Product Powwow

kev_2_04-17-2025_153743: In this
episode, I talk to Anthony Davis,

who's MD of Design at SCM Comcast.

We talk the future of tv, fixing
the web, designing with and

for ai and much, much more.

As you'll hear, he's an incredibly smart.

Strategic, creative, and humble
guy, which helps to explain

why he's so highly respected.

Finally, Anthony mentions Sky Glass,
which is their high-end TV with elegantly

integrated software and services and
Sky Q, which is their setup box, which

combines satellite and broadband feeds.

So I hope you find this conversation
as wise and insightful as I did.

kev_2_02-28-2025_101048:
To the podcast, Anthony.

anthony_2_02-28-2025_101048: Thank you.

Nice to be here.

kev_2_02-28-2025_101048: we've known
each other for a while and most of our

anthony_2_02-28-2025_101048: I,

kev_2_02-28-2025_101048: we've
done together is very much focused

on design, but we've had a few
conversations over the years around

what's going on in the broader tv,
film, and media generally landscape.

I just thought it'd be good just to
get you on, just to talk about, what's

been happening in your world, and
where you think it's going, basically.

So could you just maybe begin by
just introducing your background,

how you got into, being a head
of Design First Sky, and Comcast.

anthony_2_02-28-2025_101048: I
fell into it, to be honest, like a.

careers happen to you rather
than you necessarily direct it.

So, I studied English at university,
which leads onto nothing.

so then I was then trying to figure
out what to do and I had a really

good friend of mine had met this,
he met this girl on a cruise ship,

a dancer, she's an American, she was
living in New York and he said he

wanted to travel over to meet her.

Did I wanna come along?

And we go on a big tour of the state.

So did that pretty much
straight after uni.

And she was working at one of the
very first silicon alley as it was

called then you had Silicon Valley
in California and Silicon Alley

kev_2_02-28-2025_101048: that.

anthony_2_02-28-2025_101048: in New York.

Yeah, this is one of the first like
agencies, digital design agencies.

So she was working there.

She was like an account manager for,
I can't remember exactly who, but it

was just this incredible roster of
clients like Ford and General Electric

and Coca-Cola and some big museum,
but they all had one page websites.

You could manage like six of
these clients, no problem.

And she was working this and it was
just really the, such an exciting.

Room to be in just the sort of
level of excitement and buzz.

And she was sharing this desk with
a guy called Clay Shirkey who went

on to become a big like professor at

kev_2_02-28-2025_101048: Yeah.

anthony_2_02-28-2025_101048:
NYU and he was saying, look,

this web thing is gonna be huge.

It's gonna be absolutely massive.

And also it's all gonna
be about social media.

I dunno how he was so prescient
about that because, people would

just, we just got animated gifs for
the first time and just start to do,

kev_2_02-28-2025_101048: one Zero, right?

anthony_2_02-28-2025_101048: Oh, really?

Early.

Yeah.

And it just started to be able
to do, layouts using tables.

I don't remember that.

But it's like an Excel sheet.

You'd have like columns and rows and then
you try and merge some cells together

to create a sort of rudimentary layout.

So that would just.

Begun.

So it was really early of just trying
to turn this into something that

looked kind of consumer friendly.

And he was already saying this is
all gonna be all about social media.

So anyway, that was, that
looked really exciting.

So when I came back to London
and I thought, okay, I'm

gonna get into this web thing.

so just joined, bounce around a couple
of startups and then ended up at

Razorfish, which was a big New York
agency that spread out into Europe

and just rode that through to the.com

crash.

And then ended up client side for
a bit and then back into an agency.

And through those kind of travels,
just figured out that actually what

I really like to do was products.

what I really like to do was the
thing that was on the poster, that was

the thing I liked to do, rather than
making the poster, rather than making

the advertising, the marketing, it
was making the actual product itself.

So.

yeah, I joined Sky in 2010.

they were clients at the time.

I was looking for what to do next,
wanted to get into, product and

the iPad had just come out and it
looked more like a TV than the phone.

And so they said, do you want to come and
have a, come and spend some time thinking

about what we could do with this device.

And so I joined for a
three month contract.

okay.

I'll have a look at the iPad, figure
out what we could do with that.

And then it ended up taking me
10 years to deliver the outcomes.

I dunno.

And they got massive value for
money for that, but it was a,

massive scope creep for sure.

so yeah, that three months.

it ended up taking 10 years to
actually deliver the answer.

so that's how I ended up in it.

I really just fell into it.

kev_2_02-28-2025_101048: Excellent.

What was the answer

anthony_2_02-28-2025_101048: The
answer ultimate was Sky Glass.

That was the answer.

so yeah, we, so the way we thought about
it at the time was really around 2010.

You had three things happening in a way.

You had these new devices emerging,
for the iPhone and the iPad and just

kicked off this whole supply chain
of components that basically allowed

you to get pretty advanced computers
inside small and smaller objects and

make all sorts of interesting devices.

So that was all.

Kicking off, I'm sure you remember the
time, the excitement around that space.

And then you had, the cloud.

so everything, started to be
able to put stuff into big cloud

servers and run stuff off cloud
rather than locally on the devices.

You remember Steve Jobs had that great
presentation where he showed how Apple's

gonna reorientate itself around the
cloud rather than around the Mac, to

run, trying to have to plug in your
phone each time you wanted to sync

your photos, and then your iPad was now
out of sync unless you plug that in.

And just the nightmare of that to saying,
here, let's put the cloud at the center,

then everything would be synchronized.

That whole idea, which, you had,
software as a service company, just

that whole, that was the big thing.

So you had all these devices,
you had everything set to run on

the cloud, and then you just had.

Internet speeds were just going
up and up, and then people putting

wifi in their homes the first time.

So then these, all these devices you
would then sucked to buy for your house.

You could then get 'em online without
having to plug them in with the cable.

So that was kinda what was
going on in the industry.

So we were, we thought really then
looking at these three things, what

would you then do in terms of television?

And really the big answer to that
was, this is fundamentally changing

how you're gonna distribute content.

You can now get content much more
easily out to all these devices.

So if you think about, say, trying to,
release a TV series or even a movie.

In 2010, it was mostly through, DVD,
there were some like on demand services,

but a lot of it was just via DVD.

Even like Netflix was
mostly DVD at that point.

And just how hard that was to distribute.

you've got this, you're a Disney,
you've got a new Toy Story movie.

It's really about manufacturing and
supply chains of DVDs and packaging

and then distribution and warehousing
and getting it into retailers.

And then customers, you
try and take home friends.

The box set or the Big Bang Theory
or something like that, it's

a, it was quite a big object.

You had to then lug home and
put it on a shelf somewhere.

So the whole thing was
just very difficult to do.

But you could see with these
three trends coming that.

If you could start to distribute that
stuff digitally, then that was gonna

totally transform everything really
around how content would get distributed.

And you can see that play out now, today
you can just distribute, whatever you

want in almost instantly in unlimited
quantities globally at the push of

a button, now Disney to release, toy
Story, it is pretty much a button

press and then they can get it out
globally, around the world instantly.

And that was the transformation
we could see coming.

So then it was then, okay, what sort of
device, how could you then change the

whole experience if you can see that this
is the wave of disruption that's coming.

so in, it was 2012, so we're just
starting the project that became Sky Q.

