Ep #20: Christine Røde (The Browser Company): Dia's Morning Brief, Real Art, and Shipping Less

Tom and Jonny sit down with Christine Røde, designer at The Browser Company and the first guest under our new name, Mostly Working. They go behind the scenes of Dia’s Morning Brief, unpack why people now hand their data to AI tools so readily, and dig into the craft a designer keeps when she can ship her own fixes.

In this episode:

• Dia’s Morning Brief exists because most people stall at a blank text box, so the browser sends the prompt for you and opens a tab that briefs you on your day.
• Christine came back from two weeks’ holiday to a useful prototype that looked like ‘pure slop’, and rebuilt it around hundreds of hand-picked public-domain paintings from the National Gallery of Art.
• She designed Dia’s memory flows with lock icons everywhere, then watched user testing show that people were happy to tick the box.
• Arc shipped something new every week and ended up with around 35 branded features, whereas Dia builds and tests plenty and ships only what adds enough to the product.
• You get eight or nine nitpicks before an engineer has had enough of you, and Christine now ships those fixes herself the week after launch.
• A year ago she was asking her head of design whether she’d still have a job, and today she’s enjoying the work more than she has in years.

Chapters:

• 0:08 — Welcome to Mostly Working, and introducing Christine Røde
• 0:46 — From a Mostly Working meetup talk to Dia’s Morning Brief
• 1:21 — What Dia is, and how it has changed as AI has
• 4:08 — The blank text box problem, and a brief that arrives on its own
• 5:57 — The first prototype: beige, serifs and ‘pure slop’
• 7:00 — Finding the imagery: SVGs, big type or public-domain paintings
• 8:32 — Tom on what design adds to a prototype that already works
• 11:04 — Signal over volume: why the brief stays short
• 14:54 — Jonny on promising two or three things, and design reacting to engineering
• 16:37 — Why AI quality comes before design fidelity
• 19:01 — Onboarding, and persuading people to connect their tools
• 20:42 — From WhatsApp-era privacy to ‘you can all just read my email’
• 24:08 — Trust, and design’s part in earning it
• 26:45 — Designing Dia’s memory feature for users who weren’t worried
• 28:53 — What faster shipping does to the quality bar
• 30:09 — Why Dia ships less often than Arc did
• 32:01 — Designers shipping their own polish
• 34:12 — Working in seasons, and the hectic first month
• 38:24 — Custom tools for curating paintings and sketching interactions
• 42:08 — Real art as a brand decision, and hiring a painter
• 45:45 — What it means to be a designer now
• 48:17 — Quick-fire: free time after AGI, and Lime bikes
• 50:54 — The part of her craft Christine won’t hand to AI
• 52:09 — A rooftop garden, and the last 20% of vibe coding
• 53:31 — Final thoughts, and where to find Christine

Find us at mostlyworking.co

Creators and Guests

Host

Jonny Burch

Co-founder of Near Future

Host

Tom Harman

Co-founder of Near Future

What is Mostly Working?

A weekly podcast covering design and AI by Tom Harman and Jonny Burch. We cover both what we're seeing on the ground and industry trends, ways of working and occasional guests from the design world.

Tom: Hello, welcome back or welcome
to the Mostly Working podcast, the new

first version with our new name, formerly
known as the Near Future podcast.

Today we have special guest,
Christine Røde, joining us.

Christine has had, an incredible
career working at very exciting

companies from Meta, Deliveroo, Rekki
startup, WhatsApp, and most recently

and currently The Browser Company.

That journey has also taken you from
Norway to the US to London, lots

of exciting places in between.

So, Christine, thank you so
much for joining us today.

Christine: Thank you, Tom.

It's really fun to be here

Tom: So I thought we could, kick off by
talking a bit about, Dia's Morning Brief.

So you came to an in-person event that we
ran a few months ago now, Mostly Working

in person in London, and you walked the
group through some of the behind the

scenes of your process for designing,
creating the Morning Brief, which, as

I understand it, is one of Dia's most
successful features to have launched,

which is, an incredible statement.

Just to kick that off, I wonder if you
could just describe what the Morning Brief

is to someone who's not used it before.

Christine: Yeah, absolutely.

So I guess for a bit of context
before as well, so Dia is a browser

that we've been building for about
almost two years now, probably about

a year since we launched it publicly.

And from the beginning, we built this, as
AI was kind of becoming a thing, right?

So we were thinking from bottom
up, if we're building a, a new

browser in the current landscape,
how do you integrate AI?

How do you make AI like a part
of a web browser that can help

make your, like, daily life that
happens inside of a browser better?

And so I feel like our product has
really continuously evolved as AI has

evolved, which I-- a lot of people
knowing, listening to this podcast know

has been at a pretty incredibly rapid
pace, and it feels like every month

is different from the last in terms of
both what we think AI should be used for

and what it's capable of and what sort
of products people are building on it.

So when Dia started, our
biggest bet was really around,

like, is for getting answers.

We were using, you know, looking
at ChatGPT, and Claude was, like,

barely even a thing back then.

But it's like you're asking who was
president in nineteen fifty-five

and, like, those sort of questions.

Like, and so we thought, like, all right,
like, AI is a faster way of getting

answers to the questions you have.

And Google didn't yet have, like,

Tom: have like it's Gemini

Christine: Gemini box on the
top of every search result.

So we built really in a browser where
when you create a new tab, we route

you automatically to either, like,
Google or AI based on, like, what type

of query you're, you're inserting.

Then this has really been like now,
multi-year journey of constantly trying

to rethink as AI evolves, like what we
can do with this, how we can integrate it

into browser, like what it's useful for.

And something that I think has really
been coming up a lot, and I wanna say

like the last ten months, has been this
idea of, like, tool connections, right?

