Nathan Curtis on AI, Design Systems Architecture, and Staying Aligned Beyond Figma

Brad: All right.

Mr. Nathan Curtis, hello.

Nathan: hello.

Brad.

How are you?

Brad: I'm doing great.

Uh, let's dig into it.

What has you waking up excited these days?

Nathan: how my professional life
has changed over, like, the last

two years, just the way that our
whole field is doing design systems

different, and I feel really
empowered to build things a lot more.

And so I've been building stuff

Brad: Wow.

Nathan: architecting stuff more.

I sort of drifted from the team leader,
driving design system success, like

organizing the whole mission into,
like, strict architecture and building

tools, and it's been a lot of fun to,
continue to live on the edge between

designer and engineer in our space.

Yeah.

Brad: you just went through a
metamorphosis, uh, of such where

you've been in the weeds working
in a certain capacity for a really

long time And you're taking that,
all of that hard-earned experience

and now moving it to a different
place or wielding that or pointing

it in a, in a different direction,

Nathan: the fact that our tools are in
a place and our technology, particularly

with AI, uh, is in a place to, to build
and realize ideas so much faster and

connect dots and make things so much
faster has me really invigorated, to,

Brad: That's wild to hear from you.

That's wild to hear from you.

cause I feel like, well, we need to,
we need to go here, but it's, it's,

it's empowering, but also as people
who really care about, you know,

quality and consistency and, like,
all of, like, that stuff, you know

what a giant mess it can make as well.

So, so we should go there.

But I, I love to hear that, that you're
excited about this, that you feel

empowered to move into a totally new role,
Like, are, are you nervous about that?

Or do you feel like even though
it's, like, a new chapter…

Nathan: I don't think so.

There's a number of things, like
my world is, particularly the last

three years, is building tools to
communicate design intent through to

engineering to make component libraries.

And so, like the specs tools, as it were,
um, really have taken a different shape,

and the kinds of customers and consumers
of the data have taken a different shape.

But I'm… the confidence thing
is it's still riffing on a lot of

the patterns and building blocks
that I already had in my head.

and at times it's forcing me to be
more deliberate and explicit, and

often through that journey, exploring
where I actually understand it

quite as well as I thought I did.

and so that's humbling, but

Brad: Yeah.

Yeah.

Nathan: And, but the other thing is
it's enabling me to really spring or

break through borders that I have never,
like, thought to spring through because

capability-wise or tool-wise or just
time-wise, I didn't, I couldn't put those

Brad: Yeah.

Nathan: for us.

And that, that is really exciting to me.

Brad: I love to hear it.

That's so cool.

It, it sounds like you're, you feel
like you could do more effective work

at the right layer because you're
not, having to repeat yourself.

It's, it's like you are, you are in m-
this is design sy- you're describing

design systems at a career level

Nathan: mean, even there
are places to grow.

Like I'm, I've been in
this career a while.

I've been doing design systems
what, twenty, twenty-five years.

The, the

Brad: Well-

Nathan: design systems for that

Brad: What year did your… For the
good people who don't who don't go

all the way back, like what year was
your book on component-driven design?

Nathan: Yeah, like my
book, "Modular Web design?

uh, was published in 2009,

Brad: 2009, everybody.

2009.

Nathan: So it, and it included two parts,
and it's, it still mirrors who I am today.

The first half of the book was about how
to decompose design into smaller parts.

I didn't call it atomic design, but it
was effectively the same i- kind of ideas.

And the second half of the book was
like, okay, you've got a library.

How do you run it?

How do you govern it?

How do you document it?

How do you, like, deal with, Like…

these features and, and decompose
things that you're reusing?

And so that was 2009.

I, my first formal design system I
worked on was 2006 at Sun, uh, sun.com,

and so that's where my formal journey
from being an information architect

just making experiences into really
working in design systems happened.

Brad: It's incredible, and
you are here in the year 2026

saying, "I feel invigorated."

Nathan: Yeah, like, I mean,
one axis that you could look at

it as, like, I feel like I've

Brad: Yeah.

Nathan: visual styling, UI structure,
props and, uh, configuration and

so on, but I'm still challenged
by concepts and composition,

Brad: That's beautiful.

Nathan: towards that.

And another thing is, like, I'm now
working on how to incorporate motion

into design systems more, and that's
never a challenge that I've tackled.

And so there are concepts in motion
that are completely foreign to

me, but I'm learning an entire new
space, and that's a lot of fun.

And, and I think that mirrors, like,
what AI unlocks for people, because

now we can scale and differentiate and
explore territory in design systems

that, frankly, 97% of design systems
don't do motion, barely if at all.

Like, they

Brad: Mm-hmm.

Nathan: maybe their button scales when
you press on it, like, kind of thing.

But, like, as you're thinking about the
choreography of different pieces and

you're thinking about different kinds
of springs and what they mean, and,

like, that's all new to me, and I think
that's new to a lot of people in systems.

And, uh, being able to use AI is now
starting to unlock a lot of those doors.

Brad: that's one of the things that
I, I wanna get at this with you.

Have you been hearing people, design
system teams specifically, feeling

threatened by this moment and like, uh,
uh, "What's this, what's this doing?"

Or, "Wait, no, this is taking away…
I'm, I'm the one that, that makes the

buttons and sweats those details."

And like I think what, what you're
describing is What can happen when

you're free from that just kind of…
That's a means to an end, right?

That's, that's… I, I usually talk about
this through the lens of respect, right?

It's like, what's, like, a respectful use
of someone's time, talents, and abilities?

Like, is it the, the sweating
the border radius values?

Like, obviously, like, you need that
to be sound, whatever that value is.

But it's like, is that the best
thing to be having to do again and

again, or, or sort of dwelling there?

Or are there areas like motion or just
the, frankly, the next generation of UX,

for instance, is, like, there's all of
these places that is maybe a more fertile

and more respectful use of people's time
than just kinda having them compare…

Holding up design, holding up code,
and saying, "Are these things the

same?" Which is what we have been
doing for, like, a freaking decade.

Yep.

Nathan: first of all, I'll say it would
be naive for people in this moment to

not feel a threat uh, there are companies
that are undergoing massive amounts

of change and are making a lot of,
triggering a lot of that change based

on assumptions, instinct, guessing.

and so, you know, you have these mass
layoffs, you have the, uh, teams morphing

how they're doing their work, and
it's because we're all in this space

where we're just trying to understand
the lay of the land, and some of

Brad: Yep.

Nathan: layoffs probably are premature.

I think there are a lot of teams that
are, uh, let's say swinging too far

into adopting all the new tools and
not understanding what they're leaving

behind from the tools they have that
still do what they need to do better.

Brad: Yeah.

Nathan: so when you talk about craft
and, uh, sweating the details of a

border radius, like I do see that teams
that are already mature, most of the

teams I tend to work with happen to have
larger enterprise level design systems.

I'm biased towards design systems
that are multi-platform as

opposed to just single platform.

So like, all the stuff I
say carries certain biases.

they already have mastery of a lot of
those things, m- they're still undergoing

change on what those things are.

They're still rethinking, "Well, how
should we, prescribe an intent via the

tone of the button in certain visual
ways and using characteristics now?"

Are they using shadows more?

Are they using sophisticated
gradients as outlines of

borders, like borders to buttons?

Like, yes, they're doing
more sophisticated things

Brad: Yeah.

Nathan: the tooling's
enabling them to do that.

Brad: Yep.

Nathan: the sophistication of
expression is increasing, but the need

to decide things, the need to record
things and implement those things and

verify that you've implemented those
things, those things haven't changed.

Um, and we're still, I think, y- years
away from all of that stuff just going

from a of design intent by someone who
doesn't understand systems, and then

they get what they want on the other
side, including all the systematic

choices and architecture underneath.

Like I, I, I don't think anybody should be
threatened by that in the immediate term.

And if you're removing systems because
you wanna make those choices, that's

not a change because of AI, that's
a rejection of system thinking.

Brad: Yeah.

Yeah.

Yeah.

Yeah, and I think you're gonna find a lot
of people finding out the hard way as they

kind of go down these roads and eventually
go, "Man, uh, we're out here kicking up a

bunch of dust. We have a bunch of similar
but different things, and who the hell

made this one that's floating out here
and getting us sued or, or whatever?"

And I think that, that you're right.

It's like, it's like so long as the
concept of a company or an organization

where you're trying to organize more than
one person to go, "We're gonna go here

and not here." Like, so long as there's
still groups of people who are operating

as a group, there's going to need to be
the ability to have agreement, to make

decisions, like to, to share things.

We do this, and we don't do this, and
we need to do that work of figuring

out what are… W- where we're trying
to go and what we're trying to do

and what we're trying to build.

That's the part that is just like
so lo- Uh, this is why things like

vibe coding and all of that is like
I think that there's this like pent

up, "Oh my God, like I could just do
whatever I want, and this is awesome."

And in many respects, it does get to like,
well, what does that mean for the shape of

companies long term and like all of that?

They… You could go macro level
economics there if we want to, but

also, so long… You're still going
to have companies, groups of people

who are all trying to coordinate,
orchestrate, make something happen.

And when you arrive at, "Oh, we all
can't just vibe coder ourselves into

these seven different directions. We
all need to have an agreement." And

where do you store that agreement?

Like how… What does that look like?

Nathan: Yeah.

Brad: I

Nathan: I

mean, thinking metaphorically, I, I
see a lot of, uh… across multiple

teams that I'm working with, it's
as if we've created these behaviors

to align on a shared understanding
of where we're going, and it's like

we're always or- or like together, uh,
individually orbiting that shared space.

Brad: Yeah.

Nathan: of AI, like makes you feel
so empowered, and you can start

solving more problems, and you can

Brad: Yeah.

Nathan: going so much faster that it
triggers the risk of you're just gonna

escape velocity of that orbit, and you're
gonna go not just towards another planet,

but a whole another sol- solar system.

I do see people like exiting
like that from the collaborative

space that others are on.

