Coworking Values Podcast

Why AI Can't Feel the Room: Practical Operations with Carlos Almansa

Coworking Values Podcast

"The AI cannot feel the space. It can't feel the dynamics or the vibe. But it can free up time for you to talk to your members, to have a coffee with them, to understand and to read people."
— Carlos Almansa Ballesteros

Episode Summary
Most conversations about AI in coworking are either evangelical or dismissive. 
Carlos Almansa Ballesteros, co-founder of Nexudus and author of the Coworkings AI newsletter, refuses both positions.
In this episode he lays out a practical baseline: start with what you already do, keep a human in the loop, and never mistake efficiency for community. He also raises the question that sits underneath all of it — the Dead Internet Theory — and what it means for spaces that exist precisely because human presence still matters.
No frameworks. No magic. Just what actually works.

Timeline Highlights
00:02 — Bernie sets the episode up: practical AI, not rocket science
01:16 — Carlos introduces himself: Nexudus co-founder, Coworkings AI newsletter author
03:02 — Carlos's first coworking space, and why he's always joined one when moving to a new city
06:41 — The first 60–90 days: how a community manager makes or breaks early membership
09:09 — The London Coworking Assembly AI survey: most people use it for social media captions and don't go further
10:15 — Why the Coworkings AI newsletter exists: cutting through noise to find usable signals for operators
12:54 — The solo operator and AI: actually easier to start when you know all your own processes
15:03 — Practical use case one: automating repetitive help desk replies (Wi-Fi, printing, FAQs)
16:21 — Practical use case two: surfacing data patterns you can't see manually
17:53 — The soul-of-the-space question: automation versus presence
19:36 — Sentiment analysis: feeding community messages into AI to understand the pulse of a space
20:52 — Context is everything: how to give an AI model what it needs to work properly
23:56 — What goes wrong: people who automate everything at once and erode trust
25:21 — The human-in-the-loop rule: never hand your reputation to an unmonitored system
25:56 — Transparency: be honest about AI, always offer a route to a real person
28:42 — Can you automate community? Carlos on what AI can and cannot do in a 300-person space
30:05 — The kitchen conversation: serendipity as the baseline unit of community
31:09 — What community actually is, from random coffee chats to self-organising hackathons
33:47 — Mobile work and the future floor plan: what happens when nobody needs a desk?
35:32 — The Dead Internet Theory: bots talking to bots, and why human signal is becoming the premium

Lesson 1: Only Take AI to Something You're Already Doing
The most useful thing Carlos said in the whole conversation is also the least glamorous.

Before you touch any tool, review your processes. Do you actually know them? When you're a solo operator running a space from first enquiry to member induction, you probably do. That's an advantage.
The mistake Carlos sees repeatedly is people who want to automate first and understand later. You end up with ten tools stacked on top of a process that was working fine.

Start somewhere specific. The help desk is a good example. The same questions come in every week — how do I connect to the Wi-Fi, how do I print, where's the code for the meeting room. That's exactly the kind of repetitive, low-stakes task where AI earns its place. Automate the reply. Free yourself for the conversation that actually matters.

Carlos, on the starting point:
"The best way to start with AI is to start with something that you are already doing, not trying to implement a new process or a new tool into something that you are not familiar with."
Here is how I look at it: don't ask AI to do something you don't already know how to do. I wouldn't go to AI and say, "Make me a hit record." I go to AI and say, "How do I make a podcast intro?"

Lesson 2: The First 60–90 Days Are on You, Not the Software
When a new member joins, the community manager is doing something no platform can replicate.
They're introducing names. They're reading the room. They're noticing who eats lunch alone and who lights up when someone mentions their industry. 

Carlos joined coworking spaces in South Spain, Madrid, and London. Every time, the spaces that made it work did the same thing: they put a person in the room who paid attention.
Shared lunches in Madrid led to basketball at weekends. Hot desking in London broke the ice faster than any directory ever could.

The first 60–90 days determine whether someone stays. Carlos is clear on this. And he's equally clear that no amount of automated onboarding email replaces what happens when you walk someone through the kitchen on day one and explain — face to face — how to clean up after yourself.

That small thing sets the tone. It says: this is a shared space. We're in it together.
AI can surface patterns. It can flag a member who hasn't engaged in three weeks. It cannot do the introduction.

Lesson 3: Your Community's Messages Are Data — Use Them
Members tell you how they're feeling all the time. The problem is they're doing it across email, WhatsApp, Slack, and every other channel simultaneously.

Some of it is friction: "the Wi-Fi's broken again." Some of it is gold: "that event last week genuinely changed something for me." Most of it sits in inboxes, unanalysed, until the member quietly cancels their membership.

Carlos's newsletter recently covered this directly. You can feed your community's communications into an AI, build a sentiment analysis across those messages, and surface patterns that would take weeks to spot manually. Who's frustrated? Who's energised? What topics keep coming up?
That's not automating community. That's clearing the brush so you can see what's actually there.