Like, that-- this time last year, we
weren't really talking about this, and now

everything is agents, everything is like
working with your tools and et cetera.

So we first started working on the version
of Dia that would incorporate agents and

tools and be built on the latest versions
of Claude back in, I wanna say like the

winter, like February perhaps this year.

And the problem that I think continues to
be a problem with all this is that outside

of like this fairly niche tech audience
where we're like all connecting our tools

and we're all using agents to automate our
tasks, I don't think like any mainstream

users are really doing this yet.

And most people don't really know how
to like get that value out of AI and how

to, you know, turn this into something
that can help you be more productive.

So the Morning Brief was our first
attempt at taking we knew were useful,

like the information we thought that
we could get out of the AI, but turning

it into something that's proactive
instead of something that you have

to like write your own prompt for.

Because this like prompt writing thing,
I think is really where we currently

have a huge gap in just like user
awareness and education of what to do.

Because when you give people a blank text
box, like most people don't know how to

get the value out of that So yeah, so
we were prototyping a lot of ways that

like, all right, like, what if we just
automatically on your behalf send this

prompt to, to the agent every morning so
that you just automatically get a tab in

your browser that briefs you on your day?

Like, can that make you feel like you
are the president, and this is like your

personal assistant coming and handing you
a top-secret doc that like, you know,

means that even if you're just tired
and it's eight AM and you didn't sleep

so well and you're like nursing your
coffee, you can like feel a little bit

more on your morning and what's ahead.

So around the time that we were
building this, and this was not my idea,

this idea of the proactive, brief.

I actually went on holiday, and then
I came back two weeks later, and in

that span it had gone from, like, an
idea on the wall to a first prototype.

But yeah, like, this like fr- this
prototype that we had running internally,

it, it looked like every Claude

Tom: Bye bye

Christine: you've ever seen
except like ten months ago.

So it was even, you know, less
polished than the stuff you

can get out of Opus these days.

It was a, kinda yellowish-beige
background, and it had like all the serifs

and the eyebrows and like just like pure
slop in terms of like the layout of it.

But people felt like the
content was interesting.

we have amazing, like, prompt engineers
on the team who had been, like,

refining this, and we'd been, like,
iterating on it a lot based on internal

feedback, like what felt useful, what
was not useful, how can we get like

the signal-to-noise ratio right here.

And I don't think we're ever going
to get it a hundred percent right

because everyone has very different
jobs, very different ways that they

approach it, different, things there.

But it just didn't feel right
because of like the sloppiness of it.

So I guess this is where I came in,
and I tried to like really reframe

this as like, how can I-- how can we
turn this Morning Brief into something

that feels peaceful and informative
and delightful every morning?

Something that is like subtly different
each day, so that if I have like

today's Morning Brief and yesterday's
Morning Brief both open at the same

time, they're visually distinct.

And how can I do this in
a way that's scalable?

Because I don't want it to just
be like pure text and typography.

I wanna have imagery in there.

But the LLMs also just aren't good
enough to generate images every day,

plus that would be incredibly expensive.

So like would, - Are we building more like
SVG illustrations that we can, generate?

Or are we trying to create like
typographic imagery, something that

like is like big and bold enough
that it feels graphic even though

it's actually just like a font?

Or can I like source
imagery from somewhere else?

Yeah, so those were all explorations
that I went into, and where we landed

was, finding a source for just fine
art with expired, copyright so that I

could curate a selection of hundreds of
paintings, primarily from Impressionism,

which is just my own, personal
preference, and, show you something

that felt like different every day.

Like there, there's your tasks,
there's your to-dos, there's all your

Slack updates, and, and things to
orient yourself, but there's also

this like little breather, and,

hopefully a beautiful piece of
art there to delight you every day

Tom: Yep.

Love, love that, Christine.

And I think there's, so many examples I
remember w- at the-- when you're walking

through this project in person, and this
is probably easier if you're looking at

the, the feature itself, because the
artwork I think really is really striking

and really s- draws your attention.

But I love this, the, the comparison
that you set up of being away on holiday.

Engineers had built this core
experience that sort of resonated,

but I think this is probably one of
the clearest articulations I've seen

of the value design can add, which is
taking something that, yeah, th-the

core experience was, was useful.

There was something, there was a core
baseline there that was really solid,

but your work was able to really
take that to, to the next level.

I'd seen a, a previous example of
this back in, back in the day, from a

typography teacher who talked-- typography
teacher and art director who talked

about, getting basically a, a plain
text file showing all the contents that

should be on the cover of a, of like
a glossy magazine, and then that's the

before and then the after is how they
make, how they sort of create their

magazine cover and just visually how
radically different they are, even though

the content is effectively the same.

But for me in that example, and
I think your sort of example is

something similar, which just really
highlights the, the value and the

power of visual impact of the UI.

But I also think there's something
really nice from h-how you talk

about it and the work itself, which
is more than just the visual impact.

It's the almost the state of mind
the design is trying to drive toward.

It is this sort of calm, peaceful
experience that you want to return

to rather than this technically more
useful experience, in air quotes, where

it might show more of your tasks and
it might feel more comprehensive of

the things you need to do, but that in
itself feels more overwhelming and more

like you're staring at banner ads and
you're developing this banner blindness

for seeing this sort of annoyingness
that is in your periphery all the time.

c- I'm curious if you could talk a
bit about the, the balance and the

trade-off between those things and, and
potentially any challenges you had in,

in sort of convincing leadership or,
or product counterparts or engineering

counterparts that the, a less useful
experience on the surface is actually

more impactful to, to the end user.

Christine: Yeah, absolutely.

I think striking that balance has been
an ongoing challenge, and it's still

something that we're trying to finesse,
like how can we get the right amount

of just value and signal out of this?

and especially because it is
such a personal experience.

