bunch of stuff for like big retail companies like Estee
6:46
Lauder, Johnson and Johnson, and we got really enamoured with how
6:50
enterprise builds product. And we've all sort of built
6:53
product the same way for like the last 30 years, right?
6:56
You have somebody that that conceives of a problem.
7:00
That problem can be based in research, can be based in
7:02
intuition, but you need to go make a feature.
7:04
And so you go and you say like, all right, well, how do I, how
7:06
do I make that feature? How do I solve that problem?
7:08
There's a bunch of people that understand user experience, that
7:11
understand research, that go and do that research that creates a
7:13
bunch of comps. Those comps are an abstraction
7:15
of the thing you're going to build.
7:17
That gets handed over to a bunch of engineers.
7:18
The engineers make the thing. Everybody fights and decides
7:21
what actually gets released. And then some executive says
7:24
like, no, no, no, it's not that at all.
7:26
And then everybody asks to rework things.
7:27
And then eventually you end up with a product that it often
7:29
times takes months or years after that initial concept or
7:32
idea was initially thought of. And the distance between the
7:35
person that they thought of that user problem and the creation of
7:39
that and something that a user can touch and see is often very,
7:42
very far apart. And So what Evan and I thought
7:45
about is like, well, what if we were able to bring those things
7:47
a lot closer together? And our first sort of foray into
7:51
this was in the area of design systems.
7:53
And so design system is, is this horribly named concept for how
7:58
do you create like a central system of record for all of the
8:00
building blocks of how you make product.
8:04
And so think about it as a giant like well organized bucket of
8:07
Lego bricks. And then if you're making a
8:09
Harry Potter thing, or you're making a Marvel thing, or you're
8:12
making a Star Wars thing, you can use that same bucket of
8:15
bricks to go make all of these different products out of the
8:19
the bricks that you create. And so that was sort of the the
8:22
iteration 1 of knapsack was like, let's build a really great
8:25
way to store and house all these different bricks.
8:27
And let's wrap it all in structured data.
8:29
And now what we transform from that is like, well, you know,
8:33
instead of humans fabricating all these bricks, we want to
8:37
have AI fabricate these bricks. And then we want to be able to
8:40
use AI to intelligently construct them based on saying I
8:43
want a pirate ship or I want a castle, or I want a dragon.
8:46
And so that's what you should be able to tell.
8:48
AIAI should be able to reach into your product system,
8:51
understand how products are constructed at your enterprise,
8:54
and then make that representation of that product
8:56
for you in such a way that the person that is actually thinking
8:59
about that user problem could conceivably ship something to a
9:02
user. Brilliant.
9:04
I love it. And it particularly like a
9:07
resonates with me because, you know, being a designer for a
9:10
long time, I've really kind of like brought my design
9:14
philosophy to no design. Like unless I'm building
9:18
something for myself, I don't want the Ryan Estes experience
9:21
to be reflected in absolutely anything, you know, rather, I
9:24
want your vision to be reproduced as clearly as
9:27
possible. So it, you know, using this kind
9:30
of like methodology, then essentially the designer, the
9:34
the programmer, anybody in kind of like, you know, any
9:37
stakeholder can basically start to piece together product
9:40
without having to worry about like, is this going to cohere
9:44
with the rest of the brand standards?
9:46
What's up everybody? Let me hit you with some math
9:48
that will blow your mind. A wealth manager using Warmstart
9:52
dot AI grew their assets under management by 8.5 million in
9:56
three months. Their monthly fee 500 bucks.
9:59
That's a 1700% return on investment in three months.
10:03
And I get it, another AI tool, another monthly subscription.
10:07
But here's what makes Warm Start different.
10:09
It's not trying to find you new people to bug.
10:12
It's helping you reconnect with people who already know you.
10:15
As a founder, your biggest asset isn't your product or your
10:18
website. It's your relationships.
10:20
But most of us are terrible at nurturing them systematically.
10:23
Warm Start connects your e-mail, identifies warm contacts, and
10:27
helps you craft personalized outreach.
10:29
That doesn't sound like a robot wrote it.
10:32
The best part? You review every message before
10:34
it goes out. No automation without
10:36
permission. If you're ready to stop leaving
10:38
money in your inbox, try warmstart dot AI.
10:42
Use code AIFF for 50 bucks off your first month because the ROI
10:48
on your existing relationships is probably better than any
10:51
investment you'll make this year.
10:53
That's warmstart dot AI and use code AIFF.
11:01
This episode is brought to you by Moeburst, the AI first growth
11:04
engine trusted by iconic startups like Calm, Robin Hood,
11:08
Reddit, Uber and Google. Moeburst is a digital, full
11:12
service, mobile first marketing agency that helps companies to
11:16
scale and become category leaders.
11:19
Whether you're finding product market fit or ready to
11:21
accelerate growth, MO Burst delivers creative, data-driven
11:25
solutions that help with scale and efficiency and impact.
11:29
At MO Burst, we're not just guessing what works, we're
11:32
predicting it. As an AI first agency, MO Burst
11:35
uses cutting edge tools to analyze user behavior, adapt
11:39
creative in real time, and uncover hidden growth
11:42
opportunities. The result?
11:44
Marketing that's faster, smarter, and built to convert
11:47
from early stage start-ups to global brands.
11:50
We've helped hundreds of companies overcome complex
11:53
challenges, unlock new audiences, and optimize every
11:56
metric that matters. At MO Burst, we believe that
11:59
creative thinking, advanced technology, and data are the
12:02
core drivers of success at scale.
12:05
If you're serious about scaling your business with precision and
12:08
purpose, it's time to talk to MO Burst.
12:11
Visit moburst.com and Let's build something category
12:14
defining together. Right.
12:18
So you have this like provenance that goes into the creation of
12:20
anything, right? Like there's a million different
12:22
ways to design A button, but like we obviously don't need to
12:24
redesign a button every time. And that provenance is based on
12:28
the history of that company. So every Figma file that's ever
12:31
been used, every digital product that's ever been launched, AI is
12:34
really great at going out there and looking at all that and
12:37
saying like, what are the commonalities?
12:38
What are the things that represent like the button for
12:41
this particular enterprise? And then you want to be able to
12:45
say like, let's put structured data around that.
12:48
That gives me a set of allowable variants for that button.
12:51
And so that gives you this ability to say like, hey, in
12:54
this particular use case, I want an action button that denotes
12:57
like urgency or something like that.