Carlos:
"When you have a community of 200 or 300 people in your space, it's not easy to get to know everyone, to know where they are, how they're feeling. AI can help to surface issues, to connect people. But it cannot automate those relationships."
The distinction matters. AI as insight layer, not as relationship substitute.

Lesson 4: Never Let AI Go Solo — Your Reputation Is on the Line
Carlos is consistent on this throughout the episode. New technology, new risks. The worst thing you can do is automate a customer-facing process and walk away from it.
If AI is replying to enquiries through your website, you need to be watching those replies for months. The system might not be calibrated yet. One bad reply to a prospective member and you've made a first impression you can't undo.

The related point is transparency. When someone lands on a chat widget, don't dress the AI up with a human name — "Hi, I'm Brad." Carlos didn't say it exactly like that, but Bernie put it plainly: that's obvious, and it damages trust immediately. Be honest. Tell people they can get fast answers here and offer a clear path to a real person when they need one.
Some people just want the answer. Some need the conversation. Build for both.

Carlos:
"You're putting your reputation as a space, as an operator, on the line. You are delegating that into a system that might not be streamlined at the first place. So you need to refine and you need a human in the loop, at least for a while."

Lesson 5: The Dead Internet Theory — Why Human Signal Is Becoming the Premium
Neither Bernie nor Carlos claims to be an expert on this. But the shape of it is worth understanding.
The theory is that as AI content scales — YouTube videos, emails, datasets, code — the internet risks becoming mostly AI talking to AI. The signal gets buried. The noise takes over.

You can already feel it. Emails written by AI replied to by AI. Content created to game an algorithm that was built to surface content. At some point, nobody's talking to anybody.

And here's the thing for coworking operators: this is exactly where physical space, human presence, and real conversation become more valuable, not less. Some companies are already paying a premium for human-generated data and human-written code specifically because the AI-generated alternative is everywhere and worth less.

A coworking space where genuine connections happen — the kitchen conversation, the accidental collaboration, the hackathon that started as a hot desk neighbour saying what are you working on — that is not a liability in a world of automated interactions. It's the product.

Carlos is building software that deliberately keeps itself invisible so the human experience can take the foreground. That's the bet.
Links & Resources
One More Thing
I've been in this industry long enough to have watched every wave of technology get positioned as the thing that would finally fix community.
It doesn't. It can't.

Carlos builds software. Thirteen years of it. And the line I keep coming back to from this conversation is one of his most straightforward: the AI cannot feel the space. It's not a limitation he's apologising for. It's a description of reality.

What AI is good at is the grind. The same Wi-Fi question forty times a week. The member message that gets buried under twelve other messages on a Tuesday afternoon. The data patterns that would take you three months to see if you were looking at them manually. Handle that part well and you get something back: time. Time to be on the floor. 

Time to notice who's struggling, who's about to leave, who's one good conversation away from their next thing.

That is the actual job.

The spaces I've seen get community right didn't have a community strategy. They had someone paying attention. 

They had a kitchen people actually used. They had a culture where it was normal to ask a stranger what they were 
working on.

AI doesn't build that. But it can stop eating the time that you'd otherwise spend on it.

That's the deal. Use it right.

What is Coworking Values Podcast?

Welcome to Coworking Values the podcast of the European Coworking Assembly.

Each week we deep dive into one of the values of accessibility, community, openness, collaboration and sustainability. Listen in to learn how these values can make or break Coworking culture.

Speaker 7: Beep.

If you are listening to this message,
you will know that coworking is a last

human oasis of connection available
anywhere in the new world of work.

But seriously, I don't think
I can sit through another AI

anything at this stage of the game.

But here we are, and I'm not looking to
land spaceships on Venus or Vibe Code, my

child's blood type or manifest a seamless
integration with my neighbor's fridge.

I'm looking to get through my
workday and stay alive, and I

imagine a lot of you like me.

I just wanna know how AI can
actually help me do that.

So one of the things that comes up in our
podcast today with Carlos is don't ask.

AI to do anything, start with
something you are already doing.

So I'm not gonna ask AI
to make me a hit record.

I'm gonna ask AI to help
me with a podcast intro.

So all these little processes
and things you already do, feed

them in one by one and we'll go
into that more in the podcast.

And it is like a simple as that, but it's
also really complicated at the same time.

So.

That's why I asked Carlos along
today 'cause he writes a really good

newsletter on AI and coworking and each
kind of edition is a lesson in itself.

And he writes about this stuff
and I don't, so that's why I got

him in to do this with us today.

And there's a link in the
show notes to that newsletter.

There's also a link in the show notes
to the European Coworking Day, which

is on May the sixth, which is sometime
where this public podcast was published.

And also there's a link to the
Linked In Coworking group and all

the events that are going on there.