Like we get feedback from some people
internally that say like, "Well, this

entire to-do list is useless for me
because I actually, before I wrap up

all my days, I make a really detailed
to-do list of everything I need to do

the next morning, so I already have
my own list, and it's better than

the one that AI generated for me."

And to that, I am just, you
know, I wish I was like that.

I wish that was the sort of organizational
mental structure that I had.

But, for me, the number of times that
the Morning Brief surfaces something

that I had, you know, just hours before
committed to doing the next morning

and then not written down and then
completely forgotten about is exactly

the type of value that I get from it.

But not everyone…

Some people, you know, are actually
like very functional people

that don't require that in order
to, to succeed at their jobs.

So we're never going to be
everything for everyone.

And I guess the main thing for me as
someone who is more type B in that

respect and who is not a morning
person, is that I already have, you

know, thirty unread Slack messages
when I wake up in the morning.

I have a Linear board that's full of
things that I need to be doing, and I'm

sure everyone has various other apps
with notifications and to-do lists.

And yes, you can like bring all
those things in together, and that's

probably like your inbox because
you probably have like notifications

from all these there as well.

But I don't think that there's much value
in replicating a, a dashboard that is

just all your to-dos in one because I
think it already exists to many degrees.

And I think that's like actually a deeply
unpleasant experience to be met with first

thing in the morning when you're trying to
boot up your brain and you're, to not get

super stressed out at what's ahead of you.

So I wanted to frame this around,
you know, being a slow start, giving

you like chance to like gather your
thoughts and like bring the most

important things to you, which I think
comes again to that like idea of the

President's Daily Brief aga-again.

The way that these are written is super
specific to be around like, you know-

The headline, here's what you need to
do, or here's what you need to know,

and here's like the decision that is
needed from you from-- based on it.

It's not just like, here's everything
that happened in the world.

It should be as short, as concise, and
to the point as absolutely possible.

And I think that's where the value is.

And that automatically means that
you have to, you have to curate

this, you have to cut it down.

We have to rely on the AI to be able to
sort through everything that's happened

and try and pick out the most important
things, which it will never be perfect at.

Although, you know, the more context
you give it, the better it might be.

But yeah, I think there will always be
this tension between like the people

who want like the, the comprehension.

They just want like absolutely everything,
and they think the more, the better,

versus what I would fight for is…

I don't know if I'm projecting myself,

Jonny: Is

Christine: but I'm just like,
you know, like normal people.

you

Jonny: Yeah

Christine: the type A, Silicon Valley
CEOs who are like on top of everything

in their life, but the people who
are normal workers who maybe like

appreciate a, a slower start or an
ability to, to parse they need to know

Tom: Sten

Jonny: I completely resonate with this.

And I'm also disorganized and
actually don't want to see

a mega list of everything.

And I sort of feel like as a design
constraint, making the promise that

you even would be able to do that
would be failing at the first hurdle.

Whereas if you're like, "I'm just gonna
give you two or three things to think

about," and it's a nice vibe and you've
got a picture and it's, it's, you know, as

you say, read it with your morning coffee,
is both nicer, well, in my opinion,

and I think objectively, but also much
more realistic as a thing to achieve.

especially as you say, because people will
then be spinning up Claude and Codex and

like all of the other things that they're
doing, endless chats that are fully

comprehensive and arguably talk too much.

So nice to have a little bit less.

I did want to, ask you about, I mean,
this like artificial, went on holiday

and therefore it got built without you
of the design process or the traditional

design process, meaning like you're
given something that basically works

and you have to react to and sort of fix
rather than something where you get to

be the first person thinking about it.

I'm just wondering how typical
that is what you're seeing.

It feels, like a trend, from people
that we've spoken to and definitely

the, the work, work that-- design

Tom: design work that has been done really

Jonny: that has been happening around us.

But like, do you typically, work much
more in lockstep but like as a reaction

to a piece of technical work that is
interesting as the sort of UIs get

a little bit more non-deterministic
and, reactive and things?

Christine: I think a little bit of both.

I think there's still plenty of projects
that happen very much in lockstep or

even, you know, very design-driven.

We do, do quite a lot of design-driven
work at The Browser Company that

really, like, happen with prototyping
and ideation and, like, that whole

very classic, process as well.

But I do find that especially with
things that are so AI-focused, like,

the first step is just AI quality.

It doesn't matter if I'm like, "I think
it would be super great if we could have

a Morning Brief that, like, showed you
three to-dos and then, like, three updates

from-- on your Slack and your email."

And it's like, cool, that's a good idea.

And I can design that in Figma.

I could have started there, but it's
like it, it's completely pointless

until we've actually validated that,
like, yes, we can pull this information.

We can, like, have the quality be
good enough to be worth shipping.

With a lot of those projects, I
honestly, like, I'm obv-- I'm not

on holiday and like, you know, I'm
also sadly not on holiday nearly

as often as I might hope to be.

But even then, I might prefer to like,
yes, like I'm a part of the process,

but I'm also sort of pushing people
that, like, I don't think that we need

to that much in like especially like
visual design, but even some of the

interaction experiences until, like,
we've validated that, like, this works.

Because actually been, like, plenty
of projects over, like, the last years

that I've been at The Browser Company
where we've had brilliant ideas that,

like, we know exactly what we want
out of it, and we can, like, mock up

what it should be like, and then--

but then, like, in action, we're
just, like, not able to get the

quality where we want it to be.

One example of that is, like,
anything that has to do with,

like, suggested prompts or, like,
suggested queries, I see, like, OpenAI

struggling with the exact same thing.

Like, the little tool suggestions
they show me are, like, never

really things that I want to, on.

And we've just not been able--
Like, it's like the concept is easy.

Like, oh, we just show, like, a couple
of, like, rows of, like, queries that we

think would be, like, relevant and helpful
and that can teach you how to use AI.