12:59
And here's the 25 examples that this particular configuration
13:02
draws from. And so, you know, I think that
13:06
one of the things that is often times a big part of this
13:09
discussion is like, well, like, isn't that just a proxy for
13:11
copying? And the reality is, is that you
13:15
probably don't need to comp at all in a lot of cases, kind of
13:19
like what you said, like, hey, the best philosophy design is no
13:21
design. Instead, you should be able to
13:23
just say like, hey, I want a button to go here, and it should
13:26
be able to actually use the source code of a button that
13:29
would actually be acceptable in user land in that particular
13:33
case. Does that make sense?
13:36
Percent. I love it and I I want to
13:38
introduce to the sign for the folks that are watching on
13:40
Spotify. You see it over Chris's
13:41
shoulder. It says build using patterns.
13:44
Talk to me about that. Yeah, so we kind of generally
13:47
believe that at Knapsack there's a deep philosophical reverence
13:50
for patterns. Christopher Alexander wrote a
13:53
book called Pattern Language. It was originally about
13:56
architecture and all the different things that we see in
13:59
homes and buildings. So every house has walls, it has
14:02
floors, it has windows, it has doors.
14:04
Those are all of the different descriptive patterns that we use
14:06
to think about the living spaces that we have.
14:09
Much like a living space, a digital product also has a lot
14:12
of commonalities. You have buttons, links, heroes,
14:15
navs, whirly, Turner's lists, cards, all those different
14:19
aspects of interfaces that represent how we interact with
14:23
our our products. Also in the age of AI, you have
14:26
things like prompts and there's a lot of things that are off
14:28
screen now and all this other stuff like that.
14:29
But all of these things are able to be described as systems of
14:33
patterns that have a set of rules and a set of constraints
14:36
that govern them. And what we've done inside of
14:38
our company is created a system of structured data to help
14:42
describe those rules and those constraints and enable people to
14:45
be able to build using patterns much more efficiently than they
14:48
would be able to design or create something from scratch
14:50
every time. When I think of patterns, I
14:54
think of rhythm. When I think of rhythm, I think
14:55
of music, you know? And you know, maybe I'm an
14:59
Einsteinian, but I kind of feel like fundamentally everything is
15:03
music. Using that kind of analogy and
15:06
metaphor, what's what's good product design in terms of
15:09
music? Yeah, I mean, so it's
15:12
interesting, right? Like music is, is math, right.
15:14
And so when you look at at the way that scales are constructed
15:17
or you look at music theory, David Byrne actually wrote this
15:20
incredible book, The Theory of Music or something like that.
15:26
The title escapes me at this exact moment.
15:28
Instructs like why we hear the things we hear in music, and why
15:32
you have different overlapping sounds and patterns that
15:34
ultimately make music feel good or feel haunting or feel sad.
15:40
And all these things can be described as a system of
15:44
patterns and a system of mathematics.
15:46
Now, of course, it still requires somebody that's a
15:48
musician to make it all work. Or going back to our Lego
15:51
analogy, it takes a master builder to be able to discern
15:54
what it is you're going to create.
15:56
And I think that's the role of modern designers harking back to
15:59
your design background is that they're composers, they're
16:02
people that can build. But instead of having to just
16:04
like write music on a sheet in forms of notes, you can actually
16:07
like listen to the thing that you're going to create.
16:10
And that I think is like the brilliant idea of, of like the
16:12
new frontier of design. And I also think that there's
16:15
this really incredible thing that we're just on the verge of.
16:18
It's like just at the tip of our our tongue, which is the idea of
16:22
like, you like different music than me and I like different
16:25
music than my wife likes. What if, instead of having to
16:28
create like, one musical track for all of us, what if we could
16:33
subtly vary it based upon the preferences in the composition
16:37
that you have versus me? I was hoping we'd get here
16:42
because this is really where I see kind of like AI being
16:45
integral in this process because a lot of this probably could be
16:48
reproduced in some repository of like here's our brand standards.
16:52
Everyone's got their documents and guides.
16:53
But AI starts to like allow for that right where you can, where
16:58
if as a designer, I don't want to give somebody else a rhino as
17:01
this experience except for me. And then I want certain musical
17:06
attributes, you know, when I go to web websites, I, you know, I,
17:10
I could deal with like a folk song.
17:13
I don't want to see Coltrane giant steps in a product because
17:16
it's, it's too complicated. It's going to melt my brain
17:19
totally, you know? Yeah, I mean, so, so, so moving
17:21
from like the metaphorical for a second.
17:23
You know, we do this already, right?
17:26
We have things like dark mode, we have things like
17:29
accessibility controls. We have things like low motion,
17:32
which by the way, I really love. I have a mild amount of
17:35
dyslexia. So the ability to apply dyslexic
17:37
friendly fonts. All of these things are, are
17:41
things that we have today. They're just really expensive
17:45
and really hard to pull off. If you think about AI like what
17:48
we're doing is we're, we're taking this like pattern based
17:51
system or making it so that you can ship things a lot faster.
17:55
And that means that generally those things will become less
17:58
expensive as well. If all of a sudden it's easier,
18:01
faster, less expensive to make these experiences, why wouldn't
18:04
you make a tremendous amount more of them?
18:07
And that again gets back to the role of like the modern
18:09
designer, right? Like everybody is all about this
18:11
whole thing about like, like AI is gonna take all the creative
18:13
process out of the way that we build.
18:16
I, I could not believe that that's further from the truth.
18:18
I think that it is a great time to be a designer as long as
18:21
you're thinking about things in systems and you're thinking
18:24
about things in the way that you could build that you've never
18:27
been able to build before. Like if all of a sudden every
18:30
single feature could be segmented and you could have 25
18:33
different experiences that you exposed to 25 different
18:35
audiences. That is a a plane of design that
18:38
we've never had access to before, and that's really,
18:40
really exciting. Super cool.
18:43
So the advantage, you know, with knapsack is it's creating
18:46
structured data around these elements.
18:48
So that gives it just kind of like a universal place with AI
18:52
that you can add, you know, tweaks to the whole thing if you
18:54
want to. But also like you can start to
18:56
customize at some future point of like where you're you're
18:59
segmenting people by whatever attribute that it is and giving
19:02
them kind of a unique experience totally.
19:04
So if you. If you break it down, there's
19:06
really kind of three parts to the knapsack engine, right?
19:09
And so we call it the intelligent product engine.
19:12
The first part of it is, is the context, right?
19:14
And so everybody right now is all about like the context
19:17
stuff, right? MCP sort of rocked the world
19:20
roughly a year ago with Open AI adopting it in the spring.
19:25
It's kind of become like the new foreground where everybody has
19:28
MCP server. Context really really matters in
19:32
the when you think about like what you're trying to generate.