So you can find this podcast on the
London Coworking Assembly website

and also in all your favourite apps
and gadgets and bot or whatever

you use to listen to your podcast.

So that's enough for me.

Let's get into the show with Carlos.

bernie_1_04-28-2026_113009: Hello folks.

It is, it is another coworking values
podcast, and I've got this amigo

called Carlos in the studio today.

Carlos, what are you known for and
what would you like to be known for?

carlos_1_04-28-2026_103009:
Hi Nee, thanks for having me.

It's great to talk to you.

So.

I mean, just for, for everyone
to, to know myself a bit.

So I'm Carlos.

I run this Coworkings AI newsletter
that we started a few months ago.

But also I am one of the co-founder
at ATUs coworking software that's

been around for 13 years now.

bernie_1_04-28-2026_113009: When so.

I'm sure we get asked this all the time,
so, so to be so boring, like when you and

Adrian started Exodus you, you didn't,
can you do the little origin story?

'cause you were doing, was it e-commerce
software and then pivoted as we say?

carlos_1_04-28-2026_103009: It's been.

It's been a while.

Yeah.

But yeah, when, when we
started next, it was 2012.

And at the time the industry
was, I mean, the coworking

movement was very, very small.

we had a couple of friends that
they were to start a working space

in, South Spain Coworking company.

Talking to them.

At that point, we were doing
e-commerce, like an e-commerce platform

for people to set up online shops.

Talking to these, to these friends.

We, identified the need of
like a tool for them to.

To automate some day-to-day operations
so that they could spend time or

more time with their community which
is at the core of working spaces.

And that's how we kind of started
the coworking software over the years

kind of became a coworking platform
that then connect to all the system.

Plays a role in these type of buildings
from access control, wifi management,

printing management, and so on.

So yeah, that's in a
nutshell, the, the journey,

bernie_1_04-28-2026_113009:
And when, when I always like

asking this to people like you.

When, when was the, when was the
first time you went to a coworking?

carlos_1_04-28-2026_103009: the very first
time I went, oh, that's a good question.

I think it was to this space
in, in, in South Spain.

Before we started the the Go
software, because that's I mean.

They put together the
space with the community.

So the community or the, well, the
friends and, and such, they help like

assembling the furniture and, such.

So we spent some time
there before they open.

And then I moved like, a few
times to different cities.

So.

I always tried to, to work
from a coworking space.

It, it was a good way for me to,
to get to know people as well

when I moved to a new, a new city.

So I did that in Madrid when I moved to,
to the uk when I moved to London as well.

bernie_1_04-28-2026_113009:
When, when you, when like.

When I came to live in Vigo and I
went to a coworking space, I was

terrified and totally unnecessarily.

But like I know everyone in London, not
everyone, but you know, I, I know what

I'm doing in London and I have spent years
saying to people, whenever you go to a

new city, just go and find a coworking
space and go native, get involved.

And I, I was.

I know it was easy to jump out a
plane without a parachute than it

was to go to, I ended up in FTO
Coworking, which was lovely, and all

my anxiety was totally unnecessary.

But when, when you moved out of,
when you moved to another city, how

did you feel about, was it easy for
you to just rock into a coworking

space and go, Hey, it's Carlos.

Hi kids.

carlos_1_04-28-2026_103009: I think I
was lucky in the sense that the coworking

spaces that I joined they were relatively
small and they had this tight community.

Were, for example, I remember.

I moved to to Madrid it
was in Atlanta working.

We would have lunch together every day.

So that helps.

I mean, you get people around
food, it always have, have

to to get to know each other.

And from that, we were to, to start like,
I dunno, going to play basketball together

or just meeting on the weekends and such.

So.

It was a very nice community.

And then when I moved to,
to the uk, I think I started

coworking on, on a hot desking.

So I, I wrote an article recently about
hot desking because it gives you this.

because obviously, I mean, you are
people that you don't know, but then

break the ice and you start just talking
and then you, you realise that you

have more in common than you thought.

So yeah, there is a bit of tension,
but yeah, it's moment you break

the ice, is it, it goes smoothly
and naturally, I think, think.

bernie_1_04-28-2026_113009: We,
we do that in, we have a pop-up.

I mean, there's coworking
spaces here in Vigo, but we

have a pop-up coworking thing.

Me and a group of people
called Live GIA Organise.

And it is that it's, it's the first time
I've done proper hot desking for ages.

And I, I really love it.

'cause you never know, like the
way we run that popup thing,

there is a group of people that
come every, every two weeks, but.

Also, you never know who's gonna be there.

It's a bit of it's a really good way to
meet new people and that eating every day.

I think that's in fto we would do a bring
and share lunch every day and sitting

down with like 10 to 15 people where
everyone's brought something to share.

Is that a Spanish thing or is that We, we
just lucky in the places we ended up in

carlos_1_04-28-2026_103009:
Well, eating together you

bernie_1_04-28-2026_113009:
eating together.