I haven't seen this person in a while,
or why are they even doing it this way?

I guess we need to do it another way.

And suddenly collaborative glue
that was holding us together or the

gravity, I guess, in my metaphor,
that was holding us together And

that's-- AI is not necessarily…
That, that's not a solution.

That's not, uh, like you're creating
problems that you need to solve rather

than solving problems that you had.

Brad: Yeah.

Uh, it's, that's an, it's an interesting
lens, I think the story you just told

is such a testament to the fact that
this is a, a general purpose technology.

And I think that this is the thing that
we are in a collective moment of trying

to understand what that actually means.

Because what you just described is a
real phenomenon that's, that's happening.

And then also,

Nathan: different

Brad: y-

Nathan: directions.

yeah.

Brad: yeah.

And what I'm, what I'm seeing is another
application of it, which is we can take

a diverse group of people on a phone
call, on a call like this, and kinda go

through, "Here's what I'm concerned about.

Here's what my goals are.

Here's what my values are.

Here's, here's what I'm seeing over
here." And, and to bring all of that

together, and historically we'd be
like, "Wow, that's a great meeting."

And now we're actually able to, like,
act on that stuff, so the stuff is

actually, like, this glue that's actually
moving everybody closer together.

Or I'm taking, uh, my product work,
my own product work, and I'm firing

off GitHub issues and poll requests to
multiple repositories that I, I have

access to at the same time, and they're
all cross-linked, and they're all… So

it's like there's this, oh wow, you can
use this to, like, a bunch of people

are just gonna go off to the races and,
like, never be seen again or g- chart

their own paths and systems be damned.

And that very same technology could
be used to kind of b- bring, bring

people closer together and collaborate
more closely and create shared

understanding and shared, you know, uh,
like, agreement and all of that stuff.

It's, uh, like, really fascinating.

Nathan: I was just on a call this
morning with a, uh, let's call it

the design lead of one of the design
systems teams that I'm working with,

we had an agenda for a half an hour,
and it was just like all the different

areas that are s- creating separation.

Do we need multiple Figma files
'cause now we have too many

collaborators in the same one?

Do we need to… You now have
collaborators creating patterns and

view templates that they didn't,
never created before amid all

their components that they have.

How do we bring those together?

Oh, we have people making a prototyping
toolkit that is beside Figma and a

different way of expressing design.

How do we keep those things together?

And it was just, y- you look at
everything that we talked about, and it

was really how do we sustain the gravity
to align these things and keep us all

orbiting those as different people build
different parts of the bigger system?

Brad: Yeah.

Yeah.

Yeah.

What do you feel… 'Cause that, you,
that brings up a really important tension

that we've always seen, which is the, you
know, here's the pattern police and here's

the standards, and here's the, the people
who are the wet blankets coming in to, to

thwart creativity and invention and stuff.

So, like, given what you just said, how
do you see those kinds of attitudes?

Which I've never… You know, like,
there's a caricaturized version of that

pattern police, but we've also seen a fair
amount of it where it's just like this is

the right way and this is the wrong way,
and y- you just described quite well this,

this pull or all of this new… Like,
what, what does that mean for the literal

DNA of what a design system needs to be?

Nathan: Yeah, like I know Ben
Callahan, who we both know, talks

a lot about culture, but I think

Brad: Who's that guy?

Sorry.

Nathan: No, Ben's great.

Um, but there's a dimension of the
culture in which the design system

exists uh, you could look at as, as a
scale of flexibility or autonomy or the

ability to be creative and innovate,
however you wanna depict that, versus

Brad: Mm-hmm.

Nathan: governed and having closed sets of
choices that you shouldn't expand beyond.

And things like AI, things like
prototype, things moving out of Figma

into prototyping where you can really
express more quickly divergent things.

uh, even within Figma's the,
um, slots capability, I th-

Brad: Mm-hmm.

Nathan: some teams adapting to
that and shifting along that

scale to become more flexible.

I see other design system teams
that I work with, people are writing

about, uh, those slots are too open.

Like it actually created
more problems than it solved

Brad: Yeah.

Nathan: people can, oh my God,
put whatever they want in a card,

and we don't want them to do that.

And I'm like, "Oh my God, we totally want
them to do that." Like, it challenges the,

the, the culture of permissiveness that

Brad: Mm-hmm.

Nathan: has.

And, uh, some design systems will
hold on tighter to that than others.

each has trade-offs, each have good
and bad points, each have reasons for

being on that s- point in the spectrum

Brad: Yep.

But I think that the people who skew more
towards the lock it down, uh, terrain

are gonna have a harder time these days.

That's, that's my, uh, prediction.

Uh, you know, not, not saying that
it, that there isn't validity to it

or continued validity to just being
like, "No, you, you need to use

this stuff as is," but I do think
that you have a much bigger battle

than just yelling at three teams.

It's like…

Nathan: live in that space at all, like
being pattern police and being a strict

governor, and we, we used to call this
with one of our stakeholders, um, the

no storm, like instead of snowstorm.

They're just a no storm.

"No, no, no, no, can't do all
these things." We're like…

"Oh my God, I don't wanna…

I can't work like that." But I can
imagine a future where there is a,

um, large corporation that has a
very strong brand, and they're going

through a rebrand that is going to
change very specific yet fundamental

aspects of their visual foundation.

Wouldn't it be great if everyone else
had implemented that entire component

library and system that they had in
a manner that all they had to do was

centrally change a few things, and
they essentially imprint that new

choice across a wide swath of territory
because they had a more closed system?

Brad: Yep.

Yep.

Nathan: absolutely business value in that,

Brad: Yep.

Nathan: at

Brad: Yeah.

Nathan: of other things, for sure.

But brand consistency, uh, rate of
change of a major change, you know, all

those things are positive consequences
that people look for out of a system.

Brad: Yeah.

I think, I think that there's, there's
certainly something to that, and that's

certainly been the, the, the reason why
we're like, "We want you to be, like,

actually connected to our system." Uh, you
know, yeah, you download the new version

of the package, you get the new brand,
or you download the… You know, you get

this and you get all this stuff for free.

I do think that this moment is very
fascinating, and again, it's a very

double-edged sword, because in one
respect, yep, installing that package

and using that thing and putting
these tokens in, like that is…

You could hang your hat on that.

And what I'm seeing is this new emerging
non-deterministic way of arriving at,

it's not the same place, but gesturally
speaking, is it indistinguishable?

Yes, at this current moment, but,
like, it's also Kind of inventive

or, like, it captures the spirit of
something, of a visual language, of

brand expression, of stuff like that.

But it's just arrived at at a brand
new and totally different way.

And so that, that, like, you lock
down the system, you create these

connections, and this is how-- this
is the pipeline of how you get--

uh, how you evolve, uh, a, a visual
language through software over time.

And all of a sudden, like, uh, it
seems like this, like, new way is

kind of being born that does not
follow that same pipeline, and that

could be bad in certain respects.

But it's also very interesting in other
respects, where it's just like, "Oh, yeah,

here's this, like, internal tool that
sucks, and it's gonna be a giant pain in

the butt to adopt the system directly."

It's like, I don't know, just throw--
say like, "Here's the, here's the brief.

Go ahead and turn it into our thing
rather than this old crappy thing."

Nathan: You talked, uh, through a number
of different sort of dimensions of that,

but one of them was precision that you
get out of determinism, and like, this-- I

feel like I'm in this, mind-bending place
with AI because, uh, it's all predictive.

Like, uh, I-- for those that don't know,
I have a master's degree in statistics,

and so I am all about predictive models.

I am all about understanding variation
and understanding, a degree of precision.

We call it, like, p-value is a degree of

Brad: Mm-hmm.

Nathan: p-value of .01 or .05, which
means like, okay, you're, you're pretty

convinced that this is good enough.

and I see that kind of playing out like
you talked about, like, uh, coming from

a non-deterministic place and delivering
our designs in a way that are, are

indistinguishably good enough, and I

Brad: Mm-hmm.

Nathan: to that.

Uh, there's speed to that.

but I, I'm also a math guy.

I have a degree in mathematics as
an undergrad as well as statistics

too, and math is the study of truth.

is

Brad: Mm-hmm.

Nathan: of

Brad: Mm-hmm.

Nathan: you can prove, And so I
got a real problem with people

creating agentic tools give away
precision that is easily achieved.

So if I have a bunch of design intent,
and I know all the choices that we

made, and they can b-be all enumerated
and very specific, and it costs you

why would you use an agent to crawl a
Figma file, spend 150,000 tokens, and

at the end of it provably hallucinate,
overconfidently un- make some other

decisions that are in contrast to
the exact things you just crawled,

and frankly miss a lot of signals
because you're looking in a sea

of data and find where the signals
are, and you miss those things.

Brad: Yeah.

Nathan: you use predictive models to do
something less precise that costs more

money takes- 20 minutes can do something
in 30 seconds that is mathematically

verifiably proven the right answer?

that's me getting on a little rant,
but it gets at these two extremes

of how you think about using AI.

Brad: Yep.

Nathan: I am not willing to be an AI
maximalist, to use it for everything,

when we still have better things that we
can use to solve some of the problems.

without a doubt, AI helps you solve
so many other problems and solve

problems you would never thought
about actually solving before because

you didn't have the capability.

So this isn't to say AI is wrong, but I
wouldn't use it for everything when you

Brad: Yeah.

When, when all you have is a hammer.

Uh, that makes a ton of sense.

I'm very curious to hear your answer
to your question, which is, why do

you think people are doing that?

Nathan: Because they're being told to.

Or, like, this is gonna sound
unbelievably arrogant and self-righteous.

Because,

Brad: Hmm.

Nathan: They have aversion to tools they
didn't build themselves because they don't

understand necessarily how they work.

To them, they might be a black box.

To them, they might be complicated enough
or their intent, the tool's intent not

communicated clearly enough it, they can't
come to a point where they say, "That's

a good fit for me." And you know what?