But in practice, like, none
of them are that interesting.

So that continues to be the first hurdle
and why, like, I think a lot of these

prototypes really do need to start at very
low fidelity and focus on quality first.

Jonny: Yeah.

I suppose the other unknown is also
how well you can persuade people to

onboard and hook up enough data of
their own to be able to get to…

Like, everyone internally at The
Browser Company has added all the

connectors and all of the things.

How likely is it that someone, a normie
would be interested in doing that?

I've created a number of new Dia
profiles over the last few weeks,

just new email addresses and new
computers and things like that.

Found that,

Tom: founder, like the first

Jonny: first time Dia, switching
over from Arc and using Dia the first

time and being asked this stuff,
I was like, "Why do I want that?

I don't get, I don't get
what the value will be."

And as maybe probably seeing some
of your work online, but also just

being like, I, I understand what Dia's
trying to do a little bit more now.

I'm now like, "Okay, I will add
Granola, and I will add Notion, and

I will add all of these different
tools just to see what happens."

Like it's, there's a little bit more
of a, like, trust building up that

gonna-- gonna improve my experience at
some point along the road, even though

I don't know exactly how it will.

I just sort of believe that it will.

So has, has onboarding
been a battle for you guys?

Like it feels like there's
been a lot of iteration.

Christine: Yeah, it's funny.

I mean, I almost wanna like

Tom: For the license plate.

I apologize for how advanced their

Christine: feel when you've done them,

Tom: thing

Christine: setting up new profiles and
stuff, and that is, all my work as well.

That's basically-- That was the
following segments after the Morning

Brief is like, all right, now we need
to get people connecting the tools.

If anything, the shocking thing
to me has been that has been more

successful than I think almost
like we, we thought it might be.

Jonny: Mm-hmm

Christine: I, you know, like I used to
work at WhatsApp, and couple of years ago,

like COVID, like a years after COVID,
privacy was like the number one focus it

felt like in a lot of the tech industry.

It was like what people talked about
It was what consumers cared about.

It was the way that like WhatsApp
framed its marketing like entirely

as it tried to expand into the US.

It's like, we know that you care
about your data being private, and

we know that, you know, everything
needs to be encrypted, and et cetera.

It just felt like that
was the consumer mindset.

And I think one thing that's been really
fascinating for me to see in response

to like AI as a, as a whole is that it
just doesn't seem like this is really

a priority for consumers anymore.

And I see that with like Meta rolling
back, like encrypted messaging on like

Instagram and like other platforms.

I think like a lot of that
marketing has been toned down.

And yeah, I think it turns out that
like if you can offer enough value

to people that like it makes sense to
connect your tools, are actually like

more willing to do so than I thought.

which I think is also like…

I think that's really
interesting as a trend.

I think it's very-- it, it's
nice to be able to like build

that amount of trust as well.

I think like, for example,
being acquired by Atlassian has

actually helped us a lot there.

We were able to like, build a lot
more secure systems, like build

a lot more like enterprise secure
tech with like their support.

But yeah, it's like a trend
in consumer software as well.

I feel like there's been like a massive
pendulum swing from like twenty-twenty,

like everything's encrypted, and I don't
want like any big company to have my

data, to like now that I'm just like,
"Ugh, Claude and ChatGPT and Instinct and

Muse, like you can all just read my email.

I don't care anymore."

Jonny: Yeah.

I, I

Tom: Ja

Jonny: like, the, the, the vestige
of control that we might have had

pre pre-AI, you know, even though
we didn't really like, you know,

Google knows everything about me, of
course, and has done for many years.

But this, this, like veil of
potential secrecy has now just been

completely blown out the water.

We have no idea who knows what, and
ev- ev- just maybe the assumption

is like, "Well, everyone just knows
everything anyway, so we might as well

just get the value out of it ourselves."

of we've, we've given up collectively,
or maybe just no one cares.

That, that could also be, be true.

But,

yeah, just like this, this sense
of we are assimilating into the

Borg now, so we might as well
just relax and get it over with.

Christine: I think it comes down to value.

Like previously, the pitch was
like, "Oh, I give you my data, and

I get maybe like slightly nicer
adverts on Instagram," which is…

I still kind of prefer that
because I discover lots of

nice products on Instagram.

But now that the trade-off is I give
you my data and, well, one, like

hopefully you're not-- no one's actually
like looking into my data, but two,

like I get like very concrete value
out of it in terms of being able to

like organize my day, like get work
done more quickly, and et cetera

Jonny: The value exchange is clearer.

Yeah, Yeah, yeah,

Yeah, I love that

Tom: Curious though, because I, I
agree with that sentiment, but it

still feels like trust is central
now as it was five, 10 years ago.

Maybe that's less about privacy and
more about feeling confident that the

quality of this product or the quality
of this company, can be trusted.

So maybe it's a, a spin on the same
overarching theme around trust.

I'm curious actually about
that as a, as a frame.

So if it's less about privacy now,
are there-- like, how do you think

the role of design plays in building
trust in a consumer experience?

And is there anything specifically around
the quality of what ships that you think

is more important now or The Browser
Company thinks is more important now to

continue holding a high bar for trust and
making sure that's clear to customers?

Christine: Absolutely.

I mean, I think there's
multiple levels of this.

There's obviously consumer trust and
s- consumer security, but then there

is, like, very much like the enterprise
level of trust and security, where you

obviously have very professional IT
teams, like, vetting this and looking

for, like, specific, you know, things
that I'm not an expert at, but like

SOC 2 and et cetera, as markers of
this being, properly secure software.

And I think, like, it's been very
helpful for us, like I said, going

for, like, "Oh, we're like this scrappy
little venture-funded company with like

ninety people, and we didn't really
use to employ a security person."