Everyone sharing food.

carlos_1_04-28-2026_103009: Yeah, I
think, I mean, you could say I think

that we do that in Spain quite a lot.

Yeah.

You share food exchange recipes as well.

Their masterpieces,

bernie_1_04-28-2026_113009:
the tortilla recipes.

So when,

carlos_1_04-28-2026_103009: exactly.

bernie_1_04-28-2026_113009: Some, I'm
sure it's in one of your newsletters,

like when someone joins a coworking
space, there's this like first

nine, I'm gonna call it first 90
days, which is the name of a book.

Which has nothing to do with
coworking, but I'm, I'm obsessed

with this like 90 day window.

And what, in your experience of like
jumping around and coworking working

with coworking spaces is, I think there's
something to do with that 90 days about

how we land in a coworking space and
whether we actually end up staying or not.

carlos_1_04-28-2026_103009:
Definitely, I think, I mean, to me.

Hmm, maybe 90, maybe 60, maybe 30 days.

I think in the first couple of
weeks you can get a sense of the

space is, what are the dynamics.

So I think.

To me, or at least the experience that I
had joining Corgan space, the community

manager or the operator running the space.

It plays like a massive role
on that, on those 60 90 days.

So the first thing that I like, for
example, like joining, I remember when

I joined the first coworking space that
they, introduced you to the community.

So.

The first day you at least you
hear the names, you know the faces.

So next day you kind of start
like, starting conversations,

not out of the blue because
you've been kind of introduced.

then it's, I mean, it's more
like a natural process, but

it's, I think it's important to.

To get to know the, the space, the
community that you're gonna find there.

What are the dynamics?

What, what, what are the
expectations as well?

Remember this induction.

I mean, I think that happened to me in
most organ spaces that they're doing, but

they, they take you through the space.

When you get to the kitchen,
they, they explain to you how you,

you need to clean after yourself
and so yeah, I think the team.

like a, a big role on, on that experience.

And I, I agree that the first
60, 90 days are critical.

bernie_1_04-28-2026_113009: I,
I, I, I just shouted out 90 days.

I don't want people to go away and think
there's a 90 day framework available.

It's but there's, there's,
what, what did, so I, I don't,

I don't know how to ask this.

Like, there's, I'm gonna
introduce it like this, Carlos.

There, there's, we, we did
this London Coworking assembly

quiz last year, which was.

51 people filled it in and I thought
that, oh, I, I just wanted to get

a gauge of how people used ai.

And obviously some people
are like, you know.

Totally into it and automating everything.

Especially there's, there's a freak I
know thread of people who are really good

at using Zapier or Zapier beforehand of
just greater using whatever AI gadget.

But a lot of people, and I'm being very
general with their folks, kind of use

it to write social media captions and.

When we asked, I phoned a few people up,
even probably some people listening to

this and say like, so what stops you?

And they said, I haven't got the time
to work out how to automate everything.

And there's so much noise in the universe,
it's just easier to keep going as I am.

And I think that's pretty accurate.

And like with you a newsletter, who
are you, who are you writing it for?

Is it enterprise dudes or is it.

You know, local things.

'cause there's a lot of information
in there and it is really useful.

But how do you, how do we get to it?

carlos_1_04-28-2026_103009:
Yeah, I mean for us.

itself was a way to cut
that noise as you said.

So, there's a lot of
things going on around ai.

I started the newsletter with Fran,
so, we've done some projects together.

We talk about about AI quite a lot.

But then we, we found ourselves
in this, okay, this, there's this

thing that's new, or next week
there is another thing that's new.

So these apply to coworking in particular.

So we wanted to somehow find those
signals and kind of translate those

signals in, into something that.

Us, but also coworking
operators could use.

You'll find on the newsletter from
to specific trends that can apply

to the, to the coworking industry.

So it's not so much meant to
be for like a specific segment

of the working industry that.

For people that are trying to
get their way through, through

the noise and find the signals,

bernie_1_04-28-2026_113009: That feels

carlos_1_04-28-2026_103009: Of
these technologies, which my

opinion, I think there are two
kind of trends at the moment.

One being some people tend to
think that this is like a magic.

Portion that is gonna sort out some
other people think that this is just

noise and they need to ignore it.

In my opinion, the answer at
least to me, is in the middle.

So it's a technology that is here to,
to stay, but it's not something that

is gonna sort out everything right.

bernie_1_04-28-2026_113009: Does how
do, because when you said they're like,

they're working their way through the
noise, it feels like kind of hacking

through, which is my experience too,
like hacking through a forest with a

quite blunt machete to, to get to the
food on the other side or no, and, and.

One thing I did early on is I just
went, I'm only 'cause I, I'm only

gonna listen to these people about ai.

And that was Kofi.

And there's Tom, who I know you met, and
there's another writing group I'm in.