That's okay.

Like, I don't know how
to like do brain surgery.

Um, I'm not necessarily gonna go
and cut open somebody's head either.

But like, I don't know how to do that,
and there's, uh, I'm not gonna necessarily

try to learn all the tools to do that.

But know if that's a good metaphor.

That might be a really crappy one.

but the point being like, I, I need to
approach these things with an aspect

of humility because even the tools
that I build that frankly, to other

people on my-- that are teammates,
sometimes feel like a black box.

It's too much for some of my teammates.

But other teammates are like,
"Oh my gosh, I just saw all the

wrong things I did because I ran
it through that tool." great.

Let's correct them.

Let's be better.

Let's improve.

Let's use it to

Brad: Yep.

Nathan: that

Brad: Yep.

Yep.

Yeah.

Yep.

Yep, yep.

Nathan: I don't know.

Like, I guess I just see a lot of people
feeling like the future needs to be

maximized in terms of its agentic workflow
to minimize or remove the human aspects

of making choices at the expense o-
sometimes of actually tools that we can

build that don't require agents at all.

Brad: Yeah, it's a, it,
it's an interesting lens.

I do think that your initial answer
of we're being told to, I, I also

think that there's, flipping that on
its head, I think that there is a, a

moment of curiosity and exploration
that people, even people who have come

from more of a structured appreciation
of, you know, the other way of doing

things, it's like, "Ah, here's this.

What happens if we do it this way?"
that's an interesting place as well,

because that, that kind of curiosity
and exploration feels good, but of

course, there's, there's risk with that.

Uh, and then I, I also think just from
a brass tacks perspective, and this

is, like, really coming true for me,
is the ergonomics of it, the UX of it.

Like, I have ADHD, and going through and,
and having to change out of a context in

order to, to do something… Like, I'm,
I'm, like, really finding a lot of value

is I'm like, "I want you to… I'm looking
at a singular thing, and I want you to go

update the Notion task, update the GitHub
repo, change my branch, move over here,

create a folder, do this thing," and I'm
going to, like, be talking all the while

and, and I'm able to be, like, locked in.

And so it's very interesting because
you can design these things, even though

it's just, like, a flat interface now.

You can basically say, "My instruction
is the same as, like, saying, 'Hey,

Google, turn on the lights.'" It's
like, "Hey, Google, go run the specs

workflow," or "Hey, you know, go
run this deterministic process."

That, that is, I think, s- uh, a place
where I would like to get to, is,

is where that stuff is interpreted
as such faithfully and reliably.

It's like, how do you design the
experience of this so that we get the best

values of this, like, non-deterministic,
weird prediction machine, and the best

values, the best parts of deterministic
workflows and things that it's like,

I always want two and two to equal
four, and I want to run it through

this formula always to arrive at that.

Like, please don't make
two and two equal seven.

Like, like, don't go and invent that.

Use this thing.

This is the math part.

I love how you said it.

Like, you, like, through your
degrees, you have this, like,

very interesting perspective on…

Nathan: Jekyll and Mr. Hyde-ish

Brad: Yeah.

Nathan: thing sometimes, but,

Brad: Do you, do you actually feel
like you're, like, switching between

those modes, like, in real time?

Like, in a, in a meeting or something?

Nathan: I mean, ultimately
I'm a stats guy.

I love dealing in the var-- the nuance
of the variation and the lack of

certainty that we have with things.

Um, I find that the, um, that
brings out the best in me.

It makes me more curious.

It makes me more humble and more willing
to, um, understand that what I thought was

truth actually wasn't, it was just a bias

Brad: Yeah.

Nathan: Um, so I'm, I'm at the
heart a stats guy, but I'm still

a guy that's gonna pick math when
math is, math is relevant until it

Brad: Yeah.

Nathan: that, oh my gosh, we need to go
quantum with something, and then I don't

know what the hell quantum physics is.

Um, so the, uh, the thing I was thinking
about when you were describing, um,

your experience was, uh, sometimes how
to y- control the conversation, 'cause

there are opus tasks, there are sonnet
tasks, there are haiku tasks, and

like if you choose the wrong thing…

My, my biggest problem with AI
is, um, when I ask it to do a very

direct, simple thing, which is like,
"Hey," equivalent to, "Tell me what

the tokens are, the change for the
alert on which elements when it's an

Brad: Mm-hmm.

Nathan: error, or success message."

Brad: Mm-hmm.

Nathan: that is a mechanical task.

And suddenly, like AI's like, "Well,
but the focus order or the heading to

the description might be inverted, and
you might wanna start thinking about the

motion when this thing slides in from…"

I'm like, "I didn't ask that question."

Brad: Yeah.

Nathan: you're in

Brad: Yeah.

Nathan: space.

But

Brad: Yep.

Nathan: relevant questions.

Um, and so timing, the depth, the duration
I wanna spend to answer certain things

based on the workflow that I'm in.

And if it's me, I typically have like
three or four Claude windows open

running simultaneous tasks, and that
can get a little bit chaotic mentally

Brad: Yeah.

Nathan: I can empathize, like the,
that each person approaches those

conversations in different ways

Brad: Mm-hmm.

Nathan: different things out of their
tools, and wants to go at intentionally

different speeds or different depths or
different durations to solve a problem.

And to me, that's okay.

I think the problem that my tools
are trying to solve is time something

changes in Figma across an entire Figma
file, how do we signal that change

in a very precise way immediately?

Brad: Yep.

Nathan: that's what agents are gonna want.

They're not gonna wanna hunt and peck and
inspect via MCP all these tiny nodes and

Brad: Yeah.

Nathan: all this differentiation
across 100K of JSON data.

They're

Brad: Yeah.

Nathan: know that for warning,
we just changed the yellow token

to be one level shade lower.

Brad: Yep.

Nathan: it needs to know.

Brad: Yep.

surprise, surprise, it's using the right
tools for the job, and also i- in the

case of the weird and inventive and
non-deterministic one, trying to corral

it into, slapping it on the wrist.

It's like, "Please don't,
please don't go here.

Please don't rock that boat.

This is, this is math territory.

S- stay out of this." But also, I
kind of appreciate your perspective

on this, but just not right now.

Like, I think that there's a lot to the
UX of, of how we're all interfacing with

these things, but I think that, like,

Nathan: Yeah.

Brad: a, a very weird thing to, to say, I
think, is that I have seen so much value

in, like, a consolidated, like, window.

This simultaneously, like, excites me
and scares me, like so many things about

AI, is I'm like, I like having this
single point of entry or the single place

where I'm communicating with something,
and it's all connected, and it's just

these things are, like, trickling out.

They're being woven in together and
coming back to me through the same thing.

I'm like, that feels good.

That feels natural.

That feels, that feels a- amazing.

But at the same time, like, that…
Does that mean, oh, these other

things don't matter anymore?

Kind of far from it.

It's like we in, in, in any case, we're
able to just, like, kind of have this

command center view that historically,
for my brain at least, would just

short-circuit me, and I'd be like,
"Wait, what the hell was I trying to do?"

And so now I'm able to just, like,
stay in the place where it's just

like, do this and do, like, the
seven chores that are around it.

Including writing a bunch of
non-determinist, or, or deterministic

code, and then let's chat when
that's in place, and then we

can kinda… I, I don't know.

It- there's, there's
really something to it.

But, like, experientially, as someone
who's been using a computer for nearly

my entire life, and I'm sure you're in
a, in a similar boat, it's like this

feels like a pretty radical overhaul
of, of how we talk to computers.

Nathan: Yeah.

and it requires adaptability and
humility to, to go through periods of

change that are really uncomfortable.

Uh, and yeah, it makes
me really uncomfortable.

I yell at Claude when it's
taken, like, five minutes to do,

like, a stupid mechanical thing.

but also I appreciate that today I got a
ticket for-- Like, part of the work that

I do now is creating a schema that, the
multiple platforms all have as a data

structure to implement their components,
fact that it needs to be unbiased across

platforms, and Figma is essentially a
fourth language to web iOS and Android.

Brad: Mm-hmm.

Mm-hmm.

Nathan: I got a ticket today.

We had just finished a big chunk of
layout, uh, schema work, and we got

through absolute positioning, and we're
using terms like start and end instead of

left and right, and Figma's constraints
model is left behind in its translation.

somebody on, uh, working on
iOS was like, what's the deal?

Is this Figma bias?" They're always,
like, skeptical because of where we're

starting from with some of our ideas.

axis alignment property on
iOS is center by default.

And I'm like, "I haven't built something
in iOS ever." Like, I'm a web guy.

Like, I could have asked AI
that, but that was, like, on the

very long tail of properties.

And so that's still that bias crept in.

That I'm even that far on
an edge of cross-platform

relationships helps me continue

Brad: Yeah.

Nathan: you gotta-- That, that's really
uncomfortable for someone to call me out

when we're-- We have a skill, and it's
supposed to-- It has criteria and gates

it runs through to unbiased decisions
per platform, or across platforms.

Like we, we run our architectural
decision records through that skill.

We missed one.

It's leaky.

It happens.

Brad: Well, yeah, 'cause you're, 'cause
you're using it as an opportunity

to expand your horizons and to learn
and to, uh… Like, there's, there's

all sorts of stuff to, if you're an
interested person, if you're a curious

person, these things are just curiosity
amplifiers, where it's just like, "Ooh,

tell me more about this," or, "Here's…

What, what's that weird edge
case that-" slips between the

cracks of cross-platform stuff.

Hell, like, yeah, we've done a bunch of
clo- cross-platform systems over the year,

but, like, I would never consider myself
to know a lot of the inner plumbing of it.

We're often just collaborating
with the native developers,

and it's like, "Hey, what, what
shape do you need this stuff in?"

kind of thing, right?

So, so no doubt you're able to get
a lot closer than you ever did.

Nathan: I would have never imagined
two years ago that we'd be generating

specifications that live outside of
any tool, including Figma, and that

we can change them predictably, and
they're unbiased across platforms,

and they're, as far as the design
intent that we can think of trying

to convey through them, complete.

okay, behavior, accessibility, motion,
those are green territory for us to do.