19:35
And so there's a context piece to knapsack, but the ability to
19:38
store those contexts as archetypes is a really important
19:41
aspect of it. So if you have a tool in an MCP
19:45
server that is saying I want to have a very creative process
19:48
versus a tool in MCP server that's saying I want to have a
19:50
very rigorous compliance based process.
19:52
Those are different archetypes that you can set up rules
19:55
engines for that ultimately dictate what sort of response
19:58
you're getting from a knapsack engine.
20:00
So for example, if I'm doing just a bunch of explorations, I
20:03
probably don't really care that much about a bunch of compliance
20:07
pieces if I'm just trying to get some stuff out there and
20:09
experiment with it. But if I'm doing a final review
20:11
and I need to make sure that that is WKAG AA, that my legal
20:14
review is signed off on it, that is a very specific context
20:17
associated with that aspect of the generation.
20:21
The second side of it is, is the ability to constantly be
20:24
monitoring all these different contact sources in real time.
20:26
So you want to constantly be reaching into your Figma files,
20:30
you want to constantly be reaching into your Git
20:31
repositories, and you want the system to build artifacts on
20:35
these and update those artifacts in real time.
20:37
These artifacts are things like traditional design systems or
20:40
brand books or, you know, storybook stories that basically
20:45
showcase all of the different building blocks.
20:47
It's a really great way to browse your Legos.
20:51
And then you also want those artifacts generated for the
20:53
purposes for use of AI. So AI doesn't examine things the
20:57
way that humans do. And you want to have a lot of
20:59
artifacts that are very specific to AI use cases.
21:02
And then the final thing is, is you want to be able to hit say,
21:05
like, hey, in my generation process, there's certain rules I
21:08
need to adhere to. There are certain things that I
21:11
want to make sure happen or don't happen as a part of that
21:15
generation process in order to be sure that like I don't have
21:20
questionable content or I don't have a process whereby I can
21:26
circumvent some part of the enterprise release process.
21:29
I need to make sure that this, you know, goes through
21:32
accessibility testing or whatever.
21:34
And all three of those things combined give you the ability to
21:37
set a context with an archetype, understand exactly how that
21:42
context is being generated and created through these artifacts,
21:45
and then apply to a rules engine and an LLM.
21:48
They can say I have a predictable generation result
21:51
out of it, and this enables the idea that's kind of like
21:54
floating around the zeitgeist right now of Prototype and Prune
21:58
to actually be real using the actual enterprise context of the
22:03
company I work at. So if I work at a big bank, that
22:06
website or that product that I make is going to look like a
22:08
banking application versus if I work at a music store, that
22:14
ability to create something is going to look like an app that
22:17
that music store would run. But you said, you said prototype
22:23
and prune. Yeah, that's a A.
22:25
Like I think, oh man, yeah, it's like I'm totally reaching back
22:29
into the the backlog of my brain.
22:31
Jules dumb theme that used to work at Meta, I think coined the
22:35
phrase on Medium in a post. And I think that it's a really
22:40
interesting idea of the future of design.
22:42
And then I believe it was Scott Belsky that reiterated it.
22:45
I was the former CPO of of Adobe in a podcast I was listening to
22:49
where Scott was talking about the idea of AI is a really great
22:54
proxy for for skill, but a really poor proxy for taste.
22:58
And the idea of like how we can enable creators to make real
23:02
apps more quickly and then prune from those apps is the new
23:06
modern framework for design. Yeah, I love that in that kind
23:09
of like dispel some misconceptions that were kind of
23:12
sold to us in 2025. I mean, if you think about it,
23:15
we, we kind of kicked off the year with all the big tech
23:19
overlords saying like, hey, these, these are going to
23:22
replace jobs. We're putting hiring freezes.
23:24
We're not going to, we're not going to do this with staff.
23:27
And in retrospect, you can see what they were doing being in
23:30
excuse. Well, they were looking to, to
23:34
get budget away from vendor budget and over to staff budget.
23:39
So it's not that you're spending 1000, it's that you're saving
23:42
10,000, you know, but what we've seen is actually what where AI
23:46
is excelling is it's either it's amplifying a, a, a very good
23:51
employee, or it's automating something for everybody else.
23:53
And everybody is going to have to fine tune.
23:56
And this is a shout out to my buddy Amos with Swan, you know,
23:59
is really talking about the customization and
24:01
personalization of AI tools based on people's area of genius
24:05
within whatever their domain expertise is.
24:08
So kind of considering that who who is the prime user?
24:12
Who's your customer at Knapsack? Yeah, I mean it's.
24:15
It's the person that's making product and, and we kind of view
24:17
this as the consolidation of roles.
24:19
So, you know, it doesn't have to be your traditional engineering
24:22
product design triad, that triad model has existed for two
24:26
decades or something like that, right?
24:28
I, I think that the idea of a builder is something that is a
24:32
relatively new concept. That person can come from
24:35
design, that person can come from engineering, that person
24:36
can come from product, because the AI can supplement the skills
24:40
that they lack in any one of those particular disciplines.
24:43
It just needs to be somebody with really good taste that can
24:45
really understand a user problem and solve that user problem in a
24:47
way that works well for enterprise scale.
24:50
Yeah. Product managers predominantly,
24:54
yeah, I mean, I would say that. It's probably product is is the
24:56
closest to the problem. And there's there's a huge fight
24:58
here, right? Like I'm, I'm starting a war
24:59
with this. Like, I can feel the flames, you
25:03
know, UX people are going to be like, well, I understand the
25:05
user problem the best and engineers are going to be like,
25:06
well, I understand, I solve the problem the best.
25:08
And you know, product people are going to be like, well, like
25:10
nobody actually talks to the customers but me.
25:12
So, you know, it's, it's kind of like a doesn't really matter
25:15
sort of situation. It's one of those places where
25:19
look, there's lots of really good tools for designers right
25:22
now. Things like Figma make that can
25:23
go jump start you on a, a design project.
25:26
There's tons of great stuff for engineers right now like you
25:29
know, Kodak's Cloud Code, Cursor, Reply, Lovable,
25:34
etcetera, that all help individual developer
25:37
productivity. There's stuff for project
25:38
managers like the Linear app that is doing really cool things
25:41
with how you do like work tracking and what not.
25:43
But there's very little that exists at the core of all that
25:46
that says like how do I really make a really awesome product
25:50
that exactly matches that that particular users need and does
25:54
it in a way that is systems based and data-driven.
25:57
And that's really kind of like what we're going for.