But like, you know, we all have…

Not that we didn't think that it's
important to, like, very much like

this is a primary focus for us.

Like, we need to put security, like,
first and foremost in order to, yeah,

like, be trusted by, by consumers and
enterprises and companies, because that's

the only way that we're going to be
able to have a successful product here.

And then I think, like, as you say,
and I think-- as I think you're

alluding to, that design plays a
big role in this as well, right?

That something feeling intentional and

well-crafted and, not buggy and
not vibe coded are all things that

are going to help have an effect
on how much I trust the software.

To be fair, I noticed that also even
with smaller tools that, you know,

we try out these days, I found myself
being obviously a lot more hesitant

to just put my own personal data into
things that are small startups where…

I mean, I've worked at plenty of small
startups where I know that you sl- throw

like a slash admin at the end of it, and
they can probably look at, like, any user

data that you've put in there, right?

And like actually having, having companies
where, like, these things are logged and

tracked and have, like, serious, like,
vetting processes and et cetera before

they can just go and debug any plain text
stuff that's been entered on a website.

It's definitely, like, a defining factor
for me in what software I choose to use.

And I think-- But one thing that, like,
we worked on quite a bit when we first

introduced some of these, like, first AI
features, like I remember when we first

worked on, memory, which like asks memory
and like ChatGPT and Claude and et cetera,

like it learns from things you type, it
learns from how you use the browser, and

it saves that to your computer locally
so that like, you know, it remembers

what you're working on and who you
are and et cetera between sessions.

When we first built that, we were
hugely worried that, like, this was

going to be a massive deal breaker
for most people this is your browser,

and it means that, like, we are saving
locally to your computer securely

and encrypted, like, things about,
like, you go on and the things you do.

So the first flows that I designed
around this, it's like so cautionary.

It's like the language
is really, like, clear.

As I learned from my Facebook days,
you put like little lock icons

everywhere because that always
makes people feel much more secure.

But we ran these through user testing,
and we found that people actually,

like, didn't really engage with it,
and they actually weren't afraid of it.

They were shockingly, like,
not worried about this actually

from like my point of view.

And I think like, yeah, like from
ChatGPT and et cetera, like this idea

of like memory being core to an AI
product has actually trained people

as like, of course I want that.

Of course, I'm getting so much value
out of like the AI actually knowing

about me that I will, will check
that box with no, no concerns at all.

Tom: That makes a lot of sense.

I've been having conversations
recently with a, a range of different

companies who are, I think, in this…

sort of going through this transition
of a world where the product development

process used to take a long time, but
it had the, these natural gates and

checks and balances in place where
it was harder to ship something of an

unacceptable quality because it just
took a long time to, to get out the door.

But now that you can get things much
closer to shipping to customers much

more quickly and potentially try out
or experiment with new features, it

does raise the question of how much
do we lower the bar, or where do we

lower the bar, or how does design
quality feed into holding that bar?

but and I, I guess my hearing you
talk about this, I'm wondering about

the, the trade-off that companies are
making around trust by potentially

allowing certain features to go out
earlier that might be rolled ba-back

or changed, or they don't feel as high
quality in the ultimate experience.

And I- I'm curious if you have a take on
that or The Browser Company at large has a

position on, like, how do you decide what
is good enough quality to ship widely, and

w- what is design's role in shaping that?

Christine: I think we actually
have a very high bar for this.

I think if anything, we've only
gotten much stricter about what

we ship and what we deem to be
good enough quality for launch.

I think back in our Arc days, Arc
being our previous browser, we had

a very different, cadence here.

Like, we honestly like, by hand, build
things and ship things all the time.

And, our users, especially at the time,
got very used to, like, new features

every week, which was definitely like
a cadence that we upheld back then.

But it actually meant that at the time
we were, like, shipping these, like,

half-baked ideas, like seeing what worked,
you know, maybe finding that it worked

fine, but it neither like, you know,
radically changed a product nor detracted

from it, so we left it in, and then we
went on building, like, another new thing.

Probably how we ended up a bit of like
the Swiss knife of browsers that had

like thirty-five different branded
features, which where everyone loved

a different segment of features.

So it was very different, difficult
to know which ones were like really

the main selling points of the
browser as well, to with Dia, where

we've really tried to be intentional
about everything that goes out.

And I think people have maybe noticed
that we, we ship things less often.

I see that on our Discords and
our Reddits and et cetera, that

people miss our old velocity.

But internally, yeah, it's just incredible
how many things we build and test and

experiment with, and then kinda deem that
it's just-- it's not worth it to ship

it, and it's not necessarily because the
craft isn't there, but it's just like the

value it adds to the product isn't, isn't
great enough to, to warrant its addition.

And I think, like, that's,
that's all the more important.

And just because, yeah, something is
quicker to build doesn't mean that

everything meets the bar of going out
either On the flip side, I found it much

easier to maintain like a quality of
design and craft with, with vibe coding.

Because in the past it was usually
like, okay, you're working on a

feature with an engineer, and you're
working on it, and you're working on

it, and you're like, "Hey, just like
a small thing, do you think you could

nudge that one four pixels over?"

And like, "Can we refine that
hover interaction a little bit?"

And, "Oh, when I like click on
this and then click on that, I

kinda wish the animation between
them was a little bit smoother."

I feel like you get about like eight
or nine of these like nitpicks before

you start driving your engineer like
absolutely bonkers and they don't

want anything to do with you anymore.

To spend them wisely.

And now I've found that like actually,
like the thing goes out and then the

next week I've noticed this thing
that's going to bug me forever.

I can actually just go in
and ship a fix to it myself.

I can polish that feature myself, and we
actually end up with a more well-crafted,

product because like the designer's
degree of perfectionism, can actually

make its like way all the way through

Jonny: I'm interested in, the
design process and whether it's

being pushed to the ends, like you
have your sort of e- even past the

point of release, you can then dip
in and, and keep being a designer.