But, like,

Brad: Yep.

Nathan: that we've always been doing,
running out of challenges, and that's

Brad: Yeah.

Nathan: uh,

Brad: Yeah.

Nathan: there's like a data structure
that, that can house all that stuff.

Brad: You're just the right
person to do that work.

I'm, I'm really excited.

what does this look like for you as,
like, this as, like, a product and

you as a, a, a person creating that?

Like, like, what does that look like from,
uh, how you're situated or organized?

Nathan: Well, the, the tool that
I'm creating is, uh, a natural

n-next step from the Specs plugin
that I did three years ago.

Brad: Mm-hmm.

Nathan: the… What the Specs plugin
does is it evaluates a Figma asset

and, uh, generates its anatomy, its
properties, its, uh, layout variation,

its, uh, essentially visual variation
across, uh, various configurable

properties and so on, creates inventories
of styles and that kind of stuff.

But it's, it tries to derive everything
that it can out of a Figma component.

so- As Figma's gotten better, uh,
but slowly, like Figma's model of

data that it can use to ascribe
design intent, uh, to a component

is moving too slow, but it's moving.

It's still moving.

but really shifting from presenting
it on a canvas to presenting it in a

structured data form that an agent's
gonna be able to consume really fast.

When I… And I was inspired like a year
and a half ago when somebody said, "I

took a screenshot of your canvas specs
and gave the screenshot to AI to build

the component." I was like, "Oh my God,
they're like evaluating an image that has

all the visual specs." And I was like,
"That is not the way to do this." so,

uh, I have sort of fortified the tool.

Um, it still works as a plugin, but now it
works on the command line too, and it will

Brad: Yeah.

Nathan: you set it up one time to say,
"Process these 50, but not these 200 other

things that are corrupt in the file." And

Brad: Mm-hmm.

Nathan: redo it like in less
than 60 seconds once a day or

Brad: Yeah.

Nathan: once an hour,

Brad: Yeah.

Yeah.

Nathan: often you wanna do it.

And so you have all the data now.

and so to me, that kind of repetition
is really interesting to the

consumption of, of those decisions by
the engineers and the agentic systems

and workflows that they're building.

Two other things it's having me think
about is, well, what kind of format

are agents and humans gonna wanna
s- have to solve different tasks?

so the kinds of things I'm working
on now are, first, additional

things that Figma can't carry.

Figma is good at visual, really good or
fairly good at properties, but incomplete,

somewhat good at layout structure.

nobody uses its motion, nobody
uses its behavior, and it's

incapable of accessibility.

So if you think about all
those attributes, Figma is

varyingly good at those things.

And

Brad: Yep.

Nathan: like I have this blog post in my
head, like living on the edge of Figma,

what do you do about the things that Figma
does not as well or poorly or not at all?

And how do you merge those ideas with
everything you can derive from Figma

into what might be not one document,
but a interconnected set of assets?

Brad: Yep.

Nathan: does that need to take for
someone building a React component?

How about a contract that you can
auto-generate of the prop types?

100%. You don't need an
agent to derive that.

You can produce that.

How about a Tailwind-based set
of styles ascribe to Tailwind's

philosophy of how to build class
names and, and how to have that work?

You can generate that.

How

Brad: Mm-hmm.

Nathan: a component MD file that has
prop tables and other characteristics

and layout in markdown form
that can, a human can read too?

Yeah, you can generate that.

And so there's a lot of different
things you can project from Figma

into shapes that are further along and
closer to the shape that some consumer

on iOS or Android or wherever is gonna

Brad: Yeah.

Nathan: that stuff to be shaped as.

Brad: Yeah.

Yeah.

Nathan: and the other… So I'm
thinking about additional concerns like

accessibility to behavior and motion.

I'm thinking about, um, how to
project things into the shape that the

consumer wants, human or agent or both,

Brad: Yep.

Nathan: and, and how to
version those things.

And ultimately, like, people
wanna get rid of Figma.

I, I'm really frustrated when I hear
people be like, "Yeah, uh, we should

ditch Figma and go into prototypes."
And I totally believe in prototypes.

You and I have spent a lot of our
careers trying to figure out how to

inject prototyping into the design
process off the static canvas.

But there's the baby and bathwater of it
all, because what is Figma awesome at?

It's a strongly typed visual database
of all your design intent for certain

ca- categories of design intent.

It is awesome, and until you have a
better tool to serve as a database

of a lot of those tool choices,
why would you throw that away?

Why wouldn't you broaden your toolset
to solve more problems in interconnected

ways rather than throwing out the
toolset actually fits many scenarios

really, really well just because
it doesn't fit other scenarios?

to me, shortsighted.

Brad: Yeah.

Nathan: trying to think about the
relationship of, of Figma to other

tools and how to, um, carry the
design intent in a way, particularly,

again, 'cause I'm cross-platform.

If I was just building a React catalog,
oh my gosh, I'd just be working

in straight code the whole time.

Brad: Yeah.

Sure.

Nathan: the design ideas.

But when you're building across f-
at least five outputs, uh, web, like

say a React, iOS, Android prototyping
toolkit, basically web, that has a far

lightweight, far more lightweight package
and includes a lot of experimental stuff,

plus the Figma library, that's five
different libraries you're publishing.

You gotta have your design intent separate
from those things, as opposed to thinking

that AI's gonna all of those things
together without any hallucination or

Brad: Yeah.

Nathan: That…

Brad: Yep.

Nathan: that, the, that's where I'm
working, and that's sort of where

the space, the things I build fit.

Brad: you kind of almost implicitly
told your own journey of, like, thinking

about Yes, Figma is the source of
truth," and then it, like, kinda goes

downstream, and it's this direction.

It's… That's the workflow, right?

what you described is, like, this
is… kind of doesn't have it all.

it's covering some of this stuff,
covering other stuff imperfectly or

poorly, and then this other stuff is,
is absent, and in certain respects can

never be present because of just the,
the literal DNA of, of what the thing is.

So, the Figma source of truth camp, I
think, is having a hard time right now,

or the people that are, like, used to this
stuff, like, only going in this direction.

now all of a sudden you have
other ways of, of doing this.

And I think that th- this is one of
the other things that we're having a

really hard time understanding in this
moment, is the idea of, like, pluralism

or, or plurality as just, like, here
are different groups of human beings

that are organized in different ways,
that have those different cultures.

Again, rigid, flexible, all of that stuff.

For that React-only company or that,
that place that doesn't have, like, a

full-time staff designer, it's like,
well, here's the shape of a, of a toolkit.

if you're in a place where there's,
like, dozens upon hundreds upon thousands

of people, like, using these tools,
probably doesn't make a sense to, to

throw the baby out with the bathwater,
like you said, just, like, burn it all

down and jump into the brave unknown.

But for, like, other places,
they have a different shape.

And what if different human beings have
different needs, and we need to be able to

evaluate the, the tools and technologies
that we reach for to serve the problems

for the specific shape and configuration
of people that we're working with?

What if, what if that one right way of
doing things was never really actually the

Nathan: So yeah,

Brad: one right

Nathan: one right way

Because of all the different
constraints and dimensions of how

people and organizations are different.

Brad: Hmm.

Nathan: and so yeah, like,
does Figma fit for everybody?

No.

But, like, is there-- Should product
designers and system designers use

the same tools and solve all their
problems with the same s-set of steps?

Oh my God, no.

No.

So when a VP, like, starts saying
everyone should be using, prototyping

for everything, you are really gonna
start handcuffing a lot of people

that solve problems with better
tools that they already use and, and,

uh, potentially put constraints on
ways of doing things that are going

to erode rather than create value.

Brad: And it, and it, and really what it
comes down to is, I think, like, a real

reckoning with… I think, uh, there,
there have been dynamics at play that

make me sad over the course of maybe
the last decade or so, where I've seen

Nathan: Yeah.

Brad: just live entirely
inside of the walls of Figma.

And now it's like, and, and
it's like that was never the

case, but that's what happened.

That's, like, what ended up happening.

And, and so the… in many respects,
it's like the reintroduction of like,

"Hey, there's actually, like, a bunch
of things that you can do. You could

just remember Post-it Notes, and
remember," you know, like, all of these

other kind of fun things that weren't
just living in an art board somewhere.

Like, it's, it's like there were,
there were different activities that

we would do, and then it just, like,
all kinda just got centralized,

and now it's like, I think that
a, decentralization is happening.

And I think that, like, for Figma's
own sake, I think that they were really

trying to do everything they could to,
like, get people into their, into their,

uh, world, as a, as a business does.

And now it's like, oh yeah, like, all
this shit needs to play with each other.

All of this shit needs to be
compatible with one another.

We actually, like, have no choice
but to, like, work, uh, with and

be connected to, to one another.

Nathan: I mean, that's a real
testament to Figma's point of

view, and God, if they turned it
off, I can't imagine the backlash.

But Figma is a really open environment.

The plugin API, the REST API,
people build all sorts of

top of it.

They

Brad: Yeah.

Mm-hmm.

Mm-hmm.

Nathan: data, and they… was shocked
when somebody said, "Yeah, we, uh,

created a backdoor into Figma's plugin
API through Claude to just connect through

whatever the browser socket something
or other it was." I was like, "Oh my

God, what's Figma gonna think of that?

Are they gonna turn it
off?" But it didn't.

So there's an inherent openness to those
things, despite the fact that, yeah,

it's already a super complicated product.

And for them to con-continue glomming on
more and more super complicated stuff,

they're at the long tail of features
now, as opposed to the core cases.

Like, I don't know where they go, but,
uh, credit to them for remaining so open

and having such a good relationship and
trying to help people solve problems.

Brad: Yeah.

And I think that that's, it's
testament to what really any

tool needs to do these days.

Um, and you're seeing just a whole
new generation of stuff being

born, and I think that the more
people… The, the old mechanics

of, like, let's get people in here.