26:01
Is multiple contributors from those 3 facets of the team and
26:06
that which is brilliant because you know, just to give a shout
26:08
out to project managers, you know those guys every and gals
26:13
every product manager I've ever seen looks like they took, you
26:16
know, 12 rounds with Anthony Joshua.
26:18
Yeah, it ages you. Quickly, like the vast majority
26:20
of my my Gray hairs from when I was doing project program
26:22
management like back a couple of startups ago, like I have, I
26:27
have never like had so many empty bottles of whiskey under
26:30
my desk. And and I like it.
26:33
It is a very, very hard job because the definition of of
26:35
that role is that you're responsible for everything, but
26:38
you own nothing. And that is a very difficult
26:40
place to be in. Yeah, I mean.
26:43
You know the sales guys, you know, in the Rev OPS, they got
26:46
empty bottles of whiskey under their desk, but they're ringing
26:48
the bell. That's celebration whiskey.
26:53
Cool. Amazing What a what a great
26:55
space. Let's talk a little bit more
26:57
about the future and what's the, the capacity is particularly
27:00
around UIUX moving into kind of an AI, an AI era where all of
27:07
the the major LLMS are all getting into browsers.
27:10
So you can start to see how a lot of this, you know, UIUX and
27:14
these design decisions will be, you know, dictated and and
27:17
segmented into different groups, but also there's going to be
27:20
personal capabilities, you know, particularly around
27:23
disabilities, which is awesome. So you could set you can you can
27:26
basically tune your browser to like your preferences.
27:30
That's one area. I I see things going.
27:32
What what do you see like the end game for kind of this kind
27:36
of product in in design, moving into kind of a more AI augmented
27:40
arena. Yeah.
27:41
I. Mean so so there's a lot here
27:45
too, right, like there's there's definitely your sort of anti
27:48
patterns right where, you know, we are just a tick away from
27:53
from AI being the new ad platform.
27:56
And I think that that's that's a little bit of a scary feature.
28:00
One of the places where where AI really needs to catch up is how
28:03
does a company represent their brand in a list of AI results?
28:07
And this was Google's ticket to be in the, you know, multi
28:10
billion dollar behemoth. If they are, is, is they figured
28:14
out how in a search result you can represent a brand.
28:17
And so we, we are on the cusp of that, I think in, in AI land.
28:22
And that's going to happen, I think whether we want it or not.
28:25
And so the ability for an AI company to release some sort of
28:29
way to have a not wholly owned control of the representation of
28:35
your brand and an AI generator result is going to be something
28:38
that is very soon. I also think that there's a lot
28:42
of interaction patterns that are still very much in their
28:44
infancy. You know, yield multi line text
28:47
box is still how most people interact with AI and I think
28:51
that that's not the richest experience in the world.
28:57
I think they there's this really cool thing called AI verse dot
29:00
design that I've been playing around with lately.
29:02
I don't know if you've seen it. It has a bunch of different
29:05
ideas about how we could design better feature sets for AI
29:09
tools. And I think that there is a lot
29:11
of good design patterns there that are are interesting.
29:14
And I think a lot of the interesting patterns in design
29:16
for AI are also a lot of off screen patterns.
29:20
I think that one of the things that I've moved to in 2025, I
29:25
did it fairly early this year. I pretty much exclusively voice
29:28
type e-mail now, largely because of AI voice typing.
29:31
And that has revolutionized my productivity in terms of the
29:35
ability to to sling text across the Internet.
29:41
I think that likewise, there's a bunch of things that we're doing
29:44
that are representative of visual mediums with photos and
29:48
imagery that are really, really interesting.
29:49
All the diffusion model stuff is getting really, really good.
29:53
And those are a bunch of experiences that I think are
29:55
going to revolutionize the way that we think about digital
29:57
products as well, because you're not going to have to design the
30:00
visuals of every single interface fairly soon.
30:05
So I mean. Those are a couple of like big
30:07
examples. I think that when I think about
30:10
the evolution of this right is I also see a lot of AI going
30:13
invisible. I think that there is a bunch of
30:16
this right now, but you know, as, as you get a bunch of models
30:21
that are able to learn and adapt better, as you get a bunch of
30:24
things that are like, you know, glorified context servers, but
30:26
are doing more than just providing A context, you can
30:31
start to see the impact of AI much more, you know, behind the
30:35
scenes where like stuff is just going to get better.
30:38
And I think that that's that's interesting.
30:41
And I I think that there is there is like.
30:44
There's a lot. To unpack there specifically
30:46
that I don't want to like go on and on and on about, but there
30:50
is a difference between really, really good context improving
30:53
our generative experiences and slop.
30:56
And I think that like, as we get better at designing these
30:59
context systems, you're going to watch a lot of that slop fade
31:01
into the background or very quickly be displaced by stuff
31:03
that's actually helpful and useful.
31:06
Yeah, it just feels like the text box is just too passive.
31:10
I mean, like like your sign says, build using patterns.
31:13
AI is really, really good at patterns, particularly with me
31:16
because I'm very predictable. I only do a handful of things.
31:20
And kind of like the, the predictive use of AI is what's
31:23
most interesting to me of like understanding what I'm going to
31:26
want when I'm going to want it in front of where I actually am.
31:29
Cognitive of that, you know so for me I'm really like just
31:33
interested in like kind of biofeedback stuff of like how
31:36
different biomarkers that can actually read you know your
31:40
thoughts, so to speak. I don't want to go full Elon
31:42
neural link. Maybe you know, but there's a
31:46
there's a a a lot of steps in front of like implanting
31:49
something that can read your mind, but but ultimately we.
31:52
All we all, we all wear sensors now, right?
31:55
And so because I'm, I'm holding up my watch and my ring, and so
31:59
like those sensors can showcase heart rate variability.
32:03
I can understand my skin temperature.
32:05
I can understand all these different things that are
32:07
experiential about the way I experience my environment.
32:11
And we can start to use these to create better experiences.
32:15
Something that's thrilling and exciting should increase your
32:17
heart rate. And like, look, 'cause some of
32:19
this gets a little dystopian and like, we have to be careful,
32:21
right? There's a little spoonful or 7
32:24
of caution in here, but the idea of of a multi line text field
32:31
being displaced by other things that are more rich is great.
32:35
I mean, a picture is worth 1000 words, right?
32:37
And it's funny, I was actually sitting at a bar with Brad Frost
32:42
in Pittsburgh. Brad Frost wrote Atomic design,
32:45
big design systems guy. And one of the things that that
32:48
he said to me while we were sitting there having a drink
32:50
together was, you know, what I want for 2026 is I want to be
32:55
able to to take a cocktail napkin like the one under my
32:57
drink. And he pulled out a pen.