And there was a big debate going on
at the time, whether, the next set top

box, would it be via satellite, which
is the traditional way sky distributed

content, or would it be over the internet?

And this debate was raging back at two.

And a lot of analysis going on around
effectively, what do we think, take us

about three or four years to make this
product in three or four years time, how

much speed will the average home have?

How many homes will have wifi.

At that point, therefore, what
sort of product can you make?

And so once that analysis was going
on, we started to think about the

experience and we basically said,
look, if you're gonna distribute

via the internet effectively over,
over ip, over wifi, then you don't

necessarily need to make a set top box.

You could actually make a tv.

You could actually really
rethink the experience.

And that idea of them being able to, what
we now call like glass to glass, you have

the glass of the camera in the stadium
or the movie studio or the TV studio and

the microphones and you know exactly how
you want to capture that performance.

Like a football game or a,
some actors acting out a scene.

You know how you want to capture that.

you how you want it to
process it in the edit suite.

And then you trace that all the way
through the broadcast to a set top box.

And then it finishes at
this sort of HMI cable.

It goes into a TV and then you
lose all control at that point.

the TV manufactures all apply different
settings, different color settings,

different motion smoothing and sound is
all sometimes a hodgepodge of soundbars

and external speakers or tinny speakers
inside the, so the whole experience

kinda falls over at that point.

So we said, if we could then make
a tv, we could then connect the

glass of the TV screen to the
glass of the camera and really.

Then get the experience to come out
exactly as the creative intent of

the, producers or the directors or
the, the lighting designers, the

photographer, cinematographers,
get that captured perfectly.

Then we could do that.

And, that then meant, maybe that
would then mean, you could then

think about content coming via apps
rather than coming via channels.

And so we, in 2012, we basically
kinda mocked this all up and we

did it as a fake Apple webpage.

'cause at the time Apple was
like this, scary companies.

Steve Jobs was still around and, they kept
disrupting all these industries and they

were making claims that they'd, solved tv.

I think he said he cracked tv.

And we were like, okay, what is that?

And say they figured out actually this.

This way to now distribute
content of the internet.

Maybe, maybe this would be the
kind of TV they would make.

So we made this fake like product
page for an Apple TV and we had this

incredible, like amazing picture
sound microphones in the TVs you

could talk to, it had a camera.

So you could do things like video calling.

It had, cloud gaming, all
TV was coming via apps.

So we thought Netflix would
probably make a TV app.

We thought Disney might make an app.

Everyone thought we were
a bit crazy at the time.

There's Disney's not a software company.

There's no way that life
would make an app, but we kind

presented this whole idea.

And if you look, if you compare that
sort of Apple page to then the Sky

Glass page currently live on sky.com,

they're pretty much the same.

It's you could run 'em side by side
and it's almost feature by feature.

Almost like claim by claim.

the same thing.

But yeah, it bombed.

the idea completely bombed.

and we made a set top box in the end.

And to be honest, we
were right with the idea.

we,

kev_2_02-28-2025_101048: that?

What was that year again?

Just to

anthony_2_02-28-2025_101048: 2012.

2012, yeah.

So 2012.

Yeah.

So Sky Glass was like 20.

21, so nine years later became the thing.

But yeah, we were just
way off with the timing.

we were pitching this for 20, about 20
15, 20 16, which is when Sky Q came out.

So we thought, okay, that's, that
was the timing we were aiming at,

and we were just way off on that, so
I think it was right that, it didn't

get chosen as the product idea.

There just wasn't enough
speed in people's homes.

Not enough people had wifi routers.

Disney hadn't made an app by that point,
it was, you had to wait for like the

industry and the world to catch up
with the idea to get the timing right.

yeah, good idea.

Wrong timing.

but yeah, that was the, going back to your
original question, that was the idea, in

terms of joining, looking at this iPad and
thinking, okay, what might this then mean?

How content distribution is changing?

This, you could then make this
device, this object, and then the

whole way the content gets to the
screen would be completely different.

but it took, yeah, there's
such a, there's such a distance

between an idea and an execution.

It, that's the hardest bit always
I think of, product development is

just trying to get to a version,
one of anything is really difficult.

And so going from an idea,
there's loads and loads of ideas.

Ideas are the easy bit actually
delivering that idea is the hard bit.

And if you can get to a version
one, I applaud and admire anyone

who can get a version one of any
product actually out to market.

it's such a hard thing to do.

And then you've got that get into
version one, to version two, which is.

Even harder.

That's like the, that's
like super, like next level.

Oh my God.

Yeah.

To get to version two of
anything is very difficult.

so if you can get to that point, then
usually you're fine at that point.

Then it just, you just rolls along
and you've just gotta, it keeps

running for generations after that.

but yeah, so that's, that was the idea.

2012.

But it, yeah, it took effectively
like nine years to deliver it.

So yeah, that's why my three month.

Contract turned into like a, I'm here
15 years now, but turned into this

10 year period of arriving and then
take 10 years to actually deliver it.

kev_2_02-28-2025_101048: I could see
that was still an amazing journey.

But when you said the idea
bombed, it was a vision that.

eventually came to fruition.

What was it?

Was it held onto a vision,
all that time or, was it

anthony_2_02-28-2025_101048: Yeah.

kev_2_02-28-2025_101048: forgotten
about and then you had to dig it up

again and revive it and say, look, this
has got some relevance, or was it seen

as this is what we're raving for and
let's just track progress about when

it's gonna be technically feasible.

anthony_2_02-28-2025_101048:
What I've tended to find, like

with anything we've done around.

anything around, let's say sums of,
innovation, anything we've tried to

do around, like product innovation?

it's always been about the timing
really, rather than the idea itself.

It's you and it, in my mind, I always
think about surfing type energy of just,

you've got these waves and if you're on
a surfboard and you see the sea starting

to move and you think, okay, this is
it, there's gonna be a wave and you stop

paddling, and then it just peters out.

that feeling of just this, what looked
like it's gonna be a big wave, just

you go up a bit and then you just
come back down and then it's just a

flat sea again and nothing happened.

But you have to be, every now and again,
the wave does catch, but unless you are

paddling, then it'll just pass you by.

you'll just go up and
down and the wave will go.

We kept the idea, the fact we can still
dig out this, this fake webpage we make.

'cause we kept hold of that,
we didn't get lost somewhere.

We kept bringing that out and we kept
prototyping in the background and,

tracking stuff and, tracking trends and,
kept revisiting the idea and communicating

the idea and really just waiting for
that, that, that timing to happen.

And then you just feel it move.

within a big organization, you
just suddenly, there's a couple

of like small projects that you
think, okay, hang on a minute.

We could actually hitch
our wagon to those.

And they then accumulate maybe like
another analogy, a snowball, that kind

of, it just gets a bit of momentum and
then things start to, gather onto it.

So that's always seemed to work for
me is just to have That, surfing type

idea of just waiting and being patient
and just tracking every wave that comes

in and making sure you're paddling.

But, as soon as it feels like,
okay, that's gonna peter out.

'cause they usually do, then
just head back to where you were

and just wait for the next one.

But, you've gotta keep paddling every
time because then when it does catch,

if you are then ready and someone says,
okay, we think maybe we should be making

a tv, and we are like, ah, interesting.

You should say that.

We've got, here's a prototype,
here's a working model, here's and

then suddenly you've, you can get
a whole project running from that.