We built, you know, a better
light bulb over here, and we're

going to, like, protect it.

It's like you, you really need to be
thinking in terms of a broader ecosystem

and interconnectedness out of default.

Nathan: the old companies that, uh, sort
of the dead companies behind us, I'm

thinking like Envision, I'm thinking,
AOL, like they're walled gardens.

Like they, they weren't
interconnected, and that

Brad: Yeah.

Nathan: their downfall

Brad: It's, you know, at a, at a more
existential level, perhaps it's the

introduction of this stuff that is
ultimately the, the manifestation

of the old Stewart Brand information
wants to be free, and that technically

happened with the, the web and
with, with, you know, the open web.

But also, it's like there's, there's
still kind of these, like, little islands,

and there, you know, some, some things.

And now because we can just, like,
connect everything together so much

more and so, so much faster, that the
incentives for cooperation, collaboration

suddenly become a lot more urgent.

To, to, to live in our own island, to
compete, to say, "No, this is mine,"

versus more of, like, a sharing, attitude.

I don't know.

It's, uh, it's, like, very, very
interesting, but it's like, it's, it's

clear that the pace of technology is
facilitating something that allows

things to get out faster than they did
before, and I think that that really is

going to shape how human beings operate.

And like, like, what,
like, what's value, right?

Just like you were talking about
your Dr. Jekyll and Mr. Hyde.

It's like, what mode
will yield more success?

And obviously, it's going to be,
like, a, a balance, but, like, as time

goes on, I feel like it's harder and
harder to be like, uh, this is, like,

the one right way of doing things, or
we can push all the other out, This

world is getting smaller and faster
and more interconnected by the day

Nathan: I'm curious, like, to throw it
back to you, 'cause I, I think about these

things as a consultant, as someone who has
thought about and written about and taught

a lot of people about design systems.

Brad: Mm-hmm.

Nathan: AI has encoded so much of not
just all the stuff you and I have written,

like all this knowledge about how design
systems work, that it's kind of a threat

to us as knowledgeable experts in a way.

And, the value of the ideas that I have
now dispersed into these models that

can just convey those things without
provenance, attribution, whatever.

I'm curious how you think about, the
role those systems have as your designer

or engineer is working, and they need
something alongside them to guide them,

Brad: I

Nathan: versus the role that we
continue to play, at least for

some unclear period of time.

You know,

Brad: mean,

Nathan: about that?

Brad: a- absolutely.

But it's like that gets at, like,
well, why are you building the

tools that you're building, right?

Is to continue to…

Nathan: Uh, I'm driven to share.

Like,

Brad: Yeah.

Nathan: for my writing.

Uh, I charge enough for the tools to make
it so that I can working on them, but I'm

Brad: Mm-hmm.

Nathan: massive profits out of them.

Like,

Brad: Yep.

Nathan: the work.

Brad: and that gets to the rub, and I
think that this is the, the, the moment

in time that's unfortunately being laid
bare, which is just, like, capitalism

in general, uh, is, is ultimately,
like, the weak link in all of this,

because I think that a lot of people
have the strong desi- desire to share,

to express themselves, and to want to
help and do things, and we're often

present, or prevented from doing those
things because we gotta freaking eat.

We gotta feed our families, and
we're all trying to find stability.

And then that ends up being like,
"Ah, yes, I have this thing called

intellectual property that I need to
hold onto and hoard, because that's

the thing that feeds my family." And
it's, like, a very hard place to be

because it's like we, we ultimately…

It's like, well, what's that IP for?

Oh, it's to help people.

Well, what if we just get
to the helping people part?

And, like, the more that people get
helped, right, the more that that stuff

scales, the better off everybody is.

And if everyone was given
permission to just operate that

way, guess how much better we'd be.

So, so it's like it gets in the way
of, of, of What those dynamics of,

of just like capitalism do, I think,
is ultimately says I need to be very

mindful about how, how much I share
and like where's the threshold and

what's the thing, and how much do
I hold back, and I don't like that.

I don't like that dynamic whatsoever.

It's like I just want to feed my family,
make sure that they're stable, and I want

that for every human being on the planet.

And so it does, it gets to the
point of, of like I don't wanna ha-

I wanna be able to share freely.

I wanna be able to share
with reckless abandon.

I wanna help people and maximize
that, and that's ultimately a better

world than me with a bunch of money.

You know?

I just want enough to, to, to be
stable and, and feed my family.

Nathan: I think, yeah, I, I can,
uh, identify with a lot of what

you're describing, and I, I think
about, perhaps less directly

value that I can impart And the

Brad: Yeah.

Nathan: uh, hopefully I, I play my
little small role in the community to

build some tools and write some blog
posts and help people learn and stuff.

like, I remember you, you made me think
about also, like what you don't disclose

or you talked to, or you alluded to
that boundary and, uh, Jared Spool,

who's, been in our space for a while
and run conferences and, and was, uh,

kind enough to invite me to some of
his conferences, and I gave a talk

and I talked to him ab- beforehand.

I was like, "I hate giving talks." I
was like, "I literally wish I could

never give a conference talk." I
love giving workshops because of the

interaction and the interplay between
the smaller group of learners and

your ability to connect and see the
change that you're creating with them.

Brad: Mm-mm.

Nathan: Uh, but with talks,
it feels so detached.

And I was like, "But I have so much
to say, and I wanna fit it all in."

And he goes, "Don't fit it all in."
I was like, "What do you mean?" He's

like, "Deliberately omit things.

Remove, and be pretty ruthless with what
you remove." And I said, "Well, why would

you do that?" And he goes, "Well, first
of all, you have like a time limit, and

you have a clarity limit and a endurance
limit of the audience and all that."

And he goes, "But you also have ability
to like, they're gonna ask questions

on the edge of what you presented, and
You're gonna seem really when you have

the answer already locked and loaded to

Brad: Yeah.

Nathan: just because you didn't sort
of cross that boundary in your talk."

And that, that gets at some of
the value that I'm thinking about

around the blog posts you write,
the tools you make, really, I've

been able to then parlay that into a
sufficient amount of credibility and,

Brad: Yeah.

Nathan: the humans really
aren't paying me, the corporate

entities are paying me, and

Brad: Yep.

Nathan: like that distinction.

Brad: Yep.

Yep.

Yep.

Yep.

Yep.

And that, and that dynamic is, is
what we've been through, and that

I think continues to be a way.

It's like we're, we're able to, to…
It, it feels kind of like… Well, it,

it feels kind of like patronage, right?

Where, you know, in the days of
yore, where it's like, "All right.

I'm the pope.

I'm gonna give you some money to paint
the ceiling, and, uh, and, and off you

go." And in a lot of respects, I think
that that's, like, what the kind of

work that we've historically done is,
kind of feels a bit like that, where

it's like I'm gonna be doing this
work that is for the greater good.

But also, like, you know, there's,
there's a way of, of underwriting

that a little bit, and I think
that that's, that's interesting.

I do think that there's, like, a lot
more opportunity now for more people.

I've, I've never really liked
being put on a pedestal.

I think that the both of us,
we've always appreciated the

little hat tip here and there.

You know, that, that recognition I
think is something that all humans

desire and crave, and it's important
to keep that in check and not have

that turn into, like, ego fest.

Everybody just wants their
work to be valued, right?

And more people- Can do that.

And that's always been a frustrating
thing is, is I'm sure you've

gotten this question, right?

It's like, "Oh, like how do, how
did you like get to the point where

you're at?" Or, "How are you like
doing this stuff?" And it's like,

well, through you writing post after
post after post and doing, doing it.

It just kind of lather, rinse, repeat
of this process of just you just share,

and then you share again, and then you
share again, and then you share again.

You get excited about things, you
learn things, you put things out, and

you do it not necessarily to, first
and foremost land a client that could

then help justify doing more of it.

You do it first and foremost
because you're compelled to.

Right?

Or do you feel that way?

Like…

Nathan: I think so.

I, I think that there is legitimacy
to the grind and the consistency,

but also to finding your voice
and being able to, understand and

feel confident in your own voice.

Like, you and I have
very different voices.

but I think they're both really valuable,
and they contribute to a greater

whole of how people think about these
things, um, in really productive ways.

And I get far more, energy when
somebody, just mentions something from

a post that I read, or like today,
someone submitted a bug report for

my tool, and I'm like, "That makes
me feel like what I did mattered."

Brad: What?

Yeah.

Nathan: doing matters

Brad: Mm-hmm.

Nathan: difference, I run
towards those things, and

Brad: Yeah.

Nathan: me to do more.

It's not because I wanna, be the
coolest conference talk person.

I, I like-- I don't wanna like
show myself on-- Like, me on stage,

it's not about being on stage, it's

Brad: Yeah.

Yeah.

Nathan: find value, uh,
conveyed w- and, and, uh, when

I feel like what I do matters.

Brad: That's, that's beautiful.

What advice would you have to have
more people follow that instinct?

Nathan: I hope this doesn't
come out the wrong way.

The-- What I try to do with my
writing, for example, is find the

distinct idea that I feel like hasn't
been shared before, or a particular

point of view about an idea that
hasn't been shared before, but also

then couch it in a real-world story.

Uh, and you could say, "Okay, so you
have a formula for writing your posts."

Yeah, 'cause it, I think it works
because, uh, most of my posts begin

with a You know, three to five paragraph
hook that is about me in some silly

meeting and having some s- aha moment
or reacting to some collaborator in

a positive or negative way, or, um…
And that kind of realness then can lead

into, like, the point of value or the
understanding, uh, the idea that you

have, and being able to describe that.

So it connects emotionally.

I try to connect emotionally with
the reader through my own voice in

the way that I sometimes have really
long sentences, or sometimes I have,

like, really short sentences that,
like, have a really direct, uh,

phrase or something like that, because
that's actually how I'm feeling.

That's what's going on in my head.

And so being able to,
discover that through writing.