And it was so tempting to, and I'm a one
person operation to get into everything.

And then I know when they say like, use
these 10 prompts to 10 x your million

dollar business in 15 seconds, you know.

Like what, and then which application
to use and stuff like that.

And there's something about when
everyone in the world has access

to whatever their AI tool is, you
know, the mainstream ones, which are

about 20 to $30 or pounds a month.

Like what do you, what,
what do people do with that?

carlos_1_04-28-2026_103009: I
think, mean the scenario that

you mentioned for the one person
that is running everything it's.

It could be even easier to
get into ai and I explain why.

So when you are running all the
processes, imagine your coworking space.

You are running the space from,
I dunno, selling the space

to welcoming people and such.

So you know your processes, you
know what they're following.

The moment you scale out.

size and teams and such, those processes
cannot can get scattered and it's more

difficult to, to align the thing and such.

But when you know your processes,
a good starting point is okay.

It's, I mean, with all these AI
noise and such, it's a good moment

to post and say, okay, I'm gonna
review my processes and see if it

makes sense and where it makes sense.

try to use AI to, for example automate
some repetitive tasks or make some

processes a bit more efficient so
that I can dedicate my time to some

things that are where I are more,
more value, if it makes sense.

So think before starting thinking on
any tool, because at the end of the

day you have like so many options.

You can end up with, I don't know,
10 tools on top of, a process that

you were just running yourself.

And it was fine.

I think it's a moment or it's a good
moment to just rethink or review

the processes that you have in your
space and then where AI could help.

That at least is my experience and
what we are doing internally as well.

bernie_1_04-28-2026_113009: So.

It, it sounds like, tell
me if I'm wrong here.

It sounds like you're saying don't only
go to AI in a, in a work context only

go to AI with something you are already
doing and ask how to improve that.

Don't turn up and say, how can
I 10 x my space to, I know.

Like do ask it to do something you're
already doing is what I meant to say.

carlos_1_04-28-2026_103009: So I
think the, the best, or, or at least

the, in the best way, I think it's to
start with AI is start with something

that you are, you are already doing.

Not trying to, to implement
like a new process or a new

tool, into something that.

are not familiar with.

So let's say for example I don't
know, you are replying to customers.

So in, in your community that's a, a very
specific use case and very well known.

You have a help desk where people
submit like questions about how,

how do I connect to the wifi?

How do I print this document,
cetera, et cetera, and you get

those questions every now and
then, and they are always the same.

That's a very good use case for AI
to come in and automate the reply

so that when they, they ask the, the
next time, you don't have to to reply,

and that will save you some time.

Same goes with replying about the
same topic over and over again.

Or looking at the data.

One thing that AI can do really
well is start facing data that

might not be obvious, especially
when you have like a space.

That is bigger.

So looking at the data granularity
is quite difficult and that

for that AI is quite good.

So it comes face patterns and trends
might not be obvious on the first place.

yeah, I think applying AI to
something that you are familiar

with and you know, your process
is to me the best way to start.

then also you can expand.

bernie_1_04-28-2026_113009: The that,
I mean, it's a different type of data

to what you're referencing there, but
I, I do loads of research and Google

Notebook is phenomenal for that, and I
can, people like Marco or radio can send

me academic papers, which would be so
hard for me to digest, and I put it into.

Google Notebook, chat with it.

I make podcasts about it.

And then like with your newsletter, like
that's how I consume, sorry to consume.

But that's how I, that's how I
interface with your newsletter is

by making it into a, a, a podcast.

And then I listen to things and
then I can chat with it and it.

Just, you know, for, I think this
is great for anybody, but you

know, for me being dyslexic and a
DHD that has been life changing.

Before I forget when, so you said about
automating responses and stuff like that.

A, a constant, I know thing that comes
up is like, I, particularly for people

who have smaller coworking spaces, is I
don't want to lose the soul of my space.

If I automate everything, I'll.

Will just be, you know, some sterile
clinical detached thing, and I'm sure

you've got something to say about that.

carlos_1_04-28-2026_103009:
Yeah, I mean, to me the, the

argument in this case would be.

That if you are replying manually
emails or the same question over and

over again, you don't have the time to
spend on the floor with your community.

So one thing that AI cannot get is
talking directly to your customer.

They cannot, they AI
cannot fill the space.

It cannot feel the dynamics or
the vibe of the space, it can free

up time for you to talk to your.

Members to have a coffee with them, to
understand and to read people because

obviously it's when you talk to people
face to face and when you have that

relationship is where you can build that,
that community with within your space.

And for that, I think automating things
is useful to dedicate more time just

to spend with your with community.

bernie_1_04-28-2026_113009: It is.

'Cause then the other, in any
community context, there's.

People will email you, they'll WhatsApp
you, they'll slack message you and

all the, all the different, all the
different ways people communicate.

And sometimes those will be, I hate
you 'cause the WiFi's broken again.