So that's really how it worked.

And that's worked a number of times
now for me in my career is just, you've

gotta kind keep these ideas alive.

You've gotta protect them, but then
you just gotta wait for the timing.

You just wait for the right
timing, the right wave to come.

and sometimes it's the
whole industry has to move.

you have to kinda wait for wifi,
put enough people to buy enough

wifi routers and get enough.

Speed into their homes and you gotta
wait for Disney to make an app.

And you have to be patient and
then find the timing and then.

And then just hope you've
caught the wave correctly.

kev_2_02-28-2025_101048: Absolutely.

so you've talked about that,
that basically media distribution

story of glass to glass.

What about what's happening almost,
maybe it's not in front of the

camera, but, you've talked a little
bit about what's been changing around

how media is actually produced.

I remember you telling me about how
quickly Disney could roll out those,

those Star Wars spinoffs with the
new technology that they were using.

Could you say a little bit more
about how TV, film, and other types

of media are actually being produced
differently now than, maybe 10 years ago?

anthony_2_02-28-2025_101048: Yeah.

So this was the.

This is the other idea we had back
around again round to 2012 ish.

so we thought, okay, it's all about
content distribution being disrupted now.

That's what's happening right now.

How you get content out to
people is now being disrupted.

And then you try and think ahead to
be like, okay, once that's played out

and this sort of, this TV idea is in
existence, then what happens next?

And so then it was like, okay, if
they're at the point where effectively

distribution is solved, it's not, there's
obviously still loads of innovation

and disruption happening in streaming
and you know how content goes out.

But the basic kind of
ideas are now in place.

Once that's solved, then the
challenge then moves to how

do you actually make content?

How do you actually, if you can
distribute whatever you want in a

limited quantities globally, then
the value then moves to, how quickly.

Can you make content in what
volume and to what quality level

to then make, take advantage of
this new distribution mechanism?

everyone's effectively got an entire
cable service in their pocket.

Now, in their back pocket can take,
as much content as the highest

quality you could ever make it.

It can just absorb all of it.

You're not having to make,
big DVD box sets anymore.

The entire, output of, say
Universal movie studios is now

available in your back pocket.

So now it's about the content production.

We could, we always thought, okay,
that's where it's now going to move

to in the next decade with the decade
we're in right now is, how does,

how do you actually produce content?

and just a really sort of simple stat
for this is, and I think this is right,

I probably should double check it, but
it's something like Hollywood produces

something like 15,000 hours of content.

A year, YouTube uploads
400,000 hours a day.

So that's like the transformation
that's now happened.

everybody now with an iPhone
can pretty much, they've got a

movie studio in their pocket.

They've got a distribution, global
distribution mechanism in their pocket.

They can take a video, hit publish on
Instagram, and then that's available

globally to, millions of people.

And so now, the content
has then just exploded.

it's just completely exploded.

Distribution is not
the bottleneck anymore.

The bottleneck is how do I make,
how do I make that content?

And so as we, kinda moved into this
decade, we were thinking about,

okay, what's gonna change there?

And.

What you started to see then was
a lot of interest in bringing

digital technologies into studios.

so how could you then, create
these sort of virtual studios?

seem to be like two ways of doing that.

There was, are these
currently augmented studios?

sky, for example, for Monday Night
Football, it's in this, it's in this

studio that's a, it's a physical set, so
you've got physical, tables and floors and

walls, and you've got digital screens in
there where you can, play out graphics.

But then the cameras are also
tracked in 3D space, you have all

these markers around the studio.

And then markers on the, camera
itself, they look a bit like, little

ping pong balls with these sort
of, reflective material on the sort

of reflective material you have
on your, like cycling jackets or

your police have on their jackets.

So you have those little reflective
materials and that allows the camera to

be tracked in 3D space so you can track
it, in x, y, z, up, down left, right?

And then any sort of,
rotation of that camera.

And what you can then do with that is
then align that with a game engine.

'cause game engines, again, a
technology that's just grown

hugely over the last few years.

Game engines are effectively like
TV studios, in inside a game engine.

You've got a 3D world, you've got
objects and lighting and cameras,

and you can basically, connect
together the real camera in the

studio with a pretend camera, a
virtual camera in this game engine.

And then composite the two together.

So in the Sky Studio for example, you can
spin the camera around to look behind into

the back wall, and it's literally a wall.

It's just a big black wall with,
air conditioning and electricity

cables, and health and safety notices.

When the camera moves around to that
though, you just switch the game engine

view and suddenly you've got this
virtual space behind you, and so then

the studio becomes this much larger.

I.

Environment and it's a mix of
physical sets and virtual sets.

And you can then do that then for, movies.

You could actually then say, actually
let's get rid of the entire set.

Let's get rid of the entire set and
let's just make the whole thing virtual.

So you have these virtual production
studios, and I've been in a few of

these now and some of 'em are huge.

Like I went to warn at Warner Brothers.

And it's where they filmed, house of
the Dragon, the Game of Throne series,

and, Barbie, the movie, very funny
talking to the team there because

I think they, they spent six months
doing House of the Dragon, which is all

super dark, Misty, interior of castles,
and it's all very black and dark.

And then they then switched to
Barbie for the next six months.

And they just said the
dissonance was just like so hard.

They have to, every now and again, they
have to just go and stand outside for a

bit just to try and reset their minds.

But yeah, you go into
these things, they're huge.

They're about, a couple of tennis courts
maybe in size, and it's just digital, big

digital screens, on all four walls plus
a ceiling that can then flip up and down.

And that has a full digital presentation
on it all running through a game engine.

So you've got Unreal, for
example, just running this thing.

And again, you put a camera in there.

And you film the actors, but you're
tracking the camera in space,

which means then the set around
them, the virtual set, the screens

can then also be moving 3D space.

So it's the weirdest thing to be
in there because you effectively go

inside, you inside a castle, and then
someone presses a button and then you're

flying in a plane over London docks.

You press another button, then
you're inside Barbie land.

But if you carry the like little, that
little, those little reflective ping pong

balls, you sit on the camera, you can
unscrew it off the camera and literally

just hold it in your hand and walk around
this tennis court and all the scenery

moves with you because it's tracking
you in space and it's just adjusting.

It's like you're moving a character
around inside a video game.

You're like inside a video game
just walking around inside it

and you know you can walk over to
see this big medieval courtyard.

You can kinda walk up to the
wall and then peer inside.

Like a bucket that's on the floor.

it's completely crazy.

But what that allows you to do then is,
you can basically, your actors rather

than having to fly 'em around the world to
different scenes or different locations,

you can just keep swapping these screens
over and then, you wanna do a sunset.

You don't have to try and
time it for 20 minutes.

Of, actual sunset.

You can run it for six hours
and shoot all your different

angles and do different takes.

So it is a it that, that was a big
thing that, that, that kicked in.

and people still, using that a lot and,
it's very efficient and it just means you

can get through, stuff really quickly.

And Disney, were using this
in, a lot in Star Wars.

they could then make Star Wars.

if you've got this unlimited distribution
mechanism, then you could, as far

as the iPhone's concerned, you can
make a Star Wars a day and it can

distribute that globally, no problem.

So then it's just a question of
can we make a Star Wars a day?

And, how do we get the cost down of that?