Like, if you read my book from 2009,
you'd be like, "Oh my God, you're

sen- you're, you're so loquacious."
And I still am way too wordy with

the stuff that I write times.

But, like, able to do through practice,
find my own voice, and find, find

that, and that gave a platform for
people to soak up the ideas, I think.

Um, and, and so if I was to give,
uh, um, advice around using writing

and blogging as, don't write
the system success story post.

I've literally saved them in a folder
called System Success Stories, and

they describe how people created their
token libraries, and then created

their component libraries, and now
they have a token-infused catalog of

components that is their design system.

Uh, that's the story that everybody tells.

Instead, like, what narrower thing,
what more spors- m- more specific

thing is gonna help us move our
corpus of community knowledge much

Brad: Yeah.

Yeah.

Nathan: Move that boundary out more.

Talk about how to use easing to- uh,
like a motion design token thing with

easing and springs and choreography.

That's what I need to read, man.

There's not enough stuff on that.

Brad: Yeah.

Nathan: the boundary.

Brad: Yeah.

Yeah.

I we're increasingly going to
be like latching onto that.

So all to say it's just
like people are here.

They're living.

Y- you know, so many people are
just constantly looking over their

shoulders like, " Would this make
Nathan Curtis proud?" Or wh- whatever.

Like, "Am I doing this right?

Yes or no?" And it's, and it's like,

Nathan: thinking that.

Brad: I think so.

I, I, I could prove you wrong there.

Uh, I, but I do think that like
there's a lot of second-guessing.

There's like a lot of like, "Oh,
shit." Like, "I couldn't possibly

write because this person's already
written it better than, than I

could," and so on and so forth.

So it's just like, just, know that
just by way of you having a unique

experience on Earth, you have valuable
things that are worth sharing.

I don't know.

Call me biased.

Call me old-fashioned.

Nathan: We all are biased.

You are old-fashioned.

What's next?

Brad: Like, 'cause this is a really
strange moment, and so it's difficult

to- predict the future with any degree
of like, "Oh, yeah, I could see where

this is going." But, like, if we are
to be having this conversation three

years from now and, and I were to say,
like, "What, what surprised you about

Nathan: would be really surprised in three
years the only way that people describe

their components was in a Markdown file.

I would be really surprised.

would be really surprised if in three
years, design system component catalogs

included only 40, quote-unquote,
"core components" made by only a

select group of central people.

Brad: unfolded?"

Nathan: Like, uh, I e-expect instead
richer descriptions of components

that span a variety of formats to
describe them and combine them in those

formats and data in interesting ways.

and I expect our design systems to
spread into different people making

many different classifications of things
that we have only danced around today

Brad: What would you say?

Nathan: components.

I'm calling them… There's a class of
things I'm working on called compositions.

I expect page views, templates,
patterns, like what the… And actually

m- patterns beyond just a chunk of
the UI that's well styled with tokens.

Like patterns of

Brad: Yeah.

Nathan: patterns of sequences,
patterns of information

architecture, patterns of content.

Like all those different kinds of
recurring things, I expect systems to

have a much richer array of concerns
or classes of problems that it

solves and enables people to reuse.

And so I'd be really surprised if
systems look as, quote-unquote,

"simple" as they look today.

Brad: Yeah.

Nathan: not.

Brad: What do you think about a lot of,
like, the, the generative UI stuff and

just, like, more ephemeral stuff and, and
just the fact that, we historically have

been the gatekeepers and taste makers
for how a thing, an experience emerges

onto a screen and how people move through
it, and they're beholden to our world?

What happens when

Nathan: Who

Brad: that is no longer the case?

The people use our software.

Nathan: what world are you talking about?

Brad: I'm talking about a world
where I'm like, this page that has

17 different sections on it and 14
columns of text and the three tool tip

overlays and seven modal windows, uh, is

Nathan: page, Brad.

That, that is nice.

You got a lot going on.

Brad: And I'm just like, "How
about I just want you to look like

this from now on?" And it does it.

Nathan: not sure.

Uh, I feel like taste is a
vague and abstract thing.

Brad: Mm-hmm.

Nathan: people taste through,
descriptive terms, and let's

call it, verbal prose whatever,

Brad: Mm-hmm.

Nathan: lots of gaps,
uh, and is nonspecific.

I think the human's ability not just
to fill those gaps with instinctual

choices, but also call out those
gaps and ask them to be filled.

Like, Brad, what is they?

is the world?

Like, you, you suddenly described in a
much richer way this page with all these

bells and whistles and tooltip overlay

Brad: Mm-hmm.

Nathan: or others.

Um, you know, suddenly that our
ability to speak in a more structured

way and think about our v- how we
verbalize our ideas, uh, I think

is gonna have to improve a lot.

I'm not-- I don't wanna live in a
world where AI fills those gaps for me

without me understanding how they're
being filled and being able to perceive

what those gaps are that it filled.

I

Brad: Yeah.

Nathan: space.

uh, Dan Brown, my former business partner,
and I, 10 to 12 years ago, had this

moment where we were asking our staffs
can you please write your emails and,

like, Slack comms with better structure?

Can you think about, like, the
clarity of the simple and sort of

algorithmic stuff that you write?

Like, we didn't say it in that kind
of flat and really boring tone, but

we wanted to help people describe
what they needed in clearer terms, so

they could communicate more clearly.

Brad: Yeah.

Nathan: And like a year ago, when
I started doing, uh, architectural

decision records for the first time,
and I started having AI scaffold them

and then fill in a lot of the blanks of
them, and then I would know where the

blanks are that I needed to nuance about
them and add more clarity and so on.

Suddenly, the, that shell and that
template, that Mad Lib became so

visible to me that I wanted so many
design conversations to… Can you just

make it like an architectural decision

Brad: yep.

Yep.

Nathan: Tell me the

Brad: Yep.

Nathan: tell me the problem,
tell me the a-alternative to the

Brad: Yeah.

Nathan: what the recommended solution
is, tell me what the consequences

are, and help us make a decision.

Brad: Yep.

Nathan: suddenly, all these design
meetings that should have been Five

minutes long are still an hour long
because they don't have that sort

of structure embedded into them.

And like,

Brad: Mm-hmm.

Nathan: that AI is helping us realize
and communicate m- stronger with that

structure, and I don't wanna give that up
with of ephemeral verbalized taste-like

cues of, make it more spacious."

Brad: Yeah.

Nathan: mean?

Like, let's talk about visual contrast.

Let's talk about the negative space.

Let's talk about, you know, let's start
weaving in more deliberate terminology

Brad: Yeah.

Nathan: about how we describe
our, our wants and needs.

Brad: no, I, no, uh, no, I think that
that's fascinating because I think,

like, what you're describing is, like,
you, you appreciate a, a more structured

language, and that gives you clarity
and helps you understand things in

that when information comes at you,
and you're able to transmit it using

that structure, it reaps good results.

Nathan: But again, I have to
approach this with humility.

I remember in, like, around 2004
when I was working at Nextel, which

merged with Sprint and eventually
dissolved, uh, into a non-telecom.

But I was working with this person who
was in charge of the content design team.

I was describing an information
architecture challenge, and at the

end, she just goes, "You're, like,
talking to a lawyer." And I was like,

"I'm not sure that's a compliment." And

Brad: But I think that, I think that
that's, that, that's it, though.

It's like, you are wired in a way that
you prefer things to be in a certain

way and to be ordered in a certain
way, and that, and that's, that's,

Nathan: value them that way,

Brad: you value them that way, right?

And then there's other people who are
like, " Give me the spacious vibes, man."

what I think is really interesting about
this moment and when we come about,

uh, when we talk about the empowerment
not just of, you know, designers

and developers and how they're now
interacting with and shaping digital

experiences, but the users of the
software, it's like, you remember, like,

the user style sheets and stuff, right?

And, and the ability to just
kinda provide those site-level

overrides and those things.

I think that the fact that we can now
just be like, " I freaking hate doing

my taxes," and like, "But present it
to me as a game, and I'll be into it.

I'll do it," or, or, or whatever.

I know I've blown your mind
here, but it's like this is…

Nathan: do my taxes as a game, Brad.

I'm

Brad: The, you can do that now.

You can literally do that now.

And the thing is, is that behind the
scenes is everything that you're saying.

Back to our conversation earlier about,
like, deterministic, non-deterministic,

it's like the interface of this
stuff I think is, like, a lot, a lot

squishier, fuzzier- More open-ended
that I don't think that, like, a lot

of that, like, sameness or structure
needs to be inflicted on everybody.

It's, it's kind of one of those,
like, oh, maybe here are some, like,

smart defaults, but, like, you could
kinda choose your own adventure.

But ultimately, that, that handshake,
however the person is experiencing with

it or, or interfacing with the computer,
so long as it hits that API or that

database a certain way, and the things
get captured in a certain way, it's

like, well, who gives a crap if it looked
like, a weird MySpace page versus, uh,

something that's just, looks like a…

Uh, y- you get the, you
get the idea, right?

It's like we could have these,
like, hyper-personalized UIs.

So it's like, what does that
mean for design systems?

Nathan: even for one person, it
depends on the moment and the need

Brad: Mm-hmm.

Nathan: sh- should you be delivering a
strict schema for all your accessibility

requirements for every component,
where everything about accessibility is

compartmentalized into very concrete,
explicit properties and hierarchies?

No way.

We are nowhere close to that.

But could we do focus order if we
already know the names of all the layers?

we could do that.

So even, like, thinking about the level of
structure that, uh, fits the task, uh, um,

Brad: Yep.

Nathan: to purpose effectively.

Um, and a lot of those things, uh,
may long live in Markdown files

because it is good enough, but other

Brad: Mm-hmm.

Nathan: aren't, and, uh, that's okay.

Brad: I think, I think it's good.

It's like there's this, there's this
almost kinda like, there's like a

rigid structure, but then there's this
kinda like warm, kind of weird, fuzzy

layer wrapper to it all that just,
that's like looser, more pliable.