32:58
He's like, I just drew an app idea in 30 seconds.
33:01
I want to hold that up to a camera and use a design system
33:04
and AI to make that app real in minutes.
33:08
And that's way better than any way I could describe that app
33:11
with words. And so that is something that
33:13
like I really hope at the end of this year, I'm able to show Brad
33:17
that. Yeah, we're close, right?
33:20
So close. Yeah.
33:23
Cool, you know, and it's that's a good bridge too.
33:25
I want to talk about your podcast
33:27
designsystemspodcast.com. We just rebranded to Patterns
33:31
Podcast. Yeah.
33:34
We we've decided to to elevate a bit out of the design systems
33:38
landscape largely because as the industry's evolving, as the the
33:43
landscape is changing, it's a lot more about how we think
33:45
about patterns and systems than it is about design systems
33:48
themselves. Love that.
33:50
That's very cool. And and this is also for
33:53
builders, right? So your, your products for
33:55
builders, your podcast for builders.
33:57
What is it about builders that that makes you so interested?
34:00
I mean I was a builder. My Co founder was a builder.
34:03
We spent 20 years of our lives in the, the code mines and in
34:08
the, the fields of project management, you know, slaving
34:11
away to build, to build the experiences you all know and
34:13
love. It's a very dramatic way of
34:15
saying like we, we experience the pain that we're trying to
34:17
solve. You know, we, we have gone
34:21
through release cycles that take six months plus or a year to be
34:25
able to get a product that was designed in, you know, 2020 to
34:30
market in 2022. And the idea of the time it
34:35
takes, the effort it takes the, the gauntlet you run to build a
34:40
product, like I just want to make that easier.
34:43
I want to be able to take that person that has that really
34:45
great idea and I want to empower them to make it, and that's what
34:50
this company's really, at its core all about.
34:53
Oh, cool. I'm really fixated on this word
34:55
patterns. You got me thinking now, you
34:57
know, and it reminds me of when I was a kid in, in elementary
35:00
school, public school in Denver, Co, And it was all about
35:04
testing. They had like little IQ tests
35:06
and things like that that you had to take and that determined
35:08
your, your track and like programs and whatever.
35:12
But I remember I didn't perform as well as I could, but like all
35:15
of the, the testing was pattern recognition.
35:18
So it kind of gives you like red, blue, red, blue.
35:20
What's next? You like blue?
35:21
OK. And then it gets pretty
35:22
advanced. And I remember thinking like, as
35:25
I was looking at it, I was like, I can't, I'm not recognizing the
35:29
pattern here. But like literally as a kid, I
35:31
felt like if I could move it, like if I could change that, my
35:34
orientation to whatever they are giving me, like I'd be able to
35:38
see it. I, I, I truly, which which kind
35:40
of like, I think evolved into like thinking about everything,
35:43
a pattern and maybe falling into the Fibonacci sequence in some
35:47
kind of way. Golden spiral situation.
35:50
Yeah, some golden. Ratio.
35:53
But it does feel like everything has a pattern depending on if
35:57
you're available in a position to to recognize that.
36:00
Would you, would you fall into that camp?
36:02
Yeah, I mean, I, I think generally.
36:03
That's that's true. I mean, I look, I'm a huge
36:07
systems guy, right? I think that systems thinking
36:09
and patterns are kind of like what govern my life and how I
36:12
see the world. I think that there are a lot of
36:15
people that look at things through the lens of snowflakes.
36:18
And, and what's funny is that snowflakes are also patterns.
36:21
But the idea of like, how do you view the world is actually a
36:25
really interesting concept about whether or not you get it right.
36:30
There is systems thinking and then there's, there's like other
36:33
parts of the human psyche that are somewhat difficult to
36:35
understand, right. It's like when people are bad at
36:37
exponential math, you know, there's, there's lots of folks
36:42
that can get it and just understand it fundamentally.
36:45
And then there's a bunch of folks that just really struggle
36:48
to make that connection between the thing I'm seeing and my use
36:54
of that thing systematically. And there is like a, a Spider
36:58
man, Uncle Ben thing here, right?
37:00
Like if you're using systems to build your product and those
37:03
systems in those patterns govern the way that your product gets
37:06
built, that's a lot of power and that's a lot of responsibility.
37:10
And the governance and the control of that is something
37:12
that becomes really paramount in that process.
37:14
You can't just have anyone change your core brand color and
37:18
have that roll out to every single product in your your
37:20
product ecosystem. That is a major change that has
37:23
lots of ripple effects. And you have to be really
37:25
mindful of that pattern system and how that impacts things.
37:31
And so I think that that's the, the place that a lot of people
37:34
find themselves is they want the power, but they also don't
37:38
totally understand the, the mindful nature that you sort of
37:41
have to have about systems thinking in a similar way to, to
37:46
like not understanding how, how, you know, exponential math
37:49
causes this, you know, insane conditions in, in lots of
37:54
different environments in the world.
37:56
Like there's a, a part of human nature that is really
38:00
challenging here. And I think that's where a lot
38:02
of people face struggles in the land of AI is they think about
38:06
things as like, I am designing this single experience instead
38:09
of I'm designing the system of experiences.
38:12
And people need to start thinking about how do I design
38:15
systems of experiences, not how do I design an individual
38:17
experience. What's the?
38:19
Bridge what's what's kind of like the the limiting factor
38:23
that people are missing when they're thinking about
38:25
individual the systems, you know, I think a lot of it is is
38:29
that like people get. Into habits like there's not
38:31
really great training for this. It's funny, honestly, this
38:33
conversation came up over the holidays.
38:35
I was talking with, you know, a few my friends that are in in
38:38
the the realm of healthcare and they were talking about this
38:43
whole thing about like, you know, all right, so if you're a
38:45
tech worker right now and you're out there and you're just
38:47
starting your career, like what are you doing?
38:50
And I think that people that are grads last year, this year,
38:55
probably next year are in a Crucible.
38:59
And the reason why they're in a Crucible is like they were not
39:02
taught in college a lot of the AI skills necessary to
39:05
understand how systems work and how systems impact and influence
39:08
things. And they're also entering a
39:10
market where there's this massive change going on right
39:12
now. And the people that are really
39:14
good systems designers and really good systems thinkers
39:16
when it comes to code design, product, they're all pretty
39:19
senior in their career because they've seen it and they
39:22
understand how the systems work. They're not just like, I'm
39:24
sitting down to build this one experience, they're sitting down
39:26
to build the next N number of experiences.
39:30
And so you're in a really tough spot if you're in that place.