Or, I love you because that
event you did was life changing.

And what, what do you do with all
of that to help you understand.

The pulse of your community or you
know, what can you do with that?

'cause I, people message me in loads
of different ways around events

and podcasts and stuff like that.

And I've kind of learned how
to organise it in my context.

But if someone came to you and went,
Carlos, what do I do with all these?

What do I do with all this information?

'cause it is something to do with it.

carlos_1_04-28-2026_103009: Yeah, I think
on the last number that we, we published

in the newsletter, there was an example
where we build this sentiment based on

messages that your community sent to you.

So you can.

Populate that into an ai.

And it builds like a, only the sentiment
of the community or different people

within the community, but also it
gives you, you can brainstorm ideas.

I think one of the things that
all these models changes the way

you interact with technology.

So you mentioned about being
conversational and I think that's a big.

Point as in you can just
talk to the, to assistant.

Say, Hey, okay, I, you gave me this idea
why, should I apply to my specific case?

And then there is a back and forth and you
ended up apply with some action points.

But it's very good at building this
sentiment analysis based on, on messages.

And think on the next number
that it's, to happen this week.

We are expanding on that as well.

So if you have that data, AI is
good analysing and helping you to

understand answer facing patterns
and trends and possible issues.

bernie_1_04-28-2026_113009: So look
like attached to that is, and this took

me a long time to under understand,
is like the context is everything.

And when people, like Kofi would say, oh,
it's all about the context, which he, you

know, was exactly what I needed to know.

I didn't really understand
what he meant by context.

So when people talk about.

AI and context particularly, I'm gonna
narrow it down to like, you know, if

I run a small coworking space, what,
what does that, what does con, what's

the context of the context, Carlos?

carlos_1_04-28-2026_103009: So
the context of a, mean, I think

the context of a small, medium, or
large coworking when it comes to

ai, it's, it's similar, I would say.

so.

The contact the AI needs is the, I
mean, it'll depend obviously on the,

on the outcome that you're looking
for, but if we were to, to narrow

it down to a specific case as we
were talking about understanding

the, the sentiment of the community
based on the message that they sent.

Obviously you need to feed the, the
messages, I mean all the communications

that you have all the tickets
that the community raise with you.

also if you have like more details
around how you operate, the system

that you have in place, how they
work, You know, like frequent issues

that comes every now and then.

that will give the model context.

And then you can expand if you are
looking at different use cases, like

if you want to reply people when they
reach out the website, for example.

So then you will need to give
the context of, okay, what, what

are you offering your space?

So it, similar to training
someone that joins your team.

So if you have a person doing or
receiving people in your space at

the front desk, you will need to.

To tell this person or
to, to train this person.

How do you want to receive people?

What, what is the information that
they're likely to ask and what is

the information that you can share?

And things like that in that, it, in
that sense usually is different because

it's a technology, but the process
of training the, the technology is

similarly to when someone use your team.

If it makes sense.

bernie_1_04-28-2026_113009: It does.

And with like attached that is what
is, what is something I'm gonna

ask it like this, which is probably
the most dangerous way to ask it.

What, what is it you see
people doing when they try to.

Automate ai.

What, what, what's, what is the thing
they just should never let I AI do?

Because there must be people using
AI and they think they're being,

because I've done this just to,
you know, like they think they're

being really slick by saving time.

But actually they're, they're,
they're automating and it's

really obvious they're automating.

carlos_1_04-28-2026_103009: Yeah,
think, I mean, one thing that I think

it's important to highlight is that
these technologies that are new,

we always need to ensure that, we
implement them in a secure way as well.

because there is a lot of noise.

Everything is like changing so fast
that it's, it's easy just to add to say,

okay, I need to implement everything.

And I don't think that's the right way.

One thing.

I think people try is, people
to do is to do it like at once.

that to, to start with.

So breaking it down in smaller
problems or that you want to, to solve.

I think it's, it's a good approach
and what you should, I mean,

it depends probably on the,
on the operator or the person.

But obviously when we were talking
about having that or building

that relationship with your it's
something that AI cannot replace.

So being on the floor,
talking to them, et cetera.

Another thing that I've seen is
letting AI go solo all the way.

So not having a human in the
loop, at least until you kind

of validate how, how it works.

I think you need that human in
the loop, like the quality of.

The outcome.

If it's, if you are using AI to reply
to, let's say, to people that reach out

through your website, at least for a few
months, in my opinion, you need to be on

top of those replies because at the end of
the day, you are putting your reputation

as well as, as base as an operator.

You are delegating that into a
system that might not be like

streamlined at the at the first place.

So you need, you need to refine
and you need a human in the

loop at least for a while.

bernie_1_04-28-2026_113009: There's
with that I always appreciate it.

I dunno, I dunno what best practises as,
you know, we like to say, but when I go

to one of those things I use, it says,
we'll get a, get an answer quickly here.