Like how do we then make
more and more Star Wars?

And so that's really what you then.

I have seen, and all of those tools,
because they're game engines, they

can't run on laptops, so then you, then
that gets really interesting then of

just, how much you can just do this
with just normal consumer equipment.

And YouTube now so much content,
look, has a very professional

qualities to it because these tools
are now in the hands of everybody.

So everybody can now start, making quite
high end looking productions, quite high

production, value looking productions.

so yeah, that's the
shift that's happening.

And then of course now we have ai, which
is the thing we didn't see coming, at

the beginning of this, even five years
ago, just did not see that, okay, AI is

probably also gonna be this huge thing
in terms of disrupting how content.

Gets made or what even is content.

kev_2_02-28-2025_101048: Amazing.

Yeah, I remember.

I.

Talking to someone I knew who ran a
film school, this is about 10 years ago.

And, his core market were kind of
people who are fairly wealthy, maybe

done a postgraduate degree, and they're
wanting to learn how to make kind of

documentaries or something like that.

And, and he was based in a, I
forget the studio Southwest London

somewhere, and he was opening up two
new schools, in Berlin and Manchester.

They were gonna be teaching exactly
the same content, but they were gonna

be charging about a third of the price
because, they weren't providing any

equipment, but in, in the original school,
they had racks of, camera equipments,

editing suites, all the rest of it.

with the new schools, everyone would
just turn up with a laptop on a, on the

latest iPhone, and they had all the kit
they needed, but the, the content that

they were teaching was exactly the same,
but the way you produced it, just to

actually, absolutely plummeted in cost.

anthony_2_02-28-2025_101048: Yeah.

And you can see now with, these AI models
now with, the video models now getting

believably realistic, so you can get, you
can see to a point in a few years time

where you could have an entire like movie
studio type experience on your laptop.

you can have performers, you can have
lighting and cameras, and you can

have more and more control over that.

And you can then basically
just say, okay, oh, I want to

create a TV series or a movie.

And one person could then potentially
do that just with a standard consumer

laptop and then distribute that globally.

that's the possibilities
that are now there.

And if you think back to 2010,
that's completely impossible.

Nothing was there, the distribution wasn't
there, the production tools weren't there.

When you get to 2030, it feels
like in that 20 year period,

that has completely changed.

And that's been the really interesting
thing, just these sort of two decades of.

Transformation.

And, it's, it kinda interesting
about quite how neat it was.

arriving at Sky in 2010 with the iPad.

I think it just launched, I
think it was that year or the

year before, I can't remember.

It was around about 2010
and then Sky Glass in 2021.

And then, that's the distribution
disruption now than the

production distribution because
distribution is now solved.

Now it's about production and just
that's the bit right now where

it feels most chaotic right now.

No one really knows what the
answer's gonna be by 2030.

I'm sure it'll be a lot more
settled, but right now there's

no one really knows, right?

Everyone's just experimenting crazily
and who's gonna be win out in this

and what the model's gonna be.

And what is it gonna look like in 2013?

No one really knows at this point.

but you can imagine in five years it'll
be a lot more settled and it's gonna be,

it's so interesting just seeing how this.

How this whole thing plays out.

kev_2_02-28-2025_101048: Yeah, the
story there seems to be largely

about speed and cost, right?

Reducing, reducing time to, to
make and the cost of production.

Is that right?

Are there any, gonna come onto
personalization, but there, are there

any other, ultimate outcomes there in
terms of the changes in production?

Is it all about time and cost potentially?

anthony_2_02-28-2025_101048:
So I think there will be,

yeah, some of that for sure.

because, that's always a, it's always
the big challenge in, content production

is, time and cost is always, it's always
the difficult thing you're playing with.

but I think that's almost like
taking the old, it was always

do with new technologies.

you just try and take the old way
of doing stuff and see how that

works with this new technology.

And then eventually people
then figure out actually.

This new technology enables
something different.

Like you don't just need to, you remember,
movies when they first came out, they

were just like, like theater, right?

They just put the camera.

In a seat, in effect, like a theater
audience, just static camera,

and then filmed a play in effect.

And then eventually they
figured out actually you could

move the camera and do cuts.

And suddenly, okay, we make
something completely different.

This is not about theater anymore.

This is a different thing.

so I think it'd be the
same thing right now.

It feels like a, you can see a
lot of the ai video people trying

to recreate, traditional TV and
film, partly 'cause I think that's

what the model's been trained on.

So that's the easiest
thing to then create.

But I think what would be interesting
is what the more native AI content is.

That would be the interesting, thing
that be the, what is this new form

that, that can then emerge from that.

And I think that'll probably
be the bigger thing.

So I don't believe, I might be
wrong, but I don't believe it'll be

about like, how can we then make.

Traditional TV shows and films using ai.

I don't think, it'll definitely come
in and you can see it already in some

films and TV shows and some games.

People starting to use it
just as a production tool.

but I think probably more interesting
question is, what is content?

What type of content is getting made?

What is that content and
how is that then different?

I think that would be the
more interesting thing.

And my personal view is, I think you
think about the, modern interfaces, they

are moving more and more toward video.

I think, because I've been doing this for
so long now, like I remember, you think

back to, being in New York in the late
nineties and people started to, lay out

stuff using tables the first time you'd
be able to get like layouts on the web.

And that was very much like a print skill,
it's like a print design, like traditional

sort of graphic design around, alignment
and spacing and typography and scale and

but applying those, that sort of thinking
and that has been quite dominant for

digital design now for, the last 20 years.

we've been largely using, we
used to use Illustrator, and

InDesign for doing webpages.

we literally were using prints and that
was, that has been dominant for so long

now, but you can start to see that change.

And we likes of, TikTok and
Instagram and YouTube like so much.

Consumption now is moving to being video.

is moving away from that sort of print
aesthetic into more of a, I think

more like a sort of TV aesthetic, more
like a motion graphics, TV aesthetic.

So I can just, I personally
feel like, AI will then just

allow more of that to happen.

So be able to just take any data and
make it feel more television, make more

lean back, less of the kind of, clicking
and swiping and tapping all the kind of

button pressing that we've had to do.

that's 20 years.

I think that will start to lessen
and I think the interface will start

to become more, sort of, skills
you get in like broadcast design.

then I would think that, 'cause I, work
in a media company, so you know, you you

do get a bit, blinkered, you start to just
apply your own world to something else.

if you think about it, the,
so many apps have now started

to move to being video first.

And you can see that's where
the it, the attention is.

so I think AI would just
start to accelerate that.

So then it's really, it's not really
so much about making TV shows or films.

It's more about taking this vast clouds
and clouds of data and then how do I

then consume that in a way that feels
less like print and more like tv?

kev_2_02-28-2025_101048: Or gaming.

It almost sounds like a gaming mode.

anthony_2_02-28-2025_101048: yeah.

Or gaming.

Yeah.

gaming is, yeah.

It's so interesting.

I.

working now, so within Comcast as
well as Skype, it's all one company.

Comcast is such a big cable company in
the US the internet, Comcast provides

like internet to almost like half the US.

So you really get to see what is the
internet actually being used for,

what is actually going down through
those cables to people's homes.

And it's predominantly video,
mostly sport, and also video games.

And those are the two big uses of the
internet in terms of like data traffic.