39:33
And so I would, I would say like the easiest way to get on board
39:36
is, is going to be spend a bunch of time understanding how you
39:40
design in a place that is like a system wide abstraction versus a
39:45
specific instance of an application or screen.
39:48
And that's hard because there's not a lot of opportunities to do
39:52
that when you're early in your career.
39:54
And I think that's going to change.
39:55
You're going to watch design programs, code programs,
39:58
collegiate programs, get on board with this idea because I
40:02
don't see it going any other way.
40:03
Like we are going to go away from like this particular screen
40:07
at this particular Uri is going to show this particular thing
40:11
pretty soon. And we're just going to be
40:13
designing for the systems that go and build those experiences
40:16
in AAI enabled way. Yeah.
40:20
I love it. AI becomes almost like a
40:22
companion, you know, You know, I watched Avatar one last night
40:28
getting ready for Avatar three. I'm still I've heard it's
40:31
awesome. But you know, the way they they
40:34
like tie their braid to like the dragon and like to the, the the
40:38
lizard horse and everything. I was seeing kind of a
40:41
similarity there, you know, where there's like a union, but
40:45
each component still maintains a certain autonomy, which is I
40:48
don't know. Yeah, couple of the the bio USB.
40:53
Of, you know, that film I it's, it's an apartment.
40:58
It's an apartment metaphor, right?
40:59
Like you think about the idea of like, how do you, how do you
41:02
think about the interfaces that we all have, right?
41:04
Like those interfaces are all pretty wrote.
41:07
There's not a lot of like holy new interfaces that get created
41:10
all that often and for good reason, because like there's a
41:14
limit to human perception and, and maybe we don't need like a,
41:18
another interface, but there is this idea of like every single
41:22
one of those interfaces needs to be adaptable to a particular
41:24
market or particular use case. Database is my favorite thing
41:27
for this. I actually was super nerdy about
41:29
database for a really long time. And one of the things that I
41:34
always thought was really interesting about database is
41:36
like, look it in the context of banking, right?
41:37
If you look at it from like a retail banking app, it's like,
41:40
hey, I want to see my bank balance and like, you know,
41:42
hopefully it's going up for most of it's going down and it's like
41:46
one line on a chart month by month, or it's a bar graph or
41:49
something like that, right? But if you're somebody that's in
41:51
markets, you're watching like 30, sixty, 100 lines overlaid
41:56
with one another. And all of a sudden it becomes
41:59
the ability to like, how do I pick out certain aspects of that
42:02
data from that chart? And it's very different.
42:04
Even though it's still a line chart or a bar graph, it is a
42:07
very different way of interacting with that data that
42:10
has a wholly different purpose and a wholly different goal
42:13
associated with it. And it's the same control, but
42:16
in very different context for consumption.
42:19
And that's the thing that we all need to get, like way better at
42:22
is understanding that like, yes, we're gonna have a line, a line
42:24
graph, but those line graphs are going to all of a sudden have
42:28
lots of different purposes and lots of different goals and
42:31
intents behind them. And we need to be able to design
42:33
systems that can handle all of those different use cases.
42:36
Love that. I love that.
42:38
Do you mind if we go over some business methods?
42:42
Cool. I saw something in 2024, you're
42:45
at about a five and a half million ARR.
42:49
How did it with about 30 employees, how did 2025 round
42:52
out? I mean, it was a good year.
42:54
Like I, I'm not going. To talk about specific ARR
42:57
numbers right now, but you know we had, we had a nice up here,
43:01
it was really strong. We especially grew in our
43:03
enterprise traction, which is where we were really looking to
43:07
to expand. And so doing a lot more deals in
43:10
that like you know, 2 to 500 K range that have been really
43:15
exciting to propel our growth. And also knocking down some,
43:19
some great big logos again in banking, finance, healthcare,
43:23
we're about to get our first airline, which is really
43:25
exciting. And so being able to to work
43:28
inside of that large enterprise space was a big 2025 turning
43:33
point for us. And that's been been mainly the
43:37
focus and it's, it's gone. Well, yeah man, your trust bar
43:41
is. Bananas.
43:42
You know, T Rowe Price, Edward Jones, AP Yeah.
43:46
And Bill. Incredible.
43:48
Yeah, we've been getting some some great folks.
43:49
And I mean more to come, like I, I can't talk about stuff that's
43:53
in flight right now, but you know, getting one of the world's
43:57
biggest software companies on board here pretty soon, which is
43:59
going to be really exciting. The rhyme.
44:02
With Microsoft, I can't say I'm not.
44:06
I'm not going to mess up that. Deal at the 11th hour.
44:09
Let's just put it that way. Exactly.
44:12
In other news, Ryan Estes blows a giant deal.
44:17
I'd love to talk about pricing and how pricing is set for
44:20
particularly for AI tools because there's a lot of like
44:23
there. There's a lot of nuance and and
44:25
new pricing models. When I go to your your plans
44:29
page, there's no publicly listed pricing.
44:32
You kind of show three different tiered, but there's no available
44:36
pricing. Why is that?
44:37
Yeah. So we actually pivoted a lot of.
44:39
The model to be better suited for enterprise this year and a
44:42
lot of that was moving towards for deployed models.
44:44
I mean for deployed is like the way you do it, right?
44:49
And a big part of the reason why is because big enterprise still
44:52
doesn't understand how to implement AI.
44:55
You know, I, I like to joke that there's a bunch of, of like
44:57
senior C-Suite executives that they're all at like this pool
45:00
party, right? And they're all standing around
45:03
the pool like talking, having drink, whatever.
45:05
And the water is really cold. And every once in a while, like,
45:08
like they're dipping their toe in.
45:10
And, and that pool is obviously like the metaphor for AI.
45:13
It's only going to take like a, a push to really get the first
45:17
person in that pool. And then a lot of them are going
45:18
to jump in. And that's kind of what we're
45:20
trying to do is we're kind of trying to be that push.
45:22
And that push is using things like for deployed engineering
45:25
and professional services to make it so that your solution's
45:28
successful. One of the things that we've
45:30
been, you know, struggling with is you go and you, you sell a
45:33
big deal like a, a mid 6 figure deal to to a customer.
45:37
And that's like just software licensing, right?
45:41
And then what happens is they get four months in and
45:44
implementation stalls. And the reason why
45:46
implementation stalls is because not that there isn't like an
45:49
organizational will to get it done, but people just don't
45:51
fundamentally know how to use the product in a way that gets
45:54
them to value. And so pairing that with with
45:58
strong CX motion and strong for deployed engineering, let's just
46:02
go further, faster. And that's really kind of what
46:04
it's about for us is like how do you get your customer to value
46:07
as quickly as possible? We definitely had pain points
46:10
where we would have certain customers that were well
46:13
equipped that would get to value in like 60 days.