And I get an answer, and what
I really want is an answer.

And then there's always an
option to say, speak to a human.

And I, I think that's the way, like
being honest and saying, you know, when

it says like, hi Brad here, what would
you like to know about mega software?

You know, I'm like, oh God, that's
so obviously not you, you know?

But saying you can get frequently asked
questions here by chatting with me.

'cause you are, I feel
like I'm doing something.

And then when it says, you know,
do you wanna speak to a human?

I will take that option.

And most things you're gonna
ask for service on you.

You use a lot, you know, so
you will know how to do it.

What, what's, that's what
I think we should do.

Is that what you should do?

carlos_1_04-28-2026_103009: I,
I mean, that's a good point.

The moment, I think.

It depends on, on the case.

So you reach out to services where
it's a simple thing that any AI can

guide you through, and that's fine.

So you don't need to, to talk
to, to a person, obviously,

when it's something complex.

And then you may need to talk to,
talk to a person and you can't as

you said conversation kind of feels.

It still feels differently when you're
talking to, to a human even on a chat.

But yeah, I mean, these system are,
are also like evolving quite fast.

So.

They, they will improve and
change in that, in that sense.

And I assume also it
depends on the person.

Some people they, they prefer just getting
the answer regardless of who deliver

the answer or what, deliver the answer.

people prefer to have
conversation with human.

Yeah.

bernie_1_04-28-2026_113009: There's,
there's that, that human thing is.

This, oh dear.

I dunno what I wanna say, but there's,
there's this kind of, I've gotta

say the community word, Carlos, and
I get this sense, and it's a sense

that people feel they can automate
community more than they ever have done.

And.

Maybe we're at this level of like,
we can, we'll either become the

matrix or we'll all reconnect.

Like what is, what is,
I'm trying not to answer.

Ask this question.

'cause you must have
had it a million times.

Like what is community and
coworking Carlos, but then like

what's the effect of AI on there?

Like in that thing?

'cause I, as you will know, like I
deeply believe in the whole community

thing and I fear it's gonna be.

Further consumerized by ai,
but what's your take on that?

carlos_1_04-28-2026_103009: I mean
that, yeah, that's a good question.

I think automating community,
it wouldn't be the, the right

concept for me, but empowering.

Community managers to understand
their communities better.

I think in that AI can help.

So when you have a community that,
that it's, I know let's say you have

like 200 people, 300 people in your
space, it's not that to, to get to know

everyone, to, get to know where they
are how are they feeling, the space,

if they have any issues and such.

I think in those cases.

AI can help to surface issues,
to connect people as well.

So at the moment, the, there
is not much around that, but

you could potentially have.

On the system, like something that
help people to connect a bit better

and to get to know each other, better.

So we, we fought, for example,
the directory for many years,

but it's a static directory.

Imagine that being dynamic
so that you get like filter

introductions and things like that.

So I think surfacing and
helping people to connect.

with each other.

It's something that AI can help with.

But it cannot automate those relationship.

At the end of the day, we are humans
and we, we need that human and

face-to-face interaction, and, and that
happened within the physical space.

It happen in events.

It happen.

On a day to day, when you go to, the
kitchen to have coffee and then you

find someone, then you start talking
about presents that you're working on.

You realise that you have something
common, you decide to collaborate.

But yeah,

bernie_1_04-28-2026_113009: I feel that.

I feel that I love kitchen's
the best place in the world.

Kitchen and eating is the
answer to everything, but,

carlos_1_04-28-2026_103009:
de depends on the kitchen.

bernie_1_04-28-2026_113009: Well, that's
a whole podcast in itself, but the,

carlos_1_04-28-2026_103009: It is,

yeah.

bernie_1_04-28-2026_113009: you know,
I, I just, there are some people, or

in, in the coworking universe and they,
you know, there's like kitchens, coffee,

desks, chairs, community, and it's
like this li it's like a line item on a

spreadsheet and then there's people that
have all those things and then they just.

Kind of believe in community,
a bit like human interaction.

That will happen anyway.

Like how after going this far in the
industry, like what is that community

thing to you at this point in time?

I.

carlos_1_04-28-2026_103009: I mean
community to me, it's what in between

those people that works in your space.

I remember when we started in,
in 2012 I started like attending

these, these conferences and such.

Was this concept about serendipity.

Where people would talk about those
casual connections that happen as we

were talking, like when you have a
coffee or when you are in a hot desking

area you start talking to to people.

I think that's kind of the basic level.

you have a second level, which
is when communities start like

organising themselves and they start
like organising their own events.

they start their own projects and such.

So there are different grades
or levels of community, my

opinions in different spaces.

and it's got to do on how.

space kind of, accommodate those
communities, how they those communities

and what, or how much freedom they
give also, or not, I don't want this,

I mean, I don't know if freedom is
the right or, but how much they, they

let those communities also to kind
of grow and organise themselves, so.