And even like last year were the records
kept breaking in terms of traffic, so we

had, like an exclusive NFL game in January
for on Peacock, and it was only available

on this, the Peacock app, so you could
only get it on the internet and that.

Drove like the biggest day
ever on the US internet.

Just that one, one sport game.

It was more traffic than it
ever been seen on the internet.

But then that record kept
getting broken through the year.

And I think in the end it was
like, I think it was November-ish,

there was two things had happened.

One, there was a, an update that had been
running for Call of Duty, big video game.

they've done this big update, been
running for two or three days, so a lot

of people were still downloading it.

And then there was another,
exclusive NFL game on, Amazon.

I think it was.

Those two things coincided on
the same Thursday night and

again, another record night.

So it's been, it's really interesting
seeing just how much, live tv, live sport

and video games are really the things that
are using this, the whole, this new, this

distribution mechanism we built up over
the 10 years that's, the primary users of

that and video games are just huge now.

you can see just, they are.

Big kind of cultural moments now.

And what's so interesting is, having
worked in the TV industry, there's

always this, you always hear this
thing people say oh, no one really

wants to interact with their tv.

TVs aren't interacted.

People just wanna sit on the
sofa or the couch and lean back.

And, that is often quite true.

But then you look at video
games and go hang on a minute.

Video games are huge.

And they're mostly played on TVs
and they're hugely interactive.

they are, really per second.

The number of like button presses
and clicks happening is huge.

so they're very interactive.

They're very lean forward.

the TV screen is neither,
interactive or non interactive.

It's both.

And it just depends what the
content is, depending on,

what that screen then becomes.

so yeah, I think you're right, like
that idea around video games and

be able to interact with these big
screens, these shared screens, that's

already proven to be really successful.

And now it's just a question with ai,
can you now reformat more content,

more data to make that available
in a more interactive way, more

personalized way on this big screen?

And will you want to do that because
it's a break from your phone.

It's bigger screen, bigger sound,
and it's also a shared screen.

that's the big difference.

trying to do, say, I dunno,
trying to say like book a holiday.

you try and do that on your
phone with someone else.

It's one of the most
frustrating things ever.

if the person doing it on the phone gets
frustrated because they're trying to talk

through the options and the other person's
asking to see the phone or distracted

not, or the person not with a phone is
frustrated 'cause they can't see you.

Say, just gimme your phone.

I don't wanna give you my phone.

And it just ends up with this sort
of like slight bickering around who's

got the phone, who hasn't got the
phone, and who's clicking on what.

See, what if you could get that onto a
shared screen where it's, big screen,

everyone can see it and you can do it
in a bit more of relaxed way, then that

could work for a lot of people that could
be like, okay, this is actually a better

way to, pick a hotel or pick a restaurant
or work out what film we're gonna watch

or go through the, ordering a takeaway.

there's nothing worse than me, the
person who's doing the ordering on

the phone for the takeaway, trying
to run through the options for

everyone who wants to order what.

it's a nightmare.

So again, if you get that up onto
a big screen, that might be better.

So yeah, it's gonna be interesting.

I think just seeing.

You know how this changes.

Is it, everything you used to do
say through, Google search, it

feels like more and more now you're
gonna be using AI to do that.

And, AI can then take the data that
comes back and reformat it as a podcast

or as a TV show or as a, a game.

Like, how do you want it, how,
what's the best way for you as a

person to experience this data?

Or you as a collection of
people sat around a TV together?

What's the best way for you as
a group to experience this data?

And then be able to then get that
experience kinda generated in, in, in

a moment that feels like that could
be where we're heading with all this.

But who knows?

it's so like experimental and

kev_2_02-28-2025_101048: Yeah.

anthony_2_02-28-2025_101048: there's
so much like churn and disruption

and no one knows at this point.

It's just, but there's a
lot of money going into it.

kev_2_02-28-2025_101048: we could
be tasking a, an agent to come up

with some holiday ideas for us.

And the way it presents it back
is this is a TV show that the

anthony_2_02-28-2025_101048: Exactly.

kev_2_02-28-2025_101048:
around and watches.

Yeah.

anthony_2_02-28-2025_101048:
Yeah, exactly.

this whole AI thing, it is so interesting.

It does just fit, having, being
through the whole, from the early

websites through to today, the web,
we, it is amazing what's being built.

It's amazing what you can now do.

Digitally compared to what,
what it was like in 96, 97.

but also it does increasingly
feel quite broken.

amazing experience is, it is,
there's a lot that could be improved.

the fact all these websites and
apps, they're all different.

they've all got different
navigation schemes.

Different ways of doing things, as
because you have to try and relearn with

each website and app, you have to try
and figure out what it, what it means.

Where do I go to find this thing?

You've got all the, particularly
the web feels very broken.

You've gotta dismiss a hundred
cookie popups it feels like.

And it's not clear is this doing
anything like to, if I accept all cookies

or deny all cookies, does anything
really change about the experience?

It doesn't, as far as I can tell,
nothing seems to really happen.

I just, sometimes I hit accept
all, sometimes I hit deny or

nothing seems to really change.

It's just this frustrating layer.

And then you get all the, do you
wanna sign up for a newsletter?

get rid of that.

do you want the 10% discount?

No, get rid of that.

Do you wanna refer a friend?

No.

Here's an advert.

Oh my God.

Try and get rid of the advert.

And then eventually you try and get
to the data and that's, you spend a

minute trying to just find that data
and then you've, it's got like an

incomplete, this particular website,
it's a very incomplete view of view.

Maybe you log in, maybe you don't.

Can you remember your password?

and the history of you is just
whatever that particular service has.

So it is all, your experience
is very disjointed.

All these passwords and you've signed
up to all these privacy statements.

You don't know what you signed up to.

So that whole layer feels
like a little bit broken.

It's this idea that you could have
this agent, you know this one place

that knows everything about you.

One identity.

It's got all your history,
it's managing all your priv.

He knows what you want, what you will
and won't sign up to in terms of privacy.

So it's, you've got this single sort of
privacy gatekeeper and you just do that.

you just say to that yeah, it's
exactly, you said just, I wanna

go on holiday next summer.

This is what I wanna do.

I'm gonna get, I'm gonna go to the
gym for an hour, come back to me in

an hour, lemme know what that is.

Or look, I just want to lower
my household bills this year.

I just, that's what I just
wanna do that and can you figure

out a good way to do that?

Can you go off and engage all these
websites and call centers and apps and

just, every month just get my literacy
bill down and if you need to come

back and ask me a question, do that.

But otherwise I'd just go for it
and just give me a report each month

telling me how you've got my bill down.

you can see that world is
starting to emerge now.

these companies are startups are
starting to try and put this new

layering that just says, look, I
think we can just make this hole at.

This whole web layer, we can just almost
disintermediate that and just get you to

that cloud of data much more simply, and
you've got much more control and it's much

lower effort and you can have it presented
back to you in whatever format you want.

That world could start to emerge
and you could just see consumers

just going for that, just being
like, actually I prefer that.

Do I wanna go back to say like in five
years time and that's happening, would

you then wanna go back to actually,
do you know what I'm gonna go off

to, I dunno, 10 holiday sites, I'm
gonna put in my details 10 times.

It's slightly different, different
date pickers and trying to work

out what the date pickers are.