46:15
And then we had others that would take 6 plus months.
46:17
And trying to figure out how to make everybody 60 days or less
46:21
has been a big effort on that. And there's a cost associated
46:24
with that that is difficult to put on a website.
46:27
And so we kept having people that would come to us and be
46:30
like, hey, you know, sales on your website at your entry point
46:32
for your software is $85,000. It's like, yes.
46:36
And that also requires a significant upfront for deployed
46:40
engineering cost that they weren't necessarily ready or
46:43
expecting to pay. And that was hampering our sales
46:45
cycle. I think that you know, one of
46:48
the other things that's really complicated about this right is
46:50
consumption pricing. And so we are going to be moving
46:53
to consumption pricing in March. We're going to give customers an
46:56
option for a while. Consumption pricing is really
46:59
interesting because it completely changes a lot of your
47:02
fundamental business metrics. Now all of a sudden that speed
47:05
of adoption is the only thing that we care about because if
47:08
people are using it based on consumption, you're not just
47:10
locking them into a seat based model.
47:12
They actually have to use the software.
47:14
And it's not that our usage is bad per SE, but we don't really
47:17
know today exactly what sort of consumption per seat looks like.
47:23
And so there's gonna be some experiments that we're gonna
47:25
have to do. And we're actually running those
47:26
right now to understand more about how consumption pricing is
47:29
ultimately gonna impact things. And we're leaning our partners
47:31
and friends at different companies that do this sort of
47:33
stuff to understand more about how they're doing it.
47:36
And I think a lot of the other piece of this too is, is a lot
47:39
of AI is moving to marketplaces. And so you have big providers
47:43
like, you know, Google, Open AI, Microsoft, Amazon, etcetera,
47:47
that have really strong incentives to partner with them
47:49
in marketplace spaces. And those marketplace spaces
47:52
become this channel for AI consumption sales that is
47:56
really, really powerful. And so establishing those
47:58
partnerships, solidifying them, getting them out there, that
48:02
becomes a key driver of growth for us just so that it makes it
48:05
frankly easier to buy us. Yeah.
48:09
Well, in four deploy, I mean it, it does come down to ultimately
48:12
always is going to be change management.
48:15
You know, if you in particularly you're working with large
48:17
enterprises, which means you have this grizzled, you know, a
48:21
senior leadership board that's like, dude, I don't want a new
48:25
tool. You know, you there, there is a
48:27
coercive and inviting process to that.
48:30
Like what's what's the secret to, to winning over those kind
48:32
of key stakeholders that are going to basically deploy
48:35
everything to the rest of the team?
48:36
And you just said, like, what's the secret to enterprise?
48:38
Sales, you know, if you, if you find it consolidated, let me
48:42
know. I forget his last name, but
48:45
Chris that worked at Snowflake as their CRO wrote this really
48:49
amazing book about this called Let it Snow that talks a lot
48:52
about the enterprise sales cycle and the enterprise sales
48:55
process. And I had the pleasure of
48:56
listening to him speak in October.
48:58
And one of the things that that he talked about that I thought
49:01
was very insightful was involving your sales process and
49:04
your engineering process. I'm making it so that that
49:07
engineering product and, and sales are tied at the hip.
49:13
And that communication loop has to be really, really strong.
49:15
Because in, in an enterprise context, one of the things
49:18
that's really scary is that like these buying teams are like 20
49:20
plus people and you don't have to win over everybody, but
49:24
everybody has veto power. And so when you think about
49:27
that, like how you overcome an objection, how you overcome A
49:29
veto becomes the name of the game.
49:32
And that can be a really well timed piece of content that can
49:35
be a particularly apartment feature that you're willing to
49:39
prioritize in your road map. That can be something that can
49:44
change the game in terms of being able to win that customer
49:47
over. And a lot of this is about
49:49
people still buying into to vision, but also buying into
49:54
companies that they can match their risk mitigation goals.
49:59
And so risk mitigation comes in all kinds of forms, right?
50:02
Like case studies, sock 2 compliance, the ability to make
50:06
it through of 16 committee security process, which by the
50:09
way, we did in 2025. The idea of how you actually get
50:16
people to buy into a vision and help them propel you over the
50:20
myriad of hurdles that exist inside of an enterprise sale.
50:24
That's the key unlock. And there's frankly not that
50:27
many people that are really, really good at that.
50:29
And that tends to be like my superpower is understanding how
50:34
that process works, being incredibly resilient during that
50:37
process and then being able to pick people that are just like
50:40
me, they can keep their keep keep being dogged about getting
50:44
through that, that massive set of hurdles in enterprise.
50:51
Love it. We're about 3 minutes from the
50:53
top of the hour. Do you have a couple extra
50:54
minutes afterwards or? Yeah, let me let me shoot a
50:56
quick text to. The next caller, Yeah, please.
51:02
Cool. In that case, what we'll do is
51:04
we'll pivot to the quantum round and we'll go quickly.
51:10
I don't want to make I want to make sure that you're not too
51:12
late for your next appointment. And also I, I know you're not
51:14
feeling 100%, so I really appreciate you coming on.
51:17
And yeah, and I'm honestly, I've been, I've.
51:19
Been able to not totally cough my way through this, which has
51:22
been great. I've had like the worst case of
51:24
flu since I got COVID in like 2020.
51:26
And so I've been, I've been down for weeks.
51:29
Like I got to say, this has been this has been a tough one.
51:32
Oh, brutal, man. Yeah.
51:34
All right. We're.
51:35
Entering the quantum round round, I feel like there should.
51:37
Be like a screen effect or something.
51:40
We can make that happen. Cool.
51:42
What LLMS you are you using and in what capacity?
51:46
So I run DeepSeek locally. Just because it's very
51:49
convenient with like O llama and stuff like that.
51:51
I mostly use that for searching through TTRBG rule books very,
51:55
very quickly. And then I also I use a
52:00
combination of cloud code and cursor in my kind of like
52:05
day-to-day workflow because I voice type in Windows and in
52:11
Mac, I have a couple of different tools that I use for
52:13
voice typing, actually, the Windows native voice typing with
52:16
Copilot, it's actually pretty good.
52:19
And then let's see, I use a bunch of stuff with 3D printing
52:23
actually. And so using meshy along with 3D
52:27
scanners is really cool to be able to, to make 3D sculpts
52:32
quickly and easily. I am frankly terrible at using
52:35
3D sculpting programs. And so if I don't want to pay a
52:37
Commission and I don't own a $5000 ladder scanner.