And it's these things kinda like traveling
together that can end up helping tell a,

a, a richer story, or just, like, help
people get the, again, extract the good

bits from this new crop of technology
while still making good use of the

hard-earned structure and deterministic
stuff that has got us this far.

Nathan: I mean, that's a trick you
and I play, I assume as consultants,

is people get grounded into their
sides of black and white, and often

Brad: Mm-hmm.

Nathan: that to travel the
expanse of gray and understand

we're gonna end up in fuchsia.

So like, uh, how do you

Brad: Yeah.

Nathan: of break down that, that
strictness, that rigidness we

were talking about before to,

Brad: Yeah.

Nathan: the plane?

Brad: Is there one thing that you've
learned about- Like, you bring with

you this deep store of experience
and, and expertise, like, borne over

all of these different engagements.

Like, is there one thing that helps you
kinda lock into their specific stew?

Like, the, the… Or, or is it, like,
really just like a, you have no choice

but to, like, do the, the hard work
of just, like, getting in and closer

and having a, a shitload of different
conversations with a bunch of people,

and then that becomes known or it…

Or do you, do you, like, kinda come in
and be like, "Hey, what makes you weird?"

Nathan: Yeah, like, uh, I-- this comes
from a position of privilege, I think.

But, um, I'm at the point in my career
where the choice to decline engaging

is the first choice to, to consider.

Um, when you see an opportunity, whether
it's with a teammate or a potential

collaborator in your current job, or
whether to switch jobs or take on a new

client, of saying like, "I can't," or,
"I shouldn't," " I'm not gonna grow as

fast taking this opportunity relative
to that opportunity," and being able to

Brad: Mm-hmm.

Nathan: So I think saying no, but also
as you start to engage with somebody, the

other thing is the early signals that you
can perceive to understand the terms of

engagement, and I mean, almost literally
one of them being how a company chooses to

engage with you, like to form a contract,
to form the, uh, how they're gonna pay you

and, um, how easy they're gonna make that.

'Cause the best clients
make that, like, effortless.

Like, the fact that realize how to
get the best value of a, out of a

consultant is to not have them focus
on communicating with procurement,

Brad: Yeah.

Nathan: that's smart.

There's business value out of that.

And there, there are sort of… That's
a microcosm of how, like, just design

system collaboration with partner teams
or product groups trying to express

design ideas, like there are those
kinds of implicit contracts and You

could call them vibes, but like to
me it's, it's more how are we going--

what are our terms of behavior and
how are we gonna engage with each

Brad: Yeah.

Nathan: And I am drawn to the people
where those terms are, match my own

value system and, and way of working.

Brad: Yeah.

Yeah.

Yeah.

if they're on fire, in the
process of just getting you in

the door, kinda, kinda red flags.

But if they're like, "Here's… We know
we're bringing you in here for these

reasons, and we can kinda co-create,
like, how you're gonna be there."

Nathan: yeah, or unless they're--
if they're a government entity, and

this is like top secret clearance
stuff, I get why you're being opaque.

I'm not cleared yet.

But most of this stuff isn't
nuclear bombs and terrorist attacks.

I mean, most of this stuff is
like design and screen so somebody

can like buy a plane ticket or,

Brad: Yeah.

Nathan: purchase a, a, a wine cooler.

I don't know, something silly like that.

Brad: That's, that's your real dream, is
to open your own wine cooler business?

Nathan: No, my real dream is to, um… No,
I'm-- frankly, this is gonna sound stupid.

I feel like I'm living my dream.

I wouldn't change my career for anything.

Like I, I've

Brad: Amazing.

Nathan: and I hope I get to do these silly
puzzle-solving challenges for another

few years before I have to kick it.

Brad: for, for… 'Cause there's,
I think, kind of a new generation

of people getting into this kind
of work now, and we've kinda talked

about that, and you talked about, you
know, the, the privileged position of

being able to, like, turn down gigs.

Like, what, what advice would you have for
this kinda new generation of people who

are getting into this world or setting out
on their own to help people and places,

get better design systems in place?

Nathan: The things I find myself telling
those that I advise, uh, often, um,

are things like, run towards the stuff
that's gonna get you the most energy.

Uh, think about the opportunity cost
relative to other things that you

could do instead, um, and what, uh,
how you're balancing those trade-offs.

uh, also, I'm incredibly impatient.

Don't waste your time.

If there's a dead end or a thing isn't
working out, life is too short to just

coast along and with it and not create
something that you're proud of or create

the next iterating version of yourself
that is gonna be where you wanna end up.

Brad: Yeah.

Yeah.

No, I think, I think
that that's well said.

I, I think that's well said.

Like, fo- following your energy,
I think that a lot of people think

that this line of work, a- and it's
true that, like, there are certain

activities that come with this terrain,
and there's things that you do.

But ultimately, it's like, well, do
you want to be doing those things?

'Cause if you, you aren't interested
in doing those things, or you're more

interested in a specific technical
solution, um, I was just having this

conversation with somebody who was, like,
getting into the world of design systems

consulting, and they're, like, looking
to get their foot in the door, and it's

like they also have, like, some really
interesting philosophies around kind of

a specific way of doing this stuff, and
they've, like, thought deeply about it,

and it's, and it's really impressive.

And I'm like, "Well, here's the thing.

Do you want to be out there
sharing this solution?

Or are you more getting into design
system consulting, which is- You're

kind of showing up going, "What's the
problem?" And even if you have this,

this specific solution burning a hole
in your pocket, you kind of got to know

when that's appropriate or when that
isn't, 'cause, like, that might not be

what they want or need or I don't know.

So, so it's like there's this kind of
like vague, like really knowing kind

of why you're running towards the
world of design systems consultancy,

I guess, versus, versus like

Nathan: the…

Brad: I'm trying to do a very
specific type of project.

Nathan: I remember, um, you made me think
about a related, um, that I had with

Lou Rosenfeld from Rosenfeld Media, the

Brad: Lou.

Nathan: and He wrote the book,
Information Architecture for the World

Wide Web, which is why you and I are
having this conversation, which led

me to information architecture, which
led me through into design systems.

And, um, he gave me advice once
w- before I wrote Modular Web

Design, and he said, "Blog first."

be able to describe problems and
solutions, to describe the experiences

you have and the beliefs you hold
and the principles you abide by.

You know, break it down, man.

Like, start talking about those different
ideas and see if that kind of stuff

is gonna come together into a book.

And you know what I did?

I totally avoi- I, I was like, "Nah,
I'm gonna just write the book."

And it was too long, too dense.

Like, you know, It was n- not
a ending that, like, you… I

should've built more to that.

And I feel like design
systems can feel the same way.

You look at somebody else's book, or
you look at somebody else's system,

and it's this rich array of all
these different things, and it's

scaled to solve all these problems.

But, like, if you're starting from zero,
or if you're starting from a much, uh,

smaller place, you gotta figure out the
steps and how those steps are gonna start

fitting together into a bigger whole

Brad: Yep.

Nathan: that's what capital D, capital
S design system is really gonna be,

feel like you experiencing over time,
and it's not gonna happen overnight.

And every once in a while, I've only had
one that went truly sideways and failed.

And it wasn't my fault.

It wasn't my fault.

I tried to stop the… I sh- I,
maybe I shouldn't tell this story.

Brad: It's…

Nathan: have I told you this?

Like, uh, it

Brad: I don't know.

Nathan: three months in, and the
front end developer still on my

team didn't have access to GitHub.

And then it was about five months in,
and I was like, "Listen, UX director,

this isn't gonna work. Like, your
culture is, like, rejecting it. All

of its antibodies are, like, attacking
this," and all sorts of other stuff too.

paying.

I was like, "Stop paying us. Fire
us. Like, this is not working." Month

four, month five, month six, then the
UX director, the engineering director,

and the product VP all got fired.

Like, it was such a
bigger cultural problem.

Brad: Yeah.

Nathan: uh, fortunately, we
weren't left holding any bag.

Like, they were like, "This wasn't
your fault." Um, it was a bigger thing.

But, like, d- sometimes the culture or
sometimes the conditions that you find

yourself in don't afford you to, the
potential to actually succeed and create

Brad: Yep.

Nathan: hopefully you can short circuit
it before it wastes everybody's time.

Brad: Yeah, that's a, a telling
tale, what's hard though is

that a lot of times that doesn't
become apparent until hindsight.

You know?

U- until you're, like,
on the other end of it.

You're like, "Woo."

Nathan: Mm-hmm.

Brad: is, like, follow things that,
like, energize you or interest,

or, or interest you, you do need to
subject yourself to a bit of hard

or discomfort or, like, whatever.

Even if you're not, like, a political
person, you gotta, like, play that game.

Even if you're kinda more in the weeds
or you gotta get in there, you gotta

get that contract, you gotta, like,
do all of, like, that stuff that might

not be your wheelhouse, that's not the
main thing that's exciting you, but

it's like, it's a part of the game.

And I think that that's, that's a hard
thing for a lot of people to do because

it could be… It, it could… That
stuff can consume you, and then you

find yourself totally on a different
path than the thing that is exciting.

So it's like checking in with yourself.

It's, like, being very
intentional with that stuff.

Yes, I know that I'm going through all
this rigamarole so that we could do the

fun stuff or, or whatever, but some people
just kinda get pulled by a fast-moving

current at the expense of their own
interests and enthusiasms and curiosities.

So it's like, it's just, like, a
weird thing that you need to contend

with, and you really have to, like,
fight for and advocate for the

things that are exciting to you.

Or if you can't find that in a gig, which
has certainly been the case for me, I'd

love to hear if that's the case for you,
but it's like, to go, "Okay, this ain't

the thing that is, like, hitting all
of my buttons, but this little nugget

here, I didn't, I didn't think I was
gonna find this to be interesting."

But, like, sometimes it's, like, the
relationship between two people and I'm

like, "Oh, this is gonna be, like, my
fun for this gig. This is gonna be my,

my fun for the project that I'm going
to be excited about working on, even if

everything else is, like, kinda bland."