And it's asking me how many people each
website ask me how many people and all

these, cookies and newsletters and sign
up to this privacy policy and accept,

if I don't wanna go back to that way
of doing it, I think people will be

like, actually I don't wanna do that.

I wanna stick to this other way.

kev_2_02-28-2025_101048: Yeah.

anthony_2_02-28-2025_101048: that's
really, it's really interesting,

just, it's hard to know what's
gonna happen, but it does feel

like, this thing we built over 20
years could just start to disappear.

There's no reason that it's,
it has to be around forever.

There's nothing that says
it has to be around forever.

I'm sure it will be, I'm sure we'll
still be making websites and apps in

20 years, but I dunno whether there'll
be the main deal, I dunno whether

it'll be the main customer surface.

it could have shifted
off into this AI layer.

But again, no one knows.

this is the interesting I.

Interesting thing.

kev_2_02-28-2025_101048: indeed.

Let's just go back to, so that, that's
almost like the completely personalized

made for you by an agent type of, content.

But let's go back to, something
that's, already happening, probably,

and it's more based around the
big or whatever kind of content.

that is becoming
increasingly personalized.

we already, we're already used to, sky or
Netflix or whatever, as a recommendations

based on our viewing behavior.

But you've talked in the past about
how, films or TV shows are gonna

a lot more tweaked to profiles
or preferences or what have you.

could you say a little bit more
about what happening there?

anthony_2_02-28-2025_101048:
I think a lot of it is around.

yeah, just around accessibility to just
making, there's a lot of production

efforts to create like subtitles for
example, or audio description for shows,

and being able to do that at scale.

If you're doing 15,000 hours a year
from Hollywood, then that's doable.

it's hard, but it is doable.

But if you're doing 400,000
hours a day, then the production

effort's just too great for that.

But this is where, these AI tools could
then potentially come in and just allow

you to then have subtitles in any language
or have audio description in any language.

and then it's just a question of
could you go, would there be an

interest in, go further beyond, say
When you get those, warnings at the

beginning around say, there may be
strobe effects or, loud noises or,

it could be even like an age rating.

This has got, swearing or,
talks about adult themes.

you could, it's possible you could
actually say, okay, if you want to,

can you give me an edit where that
stuff is Then just edited out, or

suppressed in some way or altered
in some way so that, I can watch it

without having the stroke effect.

So I think in that kind of level of just
be able to just do a little bit more

customization around personal preference.

we're also seeing a
little bit, for example.

Sometimes, two people watching
it, again, it's a shared screen.

Different people have different needs.

So sometimes there's, someone
wants to listen in to headphones

with, dialogue boosted.

'cause they've got, something different
level of hearing to the person next

to them who's actually just wants it
coming through the TV speakers and

wants it full cinematics round sound.

be able to then, con control
that and these different

experiences of different people.

I think that's the sort of
thing that will start to come.

and then the other bit is then
kinda interesting just the, that

main experience, how that then.

Also starts to disaggregate
and splinter out.

So you see that
particularly around, sport.

but also I think quite a lot around
movies now as well, where, so

much of now is like clips, and so
much now is about, reactions too.

the whole like, goggle box, but
everyone is now goggle boxing.

just that, that idea
of here's me watching.

Wicked the film or doing the, dance
routine or singing the song, or here's

a clip from that film, or here's
the, interviews with the star, like

the whole ecosystem, the event that
sits around the actual film itself.

that with sport, so much of sport now
is around what happens after the game.

All the reactions, the clips, the virals,
the memes, that builds up a whole, like

that's a big part of the actual match
now, actual, 90 minute football game.

It's actually what goes before
it and around it and after it.

That's now the experience.

So that's something else
that's changing more and more.

kev_2_02-28-2025_101048: Yeah.

So something that Nick Hoon, mentioned
in a, a recent podcast I did with him,

talks about what's, the implementations
of Bluetooth ALI means in future TVs.

It will enable, different
people to access different audio

channels for the same video.

so there's also that layer
that's en enabling that kind

of customization as well.

anthony_2_02-28-2025_101048:
Yeah, exactly.

kev_2_02-28-2025_101048: And just from
a UX perspective, what's your thoughts

about how much sort of transparency
and control to give users about over

that, that level of customization?

'cause a lot of the customization
that seems to be happening

at the moment is almost.

J just picking up on behavior or
something about the profile rather

than actively giving choices to people.

Do you think that's the way it was?

It, have we learned that people don't,
even if people are given a choice to

customize, they probably don't bother
that much and they'd rather just be given,

options based on previous behavior, or
do you think, it depends on the type of,

personalization that you're being offered.

So if it's age spec, age sensitive
content, if you're logged on a kid's

profile, that just happens automatically.

There's obviously no choice.

And if it's hearing related,
you obviously do want a choice.

is it, do you see it very much?

governed by the type of personalization
that's being offered, or do you think

there are any general things you could say
about, transparency and levels of control?

anthony_2_02-28-2025_101048: Yeah,
I think what I've always tended to

see is people don't customize as
much as you think they're going to.

That's always been the case, and it's
always you do end up talking about

this quite a lot inside product teams.

There's always, someone will say,
let's throw that in a setting, 'cause

it's always the easy way to do it.

okay, we've got this feature or yeah.

do you want, yeah, particular subtitles,
just, let's just throw that into the

settings and you end up with just these
giant settings, experiences that no one

can use or figure out where it all is.

and yeah, I've even found, yeah,
you give people say used to, there

always be a thing on webpages where.

Okay, we're gonna have this customized
button and you can then change

the navigation to how you want it
and lay out stuff how you want it.

And very few people do that,
or if they do it, they tend to

do like one, one or two things.

I don't have any kids, so they
just get rid of the kids section.

Or I love sport and they put sport at
the top, and half the people do that.

They'd be like, we could've
probably figured that out through

data and just done that for you.

And you've been happy with that.

So I think a lot of the time
customization is just, is a little

bit of an easy way out for the
product teams more than anything.

'cause it's just an easier thing to do.

but customers, users don't tend
to use it, in my experience.

is that your question or?

kev_2_02-28-2025_101048: Yeah.

anthony_2_02-28-2025_101048: Yeah,

kev_2_02-28-2025_101048: my, my
guess is that, five or 10% of users

probably use those options and as a

anthony_2_02-28-2025_101048:
I think it's lower.

Yeah, I think it's, yeah,
it's usually yeah, maybe five.

Usually lower.

Yeah.

But very few people use them.

even think, say like your,
iPhone, like you arrange all the

icons how you want, have icons.

But if you look at most people's
phones, usually their first

screen is usually reasonably.

Organized.

Then after that, it's just a random
dumping ground of stuff, And as the,

as it gets simpler, where you just
say what you want, you just talk

to Siri, just type which app you
want or what does you want to do?

Then you just, do you know what I mean?

You just do that instead.

Again, I think this is where it could
be interesting, this whole AI thing

could come along where, you, it's
just much more, feels much more human.

You just conversation, just saying what
you want and then you're getting it back.

You're not having to go through
these setting screens and menus

and pages and navigation trees
to try and find the thing.

which is that's what we've
been doing for 20 years now.

so I think that's, yeah, I
just, this is what I feel.

I feel this thing coming.

'cause it just, everything
you're kinda seeing from how I.

watching users in user labs endlessly
over the years, just seeing how people

do and seeing the data, seeing how people
do things, it just feels like that way

of doing stuff is a bit more in tune with
what people want, And, it could just be

this, first 20 years of the web is really
just that's all we could do at the time.