52:44
So using those are, are really, really great to be able to, to
52:47
make that happen. I've been playing with the the
52:50
Google Photos remix thing a lot. We have a, a joke in our family
52:55
with a traveling pickle. It's it's a pickle that sings.
52:59
It's really annoying and terrible.
53:00
And it's a gag gift that shows up in random places in people's
53:02
homes. And so being able to superimpose
53:05
that pickle on various places in people's houses to make them
53:09
look for it has been hilarious. And so like, that's, that's all
53:13
like tinkery stuff, right? I think that that like in the
53:16
day-to-day side of things, I do a ton with Claude and I use that
53:21
like pretty exclusively for knapsack creation.
53:27
And then I also use a fair amount of perplexity to do like
53:30
a lot of answer seeking around, especially things in business.
53:33
Like it's really great to do synthetic research with and the
53:36
ability to do that synthetic research to help me make better
53:39
informed business decisions is really cool too.
53:42
Amazing. If you were to enter college as
53:46
a freshman, knowing what you know now, what would you study?
53:51
You know, it's it's really interesting you.
53:53
Bring this up because I think I think computer engineering and
53:55
then ended up in with a business degree and through a very
53:58
tangled process. I think if I did it over right
54:00
now, I would probably do biology, largely because I think
54:04
that like biohacking is going to end up being this huge forefront
54:11
as we get closer to like things like sensors integrating more
54:14
with our, our day-to-day lives. Plus, we're just really
54:16
fascinated by the subject. I also really love linguistics.
54:20
Linguistics is a huge passion of mine.
54:22
I think that it's really interesting to be able to
54:24
dissect the origins of language and how we communicate with one
54:27
another. Again, systems thinker.
54:29
So thinking about like the the systems of language and glyphs
54:32
and the construction of how we speak to one another is really,
54:34
really cool to me. There's not a lot of careers in
54:37
linguistics unless you are in AI, and so I am, Yeah, it'd be
54:43
one of those two things, I think.
54:46
I I love linguistics too. You ever read that book by
54:48
Jane's? It's the origin of consciousness
54:51
in the bicameral mode. So good.
54:52
Yeah. So good, absolutely, so good.
54:56
You know, it's 'cause it's kind of thing.
54:58
You can't really prove any of this stuff, but boy man the dots
55:01
look like they connect. Really gets you thinking.
55:04
I I think that. There is a lot of really cool
55:06
stuff. Yeah, totally.
55:09
I could go on forever. Right on.
55:12
Cool. Is there ACEO that you follow
55:15
closely or admire? Yeah, I think what?
55:17
Mellie Hicks at Canada. Has done and continues to do is
55:20
is pretty, pretty darn remarkable.
55:23
You know, Canva was something that took a little while and her
55:28
perseverance, her leadership, her ideas of the future where
55:33
the creative process is democratized.
55:36
I used to run a creative services team.
55:37
It was, I'll be at very small, I think it was three people.
55:40
But being able to see first hand how Canva changed the way that
55:43
people think about making assets inside of their organizations is
55:47
really remarkable. Yeah.
55:50
It seems like the whole model just kind of was the origin was
55:53
like Adobe is too annoying to use and too confusing in the
55:56
same way that like Google Sheets was just like, dude, Excel is
55:59
too much. It's like the simplification
56:01
kind of like what you've been talking about design thinking,
56:04
you know is there's like it creates massive business
56:07
opportunities and canvas AI product is actually pretty good
56:10
too. I, so I make a lot of TTRPG
56:12
assets. It's obviously a theme in this
56:15
podcast. I made a fake newspaper on Canva
56:18
and I had everything written that I wanted in the articles.
56:20
So I was like format this like an old school newspaper and it
56:24
did a fantastic job making a really simple like asset handout
56:28
that I could I could use for a game that before I would have
56:32
had to Commission or I would have never been able to do
56:33
myself. Well, I'd love to give shout
56:37
outs to charitable organizations.
56:39
Is there a charitable organization that's near and
56:41
dear to your home? The Wikimedia Foundation
56:42
forevermore. I think that Wikipedia is one of
56:45
the greatest assets and repositories of human knowledge
56:47
is that has ever been created and we would all be poor without
56:51
it. It's one of the few things that
56:53
I use every single day. I also, because I live in the
56:56
Pacific Northwest, protect our winners is a big thing for me
56:59
and my family. You know, it's it's
57:02
simultaneously climate based organization as well as one that
57:06
looks at things like glaciology, snowpack, skiing.
57:10
It's a big, huge passion of mine and my family's and they're a
57:13
great organization. Audible.
57:17
Any opinions about Grokopedia? Oh, I mean no.
57:26
Comment I feel like I I. I don't know.
57:29
I like I'm. Especially given the recent.
57:30
Changes I I heard some rumblings about like the, the, you know,
57:34
acquisition side of things this past week.
57:37
I'm not educated enough on the topic to to dive in.
57:39
Really amazing. Chris, thank you so much for
57:43
joining us. And going a little bit over
57:45
knapsack.cloud is the product. What's the best place for people
57:49
to connect with you personally online?
57:50
Yeah, track me down on LinkedIn. LinkedIn.
57:53
Slash in slash Chris Straw. You know, I don't really I don't
57:56
really Twitter anymore or X anymore and I haven't really
58:01
been able to embrace like the short form content the way that
58:03
I was, I was hoping to. So it's really LinkedIn is worth
58:07
finding tremendous. Chris, this has been amazing.
58:10
Thanks. Thank you so much for joining
58:12
me. Yeah, absolutely.
58:13
Hey, thanks so much for having me on.
58:14
It's. Been great to chat.
58:16
Hey founders. Let me ask you this, are you
58:19
building with AI or getting left behind by it?
58:23
If you're serious about scaling smarter, faster and bolder, you
58:26
need to check out AI for founders.co.
58:30
It's your go to resource for cutting through the noise and
58:33
actually using AI to grow your start up.
58:36
Each week, AI for Founders drops a sharp, no fluff newsletter,
58:41
dives deep with expert guests on the podcast, and hosts monthly
58:45
workshops where you can get your hands dirty with the real tools
58:49
and strategies. Whether you're precede or Series
58:52
B and beyond, launching something new or just staying
58:56
sharp, AI for Founders gives you the edge you need to win in 2025
59:00
and beyond. Hit AI for founders.co to
59:04
subscribe to the newsletter, tune into the podcast, and grab
59:08
your spot in the next workshop, and it's all free.
59:11
Build smarter, move faster, think bolder with AI for
59:15
Founders. Now back to the show.