You don't have to say, "Well,
we're onto the next thing."

You can-- You have autonomy there.

Also, design more important anecdotally
at the start of the process to gain

that confidence that the thing that
we're building is the right thing,

and it sounds like at The Browser
Company, there's like a, the super high

expectation that everything is justified.

I'm just really interested in practically
the process inside, inside you maybe,

but inside The Browser Company that
allows you to hit that cadence and

work on these slightly slower projects
and, and like carve out that time.

do you have to really advocate for that,
or is that just a belief from the top

and, and, and you get to just do it?

Christine: Bit of both.

I find that, like, so we work in
seasons, or quarters, if you will.

So projects tend to be like…

Or I guess focus areas tend to
be like three months at a time.

And I find that there's like a certain
rhythm to them, where as a designer,

usually like the first three to four
weeks of a season is incredibly hectic

and stressful and overwhelming because
we basically start a season, we have

a new, like, focus and new goals for
what we're going to hit, and then we

don't know what we're building yet.

We have a bunch of ideas.

We don't- we haven't validated them.

We don't know which one of them is right,
and we don't know what's going to work.

And I have to kinda like figure
out, like, what are we building?

What does it look like?

What are the ideas
that, we should pursue?

while I have like three to five engineers
kinda like, "Do, do, do, do, do."

Like, "When?"

Jonny: Waiting.

Christine: Right?

Like, you know, they need stuff to build.

Like, "What, what can I build?

What should I jump into?"

that usually means, like, a lot of
trying to figure out, like, what

are the low-hanging fruit that,
like, we absolutely need to do?

How quickly can I turn out a couple of,
like, directions and mock-ups for things

that seem like they are like, you know,
high certainty, like fairly low effort,

obvious wins that will, like, take me
very long or very little time to design,

just that I can buy a little bit of time
to, like, then start thinking about,

like, the bigger ideas and exploring the
bigger ideas, and that's like everything

from sketches to prototyping and, like,
getting, you know, a dozen people in a

room and brainstorming and et cetera.

So that's usually the time where I
feel like I, I worked long hours, and

I'm very stressed, and especially when
you're in that, like, phase of, like, I

just-- I don't know yet what we're going
to be doing this season, and it's…

I may never find out.

Like, what if I just, like, don't
come up with any good ideas, and then

I've just let the entire team down?

But then, you know, like, you g- do
more work, and you get more certainty,

and you end up with some directions
that feel like they're, like, kind

of useful, and then it becomes more
around, like, turning those things into,

like, like, solving the problems and,
like, designing the UI and then doing

all the assets that are required in it
and then working with the engineers to

get it built, and that's when, like,
it tends to slow down quite a bit.

And so that's more, like, the, the time
of the season where I feel like I have

more bandwidth to go perhaps, like, a
bit outside of, like, our roadmap like,

polishing, like, existing things we have,
thinking of new ideas that maybe, like,

no one specifically, like, told me to,
but now that my brain is, like, churning

on these in this problem space, they,

Like, tend to emerge more naturally.

So I think, like, that rhythm kind
of lends itself to a combination of

things, and yeah, like, during those
first few weeks of any new quarter, I

definitely have, like, zero bandwidth
to, to think about that stuff, but

then, as the projects roll on, you
kinda, like, find your own rhythm,

and you find some breathing space
to, to push for other things as well.

Jonny: I love it.

I love it.

And, really relate to the have a number
to hit and we don't even have a sense

of what we're gonna build towards it,
and the number needs to start moving.

Oh, not necessarily that seasons have
numbers, for The Browser Company, but

that

Christine: but

they, they do.

Jonny: of every quarter, where
you sort of have to, like, think

on your feet pretty quickly.

and then you buy yourself time to, like,
polish and to, to think more deeply later.

Yeah.

Is--

Christine: Exactly.

And it's like twofold, right?

Like any, any company is obviously
going to care about retention

and like, you know, growth and
like all those sort of things.

And there's like certain things
that you can hit to like try

to increase those numbers.

And then-- But you also
just need a good product.

You just need to like improve the
product, and that's like the, the

vague, scary space that, you need to
navigate at the same time that you're

like gunning for the obvious wins

Jonny: Yeah, yeah,

for

Tom: true.

So true.

Christine, I want to ask about tools.

We like asking about tools on the podcast.

I remember in that walkthrough that
you gave in person, there were a

few moments where you built custom
micro tools to help with different

aspects of the design process you were
working through for the Morning Brief.

One of them specifically I
remember being about curation.

You built sort of a tool to help ingest
these fine artworks and help you decide

what should be sequenced to come up next.

But there's also an example of a
tool that you created, I think it was

for sort of typography level design
details, and you were sort of tweaking.

You couldn't decide which font to use
maybe, and you were sort of exploring.

You'd created a custom UI to sort
of try different options, and then I

think you ended up deciding to ship
it to customers because it actually

felt, like, useful for everybody.

but I'm curious how, since this project,
how your thoughts have changed, if

at all, on where custom design tools
fit into your process, and especially

on that point of the weight of stress
at the start of a season versus

once you, you're up, up and running.

Where, where do t- where do custom
tools fit into that, if at all?

Or are they just a sort of nice
to have at the end when you're

sort of polishing the details?

Christine: Yeah, no, let's see.

I think, I think like for those custom
tools, they've mostly been when I

get to like a real drudgery part
of the process or where it can just

like help me iterate more quickly.

So I guess that inherently does fall
more towards other end of the design

process and the examples that you noted,
like the tool I built to pull in from

the National Gallery of Art because I
started with a manual spreadsheet and

found out by row three that this was
very slow and very impractical, and

that's exactly what AI is really good at.