Really, it was, you look back on that
20 years time we look back on that and

you're like, oh my God, that was, do you
remember all that clicking and scrolling

and swiping and typing and tapping?

We used to do anything.

it feels, it's like looking those old,
computers that all, you know, loads of

buttons and punch cards and it, look,
they look so industrial and mechanical.

It could look like these interfaces
we've made could look the same

thing was just, look how many
buttons were on the average screen.

It was like.

Hundreds of the things, it's
like this giant switchboard.

You have to click all these things and
tap all these things to get things done.

And that might just end up looking
like completely dated in a few years

time because I don't think that
feels natural to a lot of people.

you can see, you can just
see it in a user lab.

You can see the frustration people have.

Just trying to work out what that
icon means, or I've clicked that

and it didn't do what I expected it
to do, or where's that thing gone?

There was a button and then it's moved.

I'm sure it was there.

Or maybe that's another website.

I can't remember now.

You know that whole world
is just frustrating.

kev_2_02-28-2025_101048: let's
talk more about how, what people

are doing with TVs in their home.

people are well, aware of kind of multi
screening about people just watching

TV and having the laptop and the phone
and their tablet out at the same time.

there's a bit, been a bit of a
transition of watching YouTube

on phones on onto the actual tv.

There's you also talked about, the
frustration of sorting out a takeout

order or a, holiday or something and
passing around phones and maybe projecting

phones onto TVs and things like that.

what kind of changes in consumer
behavior are you seeing around the tv?

anthony_2_02-28-2025_101048: So I think I.

Ultimately it's, it's just moving
to more interactivity, it, that's

just been the trend over the
last, I don't know, 15, 20 years.

I think games consoles have
been at the forefront of that.

they've really driven just using that
screen for a lot of interactivity and

you're now starting to see that come
down into more like casual gaming.

So we're seeing a lot more
casual gaming on the tv.

And like Netflix is like to move
out into that space as well.

Running casual games on the tv, so
it's not the high end, time consuming

AAA classic video games, where you
need to invest hours and hours to of

gameplay to learn how to do it and
then to master it, these are much

more coming come out type games.

so we're seeing more of that.

And then it's really then just.

What e extra capabilities do you then add?

we put a, we created the, camera
accessory for Sky Guys called Sky Live.

again, that was in the original
like mockup we did of the

Apple, fake Apple TV in 2012.

We had a camera on it.

and so yeah, we launched that
a couple of years ago and,

using that for motion gaming.

having, like the old, we being able
to, stand up in front of the CV and

move around and jump around and be
more, more physically active and

you get a different type of game.

It's a sort of fun party type experience.

but also that fitness, we, we, got this
great fitness app that does, uses the

camera to, and AI to then track you.

so you can then do rep counts
of, how many press ups have

you done, or correct your form.

just, you need to, you need to move
your, shoulder, more of your elbow.

it can give you that feedback.

So it's like having a personal trainer.

so people are using that now, So doing
fitness at home, what more gaming at home?

video calling, just more and more in
interactivity and it definitely got a

bit of a bump through lockdown because
people, were at home a lot more.

So it's to do, a lot of other
activities start to come, a lot

of, so many people like, work from
home now a few days a week or.

Homeschooling and people just got used to
that idea of, more activities in the home.

And the TV is just a
great screen for that.

It's just big large
format you can share it.

so yeah, we're just seeing that
is, that's the kind of macro trend.

Still people are watching tv,
there's still that lean back.

people sat on the sofa experience, there's
still a lot of linear TV getting watched.

people still, sit down and watch strictly
as a family, live, that is still an

experience, so there is, it's very slow.

these graphs moves very slowly.

You always think, okay, this
is, like you, it just slowly

moves from one to the other.

It takes years of just
behavior slowly, adjusting.

But yeah, the TV still, it's,
it is definitely not going away.

I think, as YouTube is saying, it's now,
now becoming their most popular screen.

there's something about that big screen
format, the big picture, the big sound.

Be able to share it with other
people that, it is compelling.

And as that TV just becomes more and more.

Capable thanks to, be able to
stream data onto it now, and you

start to add more sensors to it, be
able to connect it with the phone.

as, Nick was saying about Bluetooth,
more these technologies that allow

to connect these devices together.

I just, you can just see that
screen continue to be a big deal.

They're getting bigger as well.

that's the other thing,
other obvious trend with TVs.

They're just getting larger and larger.

you're always just now 77 inch TVs are
now becoming almost like standard now.

they're just, these things are
just getting bigger and bigger.

People just want the biggest
TV they can fit into the space.

That's the basic rule of thumb.

kev_2_02-28-2025_101048: And what
about co watching, people watching

the same show in different locations?

Is

anthony_2_02-28-2025_101048: Yeah.

That's the,

kev_2_02-28-2025_101048: off?

that became a thing around lockdown.

I remember, but has that
become a big use case?

anthony_2_02-28-2025_101048:
it is definitely a use.

I don't know, it's a, Big use case,
but it's definitely a use case.

and we have a whole service that we
have a watch together service for that.

So you can then, watch, separate
households and then we'll just sync up

the content because that's always been
the big issue, is that the content gets

out sync slightly, the experience off
it's ruined because, you're cheering for a

goal that hasn't appeared on the other tv.

So making sure it's all synced.

Or if someone presses pause to go
make a cup of tea, then it pauses

everyone, who's watching in every house.

So we call all that working.

so yeah, that's, it is probably, I
wouldn't say it's a big use case,

but it is definitely, popular.

But I think that what you're seeing as
well is a lot of slightly asynchronous

kind of co watching where people will.

Watch stuff maybe independently, but then
immediately there's then a whole social

experience builds up around that on
social media, so there's the whole like,

the whole spoiler, avoiding the spoiler.

Like people that also drives
a lot of kind of audiences to

view stuff around the same time.

Because you want to be in that moment,
you want to get your reactions to

that, posted up onto social media.

You don't wanna be like two days
late with your reaction to what

happened in that sport team or the,
who got voted out of love island.

you what you need to be in
the cultural moment for that.

So that's the other bit of co watching.

It doesn't necessarily have to be
like we're all literally watching

the same video at the same time.

To the, millisecond that is a thing.

But maybe the bigger thing is just as a
society, there's a lot of kind of cultural

currency around, around television and
just being in that cultural moment,

that I say at the Super Bowl, it's on
a few weeks ago now, you know that.

the halftime show, like all
the kinda reactions to that.

That's very much of the moment.

And it's all died away now.

if you wait two weeks to watch that and
then come out with your, here's my take on

Kendrick's halftime show, it's too late.

That moment's come and gone, you know?

kev_2_02-28-2025_101048: how do,
does Sky Comcast get involved that?

stuff that's going on.

WhatsApp or Instagram or what,
however, people are chatting about it.

And do you deliberately, for example,
release little shorts that people can use

in that, interaction on other platforms?

Or do you just let it happen?

you are providing the content
and the chat happens elsewhere.

anthony_2_02-28-2025_101048: Yeah.

The, yeah, all the social
platforms are really important.

yeah, they're really important channels,
so a lot of content goes out onto there.

sky Sports has, big,
Big platform on YouTube.

but you, sky News very popular on TikTok.