Nathan: Yeah, I, I really appreciate
some of the, the leaders I've been

brought in to help establish a design
system in their orgs, and the degree to

which enable, like, the people that are
gelling to find their way and form their

stuff, even if it's not necessarily the
vec- the direction, um, that, that they

expected things to go in, and they sort of

Brad: Yeah.

Nathan: other directions and end up in
other places, but that it still ends

up creating sufficient value in the
general direction that they expected

it to go, but they're hands-off.

Like, I, I really have been
blessed to, to have some of those.

and I mean, like, I even recall, like,
the project we worked on kinda together

recently where I wasn't available, but
it also wasn't the ideal project for me.

But I was able to say, "I am available
for this, and would it help me or help you

for me to just participate like X and Y?"

Like, "Do you want me to hang around
and give some input and be a little bit

more distant from everything?" And I
think it, for the very limited amount

I was able to participate, created
a little bit of a difference, but it

Brad: Yeah.

Nathan: wasn't necessarily
a cookie cutter thing,

Brad: Yep.

Nathan: okay.

Brad: Yeah.

Nathan: charge.

I wasn't, uh, tasked with any
particular domain, but that's fine.

Like, I'm still creating
a bit of value, as it

Brad: Yeah.

and you did.

You absolutely did.

But it is, it is good.

It's like there's, yeah, here's the
thing that, like, I've shaped for myself,

my, my offering, my normal thing, or
here's, like, the, the way I'd prefer

to help, and yet there's still the,
"Sure, I'll help you, load this couch

into a moving truck," or something.

It's like, like not something you're
doing every day, but you could still

be helpful in just weird, varied ways.

And I've enjoyed, like, being
that, like, amorphous puzzle

piece for different places.

It's like, "What, what are you
struggling with? Do you need a lot

of help? A little bit of help?"

Like, weird kind of helps.

Nathan: that's another piece of advice
too, is how can you put yourself in such

a diverse and rapidly changing array
of relationships and, and contexts so

that you can pick up all the patterns?

Like, part of what you get when
you consult with me is, like, I've

worked on over 100 design systems.

I've seen and I can feel the nuance
of the boundaries, and patterns, and

cultures, and, like, feature sets, and
architectures, and all that kind of stuff.

And so, like, I come equipped with
that, but that's also because,

like, I was drawn to work with
Brad and these other few people.

Even though I couldn't draw, I still
learned and, and observed and, and

saw how they interacted and how
they leveraged their own strengths

and, and all that kind of stuff.

And putting, uh, myself in a position
to be able to bounce between those

things, I think also really helped,

Brad: Yeah.

Nathan: be in

Brad: Yeah.

Nathan: learning one way with
a few collaborators for a

Brad: Yep.

Nathan: time.

That,

Brad: Yep.

Nathan: me, wouldn't have been good.

Brad: Yeah, yeah.

And it's good to, it's know thyself,
because you get a lot of, like, depth

and, and rich knowledge from going,
going deep and, and with, like, a,

a, a singular goal, and then there's
also real value in, in breadth.

And so, yeah, just kinda knowing
what gives you energy and what feels

exciting at that time and, and moment.

I, I too like the variety and the ability
to learn from- A bunch of different

people all at once that, that really
is helpful, and I've enjoyed that.

And so if that kind of breadth and
curiosity for a lot of different kinds

of experiences is something that you
crave, then well, then consultancy is a,

a pretty damn good way of, of doing that.

But then I, I also really admire, and
I'm sure that you feel the same with

the teams that you've worked with.

Some of these people have been doing their
thing at that same place for, like, 15

years, and you're like, "Holy shit." Like,

Nathan: there, there's a respect to

Brad: that's

Nathan: some of

Brad: what I'm saying.

It's like, it's like, it's amazing.

Like, I have, like, the utmost respect for
somebody that's like, "This has been…

This is my job.

Like, this is what I do."

Nathan: 2011's a long time ago, 15 years.

I can't imagine that.

But, well, actually, I worked
with an insurance company.

I can imagine 15 years.

Um, like, the, it's also that,
like, in the roles that you choose,

you have to know what you're
seeking and what you're not.

I'm not seeking being the
figurehead of the successful system.

Like, I've worked with some design
leads over the years that have worked

it for years, and they are the leader
that has taken this thing and grown

it and scaled it to that thing.

Brad: Yep.

Yep.

Nathan: to them, they're, they're
the, the hub of all that success.

I don't need that.

Brad: Yeah.

Nathan: uh, but it's g- I really
admire a lot of the people that have

made that kind of change within what
can be a really chaotic environment,

Brad: Yeah.

Nathan: for some people, that's the jam.

It's not my

Brad: Yeah.

Nathan: some, it's definitely the jam.

Brad: I think that there's something
to that, like, longevity in something.

In your case, a vast and varied
experience, but a longevity nonetheless,

that I think is, is going to also be
something we increasingly value in

this age where everything is immediate.

It's like, oh, when you find out that,
it's like, people have been on this ride

for, like, seven, eight, nine, 10 years,
and there's, like, a lot of reps in there.

I think that we're, we're going to
suddenly realize this, that value

in those, like, longer efforts.

That's, that's what ultimately breaks
through that just hype-y, noisy stuff.

It's like, nope, we've been focused.

We've been doing the work.

We've been iterating.

We've been improving.

We've been, like, thinking deeply
about this and through the lens

of a changing landscape over time.

all of that rich infrastructure and
all of those decisions and all of

that agreement and all of that, that
richness that comes with the territory."

So

Nathan: Yeah.

Brad: Here's hoping at
least, or we'll all be dead.

Nathan: Hey, man.

S-someday.

Someday

Brad: No, that is true.

it has been a real pleasure finally
getting a chance to, to like, kind of work

closer to you to like see how you operate.

It's like, it is really wild to,
to have this similar but totally

different style kind of to it.

And so it was really, really fun to
be able to, to work together and,

and work a bit closer and get to
swap notes and all of that stuff.

That was, that was really a treat.

And I also just appreciate your
longevity through all of this.

Like, the fact that you're able to
be here kind of carrying this stuff

through and, and still finding lots
of things to be excited about and

enthusiastic and interested in.

It's like, it's awesome.

So I'm, I'm glad that you're
continuing to have a good time.

Nathan: Yeah, you too.

Don't stop.

Brad: Yeah.

Uh, well, before we wrap up, uh,
the only other structured question

on the show is, is what music would
you like more people to know about?

Nathan: Oh, gosh.

Um, I am not a music aficionado,

Brad: That's okay.

Nathan: you know, like, maybe I'll
talk a little bit about Frostapalooza.

When you put on Frostapalooza in 2024, I
believe, in Pittsburgh in, in September,

was able to go because I live in Ohio,
so-- Or grew up in Ohio, so I went home

and then came back and, and hung out,
uh, at your gig and enjoyed the show.

But it made me-- It was among--
amid a time when I was starting

to learn how to play the piano
uh, not just learning music.

I've never done instruments before.

And so I haven't really d-dove in much
to it, but I still occasionally do, and

I'm playing the same song over and over.

it's called "On the Nature
of Daylight" by Max Richter,

Brad: Okay.

Nathan: pretty solemn piece that
once you hear it, it's in, like, so

many movie soundtracks, like Arrival,
that it'll sound very familiar.

I don't know.

It's, uh, like, since I'm learning to
play it, it's got an increasing level

of sophistication, but it also has
very recurring measures, and so on.

Brad: Oh, wild.

Nathan: but how it's, like, helped
me on my journey of learning how to

play piano because of how I've been
able to layer on the techniques as

actually the piece itself becomes more
and more complicated as it progresses.

it's, it's been, uh, a lot of fun.

But that and also the other, uh,
song was "Melancholy and the Infinite

Sadness," the classical sort of

Brad: Oh, yeah.

Nathan: the Smashing Pumpkins
double album from '96.

Brad: Oh, yeah.

Nathan: anyway, those are the,
those are the tunes I'm hearing,

Brad: Beautiful.

Oh, that… Beautiful.

That's incredible.

I love… I, I'm really excited
to listen to that as you describe

it as, like, it sounds like as the
piece unfolds, it gets more complex,

more nuanced, or there's more to it.

So it's like, it's like starting
at closer to, like, a 101 level,

and then, like, ending up in,
like, master's degree territory.

Nathan: Yeah.

It's a pretty-- The M-Richter
piece is pretty straightforward.

Like a, a pianist would be like,
read that once, and I'll know how

to play it." It's not, uh, "Mary
Had a Little Lamb," but it's not

also-- it's not Beethoven or Mozart.

Brad: sure, sure.

Nathan: it's-- it, uh, tone is, is good.

Brad: That's amazing.

Oh, that's exciting.

I'm really, I'm really excited to, to
know that you're, that you're digging

into that, that world, and, you know,
math, statistics, all of that stuff.

I mean, there, there really is so much
correlation between these things and, and

it is just kind of equal parts, art form
and, and technique, and there are rules

and, structure and rigidity in certain
respects, but also not, and it's, it's…

There's a really cool
game board to play on.

So, oh, that's really, really cool.

I'm, I'm really glad to hear that.

I feel like in, in, in some small way
helping just kinda, like, poke that, that

fire, poke those embers a little bit.

That's, that's awesome.

okay, and then last thing,
like, where could people follow

along with what you're up to?

Where, where do you hang out?

What would you, uh, like
to point people towards?

Nathan: Uh, I'm on-- mostly posting
on LinkedIn and writing on Substack

about, uh, the Specs tools that I've
been building, uh, which you can find

them, uh, on Figma, but also just
at Specs plugin and the, the Specs

repo that I publish stuff through.

me there, and let's, let's, let's
chat and keep sharing ideas.

Brad: Yeah, amazing.

Well, thank you so much for coming on,
and thanks everybody for listening.

So,

see you later.

Nathan: Thanks,