And also the command line tool
saved me hours of doing that job.

but I mean, I do fi-- I do use, Claude
Code and ChatGPT in the process, and

we're actually currently at the very
beginning of, of a quarter, and I am very

much in the messy middle phase right now.

and even though they're not like specific
tools, I've been, I've been doing a

lot of sketching around like different
interaction models, different like

modalities, like, you know, without
going too much into detail, like if

there's like a dozen different ways
we could potentially like have this

interaction work, like how do I determine
like what the right interaction is?

And that's always something that
I've kind of explored with thumbnail

sketches and like then maybe like
little prototypes and things like that.

but this week I've been trying to
outsource quite a lot of it to, to

like Claude Code, which has been really
interesting to see that I can like take

these like thumbnail sketches that I
have of like, oh, you click here, and

then it expands like this, and ha-- like
rely on it to like turn this into like

a gallery of little like interaction,
like interactive low-fidelity things.

So that means that rather than me trying
to present thumbnail sketches to people

and like run them by it or spending
like hours like animating each of these

individually, very quickly like turn my
ideas into something interactive that is

easier to communicate of doing that job.

But I mean, I do fi-- I do use, Claude
Code and ChatGPT with the wider team.

So that's not really a tool per
se, but I think it's like a very

useful way of being able to speed up
parts of the process without really

losing the intentionality behind it.

It's just getting to something
presentable much faster.

Tom: Great.

Yeah, that sounds really good.

I think there's, this is sort of maybe
a, maybe it's another topic change, but

it sort of feels related in my mind.

I remember you saying about the decision
to use real art in the Morning Brief

as obviously built creation tools
around that, but it also as a product

decision was also a major brand decision.

And I think when we spoke before, you were
saying how The Browser Company just hired

a brand designer who also happened to be a
very talented oil painter and fine artist.

And I guess I, I don't know if this is so
much of a question as so much of a, an

observation of the idea that bringing
design in almost as a reaction to AI slop

has had this sort of counterbalance where
it's pushed hiring decisions toward people

who bring in a level of human craft that
just wasn't, either wasn't possible before

or wasn't a big priority for the company.

And I wonder if there's any sort of
anything in that realm that you're,

you could share more on in terms of
any other sort of craft, human craft

importance in the hiring profiles of
people you're looking for or, how

that weighs in the design process
and, and where people push that.

Does that make sense?

Christine: Yeah.

Yeah, I think it does.

I think it's like-- I mean, you're
touching on something very real,

which is that as AI becomes capable
enough to take over many parts of our

jobs, people we look for and the jobs
that-- the roles that we gravitate

towards are, like, people who can
fill in-- people who can do things

that, like, AI absolutely cannot do.

And the thing is that, like, I mean,
I keep seeing demos where it's like

Claude is actually a very competent
painter now, and it's like you give it

MS Paint, and it fills in pixel by pixel,
and it's, like, crazily impressive.

But I think, like, people still also yearn
for authenticity and, like, handcrafted,

items and things that only human can do.

So I think there is something, like,
deeply validating about, like, whether

it's, hey, like, this is a real Matisse
painting, or Pissarro painting and,

you know, like, maybe you've never seen
it before, and now you're getting an

art education at the same time that
you're getting your to-dos for the day.

Or just like, hey, like, a real person
painted all the graphics that we're

using from the marketing perspective.

That adds, … It makes
it more legitimate.

It makes it more authentic.

It makes it feel at least, like, warmer.

Even if, like, AI was able to create
a painting just like it, I think

there's something nice about things
that feel like they are still made

by humans, and the time went into it.

And it gives me, something that,
you know, is less easy to copy.

I have seen, like, a big trend
in people using, you know, like,

paintings with expired copyright
and all that sort of thing as, as

imagery, and I didn't invent that.

I mean, it's been a thing for
book covers for a long time.

But then being able to have someone
who's in-house who's doing this for

just us, like, unique artwork that
no one else has, I think, like, is a,

is a subtle differentiator as well.

Ian's doing an amazing job.

I think we've just done some
advertising campaigns that's,

like, all based on his artwork

Jonny: amazing.

Before we get into quick fires, I
would love to just get your sense of

what it's gonna mean to be a designer
ongoing in this world where we

have, like, this everyone's a design
engineer, we can build anything we want.

At the same time, people really
value craft and, and a human touch.

Like, where do we live?

Where do we end up living?

And this is not the first person
we've asked this to on this podcast,

and I s- you won't be the last.

But, really fascinated in your
take from what you're seeing.

Christine: I guess I'm, I'm optimistic.

I feel like I'm in a place
where I'm enjoying my job,

like, more than I have in years.

I feel capable of doing things in a
way and working in a way that simply

wasn't possible before, even as someone
who's always been, like, semi-technical.

And it hasn't always been like that.

I feel like it's been a bit
of, like, a rollercoaster.

A year ago, I was definitely, like,
in one-on-ones with my-- with Dustin,

our head of design, being like, "Oh
my God, like, do I even-- will I have

a job in, like, a year from now?"

Like, is it just-- are we
getting pushed out of this?

But I feel like that hasn't been the case.

I, I think we've seen quite a lot of,
like, people realizing the need for

designers almost more so as people
vibe code their own apps and hit,

like, limitations in terms of, like,
how good they're able to make it feel

and how professional they're able to
make it be, a skill set perspective,

which is different from the engineering
problems, and maybe we'll solve that.

I know there's lots of people working
on building taste for artificial

intelligence and, you know, like
pushing that to the same level.

I do think that there is, like, an
element of if we are to believe that

LLMs are purely, you know, taught by
things that already exist in the world

and trained on existing solutions.

And yes, like a whole bunch of this
job is, like, reiterating, like,

problems that have already been solved.

But there is also, like, need
for invention and originality.