It's a very important channel, to get
news out there because, you're just trying

to reach the audiences where they are.

so you often have to reformat the content.

You have to start thinking about vertical
video rather than horizontal video,

and you have to edit it in a different
way and put different graphics on it.

Think about people, watching
it without audio, which happens

quite a lot on social media.

So yeah, it's a different, it's a
different thing, but, a lot of the

production, setups now you go into
the galleries as they're called,

it's these big NASA like control
rooms where, they, are managing the

production of a, like a live show.

you'll have the desks in there
dedicated to the social channels.

you're putting stuff out to
social channels as well as out

to classic broadcast channels
as well as getting it ready for.

catch up on demand services.

it's just these things are so many
desks now just putting it out into

all these different formats, because
that's where the audience now is.

They're now out across all these screens.

'cause again, the
distribution is so easy now.

you just, it's just that production.

The production is now the challenge, like
how big can these edit galleries get?

How many people can you have
on hand to run all these

tools and get content out to.

All these devices
scattered around the world.

Yeah,

kev_2_02-28-2025_101048: Yeah.

Great.

let's move on to how your insect
integrating AI into your practice.

Your, how is your team AI now
and how do you see that evolving?

anthony_2_02-28-2025_101048:
so we're using AI not so much

in our, own workflows yet.

the tools aren't quite there yet for us.

we've been, I think we've been
using AI more in the customer

experience for a long time now.

we've been using, voice for example.

Using machine learning AI
for well over 10 years now.

so you could just say whatever you want.

It's a voice remote or use a wake word to
wake up a TV and without the remote and

search for something or control something.

We've been using that for a long time now.

And we also use AI for, for example,
for creating, simple kind of edits.

so, if you say for example, if
you join a football game, like 20

minutes in, you want to catch up.

or, a golf and it's an hour
in and you wanna catch up.

We then use ai.

We'll then basically
create, editable highlights.

So you don't need a human
editor to do that for you.

The AI will be watching the game.

We'll understand this is a goal, this
is yellow card, who the player is, and

it'll then create that clip, put all
the text around it, this is the red

card, this player at this timestamp, and
put that into a, an interface for you.

So you can just come in at 20 minutes
and just go what have I missed?

Watch the main game and then just
catch up on the highlights to one side.

and then we even did in for the Olympics.

last summer, then same idea,
and then added in, a narration

on top of that as well.

So he had this, very famous US
broadcaster, took his voice, trained a AI

model with his voice, and then you could
basically say, okay, I'm interested in

these four sports for the Olympics, and
just gimme a highlights package each day.

And this AI would then go and assemble
the highlights package for you.

It would then add his voice.

It would then create a script, a
commentary script in his style.

kev_2_02-28-2025_101048:
AI would create that.

anthony_2_02-28-2025_101048:
The AI would do this

kev_2_02-28-2025_101048: Wow.

anthony_2_02-28-2025_101048: and
then take his voice, and his voice

would then read out the script as
the kinda script he would write.

His voice would then read that out.

It'd say, Hey, Kevin, here's
your, summary of the Olympics.

Here's what happened yesterday in the,
swimming, fencing and triathlon, whatever

the sports you said you're interested in.

And, the voice would then talk
you through what happened.

And then you get this like little,
like three minute sort of highlights

package and you get like hundreds
of these, thousands of these for

different, each customer of different
names, different sports they're into.

so that was all an
entire like AI workflow.

yeah.

So it's yeah, that's, so it's more in
the, that production side where we're

doing stuff less in the pure design side.

The, for whatever reason, the
design tools aren't quite there yet.

The Figma is doing stuff, some
of the other tools are starting

to bring in some AI, tools.

but yeah,

kev_2_02-28-2025_101048: are you
looking for when it does get there?

What, what would getting
there look like for you?

anthony_2_02-28-2025_101048:
so it's hard to know really.

I think it's, it is really more
around what is the experience

we're going to try and.

Make, so say for example, like
trying to do that, that Olympics

highlights clip, there isn't really
a tool that allows you to do that.

You're basically stringing together
like 20 different AI tools and lots

of code to then pull that together.

And there's still some humans
involved in some of the production

process of em, the raw video files.

And you, there's not a
simple way of doing that yet.

So it really because you've, you,
what you're there doing is you're

creating a novel experience.

No one's tried to do that experience
before, so there's no one's

obviously made a tool for that.

So it really have to work backwards from
what's experience you're trying to make.

and so that's the bit, I'm, I just,
until we figure that bit out, it's

hard to know what tools you then
need, but in terms of that experience,

it's in terms of the skills, it
still feels like you're gonna need.

The kind of basics, you're still gonna
need people who can write, you still

need people who can draw in effect.

you're still, people who've got a good
eye can photograph, can do lighting

design, can understand motion and
animation, the very old school skills.

I think you'll still need those,
but I think it will be less,

less kind of direct manipulation.

So it feels like a lot of what we do
right now is almost like, puppeteering,

it feels so you engage with these
websites or apps, they are interactive,

there's usually one or two set journeys
or one or two set layouts where you

click this button, there's one or two.

Paths that go from cleaning that button.

And it's all very handcrafted.

you literally, someone sits
down, draws the wire, frames the

layouts, and describes the flows.

If I click this button, then what happens?

And someone codes that by hand.

So it's all very, directly controlled.

It's all, it is like doing a pixel movie.

you're controlling every single
movement of the eyebrow, every single

look of the eye, every reflection.

Everything's like super, curated and
controlled and a human hand in everything.

You can see with ai, it's a
lot more, it's in create, it's

generating it in the moment, right?

It's generating it in real time.

And so you can't control exactly what
text it's gonna produce or exactly

what the image is gonna look like.

You can influence it, but you can't
direct it to such a tight level.

So I think that'll be the interesting.

Thing is then how do you then get these,
people talk about like prompters now,

like skills in prompting, it's that,
it is that idea of like, how do you

get good at being able to influence a
performance, orchestrate a performance?

'cause all these ais, you might to
do one thing, you might have 30 ais

working together with one AI acting
as their main orchestrator, the human

to create like an experience that
humans would want to experience.

You'll need probably a human
in the loop at some point to

be able to craft what that is.

but it might be more of like a sort
of, theater director or film director

type of role where you are less,
it's not like an animator where

you are hand moving everything.

It's more like you, you setting
the conditions for a performance

and you're trying to set a vision.

You're trying to influence how it
works, but ultimately you can't directly

control the actors, the performers.

You can't exactly.

Modulate everything.

you have to just create those conditions
and let the performance play out.

And it'd be different every, like
theater is different every night.

It'll be slightly different every time.

It could be very different with
ai, but you want to try and have

some sort of control over that.

So I think that'll be the interesting
skill set that will emerge, I think is

this ability to like influence, rather
than like hand puppet, and experience.

kev_2_02-28-2025_101048: Yeah, and it's
good to hear you focus on using AI to

create new things, new formats, rather
than, streamlining, business as usual.

So I think that's good to

anthony_2_02-28-2025_101048: Yeah.

kev_2_02-28-2025_101048:
Anyway, Anthony, thanks so much.

I think I need to let you get
back to the rest of your day,

but, that's been fascinating.

Thanks a lot.

anthony_2_02-28-2025_101048:
Thank you Kevin.

That was great.