I hope that that will just be more and
more of our role is that we get to skip

some of the tedium that's just like,
"Yes, this requires a bottom sheet.

Let me design the seventh bottom sheet
that I have done eight this month."

And, more focus on, like, the difficult
problems that maybe only humans can,

can think of original solutions to.

Jonny: I agree.

Tom: Yep.

Yep

Jonny: intelligent to say than that.

Tom: That seems like a, a great place to,
great place to wrap up the conversation

and move to some, some quick-fire
questions we like to ask at the end.

But, but thank you so
much for that, Christine.

Very, very interesting, very insightful,
and especially your perspective on craft

'cause I think of, of designers out there
who are strong at visual craft, I think

of you being at the top of the list.

So, really interesting to hear your take.

So, yeah, let's ask
some, some quick fire…

Yeah, of course.

So, some quick-fire questions.

So, I'll-- Jonny and I
will take it in turns.

I will start.

My question for you is: When AGI takes
over everything, what will you do

with all, with all of your free time?

Christine: Probably write.

Write or read and, probably watch a lot
of television, play video games as well.

But I would like to do more, more
writing, and I haven't quite had the

time to that, to the degree that I want.

Tom: So what kind of writing?

Christine: Fiction
novels, that sort of thing

Jonny: You feel like you've got one in

Tom: Hey.

Well, watch this space

Christine: Yeah,

Jonny: to

Christine: I have some ideas.

I have some drafts that have never gone
anywhere, and I feel like there's-- it

would be nice to, have more time on that.

If

Jonny: Nice

Christine: that are written by, by humans.

Jonny: I think more so, right?

By your, by your, by your definition,
like wants to read-- I'm not gonna

spend 40 hours reading AI slop.

No way.

No way.

I wanna read Christine's next novel.

Tom: Same, same

Jonny: My question to you is, what
is one product that you love?

And it doesn't have to be, a
digital product, of course.

And it can be something you
created or it can be something else

Christine: Lime bikes have really
changed the way that I get around London.

Things that used to take like 45
minutes, now I can like be at Oxford

Circus in 12 minutes without ever
stepping foot on the tube, and I feel

like that's a proper game changer.

So I'm gonna go with Lime bikes.

Yeah

Jonny: And the, the, the
constant mild risk of death, just

makes it all the more exciting.

Christine: Thrilling.

I feel like you never feel more
alive than when you're risking

your life at least a little bit.

And

Jonny: two buses on The Strand

Christine: Well, you know, I have my…

I would never, but, you gotta
take the risks within control.

Jonny: Yeah.

Amazing

Tom: my final quickfire question.

Actually, before we started recording,
we were talking about Adobe Illustrator

and copying SVGs into Google
Slides and into Figma, et cetera.

But I'm curious on the broader
question of how your craft is evolving.

What is one aspect of your craft that
you would refuse to give over to AI?

Christine: Really, I like the visual
design and the sketching and really

the design part of this process.

I'm happy for AI to fill in the
blanks and to push things forward

faster and build things for me.

I've-- I also like riff
on ideas with AI a lot.

Like, I might give it like, you know, say
we're debating what entry point to do.

I might like list like seven different
ideas I have for entry points and like

help it like chat through with me.

But I have yet to trust AI to
be like, "Oh, right, I need an

entry point for this feature.

Can you just like give me-- Can you
just design ten options for me?"

I feel like I still very much wanna
be in the driver's seat as far

as like forming the, the look and
the behavior and the intentions

behind all these product decisions.

Tom: Great answer

Jonny: I have one more, which is, and
I-- this is a bit of a cheat because

I know-- known each other for long
enough that, I see you building

stuff constantly even before AI.

But, is there something you're
building yourself right now

that you're excited about?

Christine: Well, I think in the, in
the real world, I've always had a big

knack or interest in interior design,
so probably my biggest project from the

last year has been around building a
rooftop garden and decorating my flat.

But I'm not sure that counts as a

Jonny: That

Christine: as an answer.

Yeah.

That's probably been, like, more
the primary focus outside of work.

I have, I have definitely
dabbled in vibe coding.

I have not successfully shipped
or seen any of it through.

I find that with vibe coding, the
first eighty percent is really fun

and really easy and really quick, and
then the last twenty percent takes,

you know, several full working weeks.

And I, I have enough with my, my
actual job to occupy most of my

time in front of the computer.

So, but yeah, if I, if I didn't
have a job, there's certainly more

things, but it's nice to spend
time outside instead when you

Jonny: Yeah.

Christine: when you get

Tom: Fun

Jonny: has been the summer for it.

Yeah, absolutely.

Tom: Great

Jonny: shall we

Tom: Thank you.

Yeah, let's wrap up.

Christine, this has been great.

Thank you so much for sharing your
wisdom, and just having a really

interesting conversation about what
this all means, and how you're--

how, how design is evolving.

Any-- Do you have any final, thoughts
you wanna share with any listeners

Christine: I assume that the audience
here are all maybe people who are

already quite interested in AI and
automatically drawn towards it.

But I guess I do want to push anyone who,
who isn't to, try out these tools more.

I know from like my own personal friends
who are not in tech, who are all using

ChatGPT for, you know, the simple things
but have yet to like incorporate it fully

in, in design work and et cetera, that
I do feel like there is real untapped

potential here for a lot of people.

And, if there's anyone, especially like
women and people who are junior in the

industry who'd want like any coaching or,
have questions about that, I'm also happy

to have people reach out to me by email.

Jonny: Awesome.

How do we find you on the

Christine: My website
is christinerode.com,

so that's probably the
easiest way to contact me

Tom: Amazing.

We'll put that in the show notes.

Let's wrap up here.

Thank you so much, Christine.

It's been fantastic.

And thank you for listening
and see you all soon