From a random conversation in
the kitchen with a coffee to

a group of people organising.

I know a hackathon on the weekend.

I think there are different levels,
but all that to me is yeah, is

what I would call community.

bernie_1_04-28-2026_113009: Yeah.

You said serendipity at the
beginning there, there's do, do you

remember Ronald from seats to meet?

Did you ever meet him?

carlos_1_04-28-2026_103009: Who, sorry.

bernie_1_04-28-2026_113009: Ronald from

carlos_1_04-28-2026_103009: Said to
meet I think probably a long time ago.

Yeah.

bernie_1_04-28-2026_113009: Is it?

carlos_1_04-28-2026_103009: met
there in one of these conferences.

bernie_1_04-28-2026_113009: I I
thought that like randomly, I, I

thought he would like, he, I think
he would've met him along the way.

'cause he's definitely been a, a cowork
in Europe in the, in the old days.

But he had this whole serendipity
conversation and I, I'm gonna

have to go and look up his book
when we finished this, because

he published that in like 2014.

And I, it'd be really interesting to see.

A lot of it was taking shape
at that moment in time.

How we would be like, software would
enable us to be in spaces together and

we'd have more time to be with each other
and, and then like, 'cause so much will

be, this was even before anyone knew
what chat GTP and Claude and Gemini were.

I The actual last thing that's really
valuable is like a face-to-face.

Conversation, interaction and in
the context of what he was talking

about, business deal, you know?

carlos_1_04-28-2026_103009: Hmm.

I mean, one concept that we were
discussing last week in another

conversation is what happened if.

All this technology kind of enable people
to work from their mobile devices, and

you don't depend on a laptop or you
don't have to to be sitting on a desk.

How the physical space changes
where you don't have desk anymore.

So,

bernie_1_04-28-2026_113009: that is
more pertinent than you may think.

'cause I do, I walk more and more
around on my phone and I actually,

I actually get more work done on my
phone now by walking and talking.

Then I do, I use a thing
called Voice Notes.

And amazingly you do, you know,
contingent works, which is in Bromley,

the, the, the, the makers of Voice
Notes are based in contingent works

in Bromley, which is like a, a freak.

Oh my goodness.

How, how are they there?

But that app I just chat into and it
works really well for the way my, you

know, my, the way my brain works and.

I just sort of organise everything
so when I come back here to my great

big computer and loads of screens, I
can just, I know what I'm gonna do.

And but until you said that just now,
like obviously if everyone's walking

around in their mobile device, that's
gonna really dramatically change

the floor plan of coworking spaces.

There's one thing I wanna
ask you before we go.

And I heard you talking to Hector about
this, that dead internet theory is that.

Like, can you explain that?

'cause that is really
dark and interesting.

carlos_1_04-28-2026_103009: The, I
mean, that was something that I thought.

Put together on the, or it was on the
newsletter and we discussed briefly.

I'm not, I'm not an expert
on that topic to be honest.

So yeah, I don't want to, to say
anything that is not accurate.

bernie_1_04-28-2026_113009: But
what is the I, I'm not asking you to

agree with it, but what is the the
idea Is it that we like is just that

like LinkedIn is just bots talking
to bots, it's like a comedy ending.

carlos_1_04-28-2026_103009: I mean that,
I mean, for everyone listening, yeah.

That's a theory that as AI is kind of
grow and such, you would end up with an

internet that it's only bots or AI agents,
for example, talking to with each other.

I mean, to some degree it's,
it's happening with content, so.

You got much more content that is
produced by AI in different fields.

You, you see people, for example, I was
listening to this to this programme,

the radio some people talking about
research on, on health and such.

They were buying data sets before
charge GPT because they don't have

like AI generated data and such.

and I heard another company recently that
is source code for apps produced only

by humans, like, and things like that.

So I mean, the theory of this death
internet is just that, that you end

up with an internet that is only.

AI talking to, to ai and I mean,
to some degree also, you can feel

it like sometimes with emails.

So you got an email that is written
by, by ai and then you, you answer with

another email that it's produced by,

bernie_1_04-28-2026_113009: It's so scary.

It's, it,

carlos_1_04-28-2026_103009: is.

bernie_1_04-28-2026_113009:
it, it, it's tempting as well.

Like I, I don't get a lot of emails, but
if I had to deal with like a lot of emails

every day, I would, you know, definitely
be leaning more to those lukewarm

prompts in my, in my Google Gemini.

Okay.

What, is there anything you,
is there anything you want

to shout out before we go?

'cause it's definitely lunchtime here.

carlos_1_04-28-2026_103009: no, I
mean just for anyone interested in

AI and coworking, just take a look
at 10 years later, I mean, we want

to basically every other week, every
two weeks we share some content.

Recipes, trends and such
around AI and corings

bernie_1_04-28-2026_113009: You,

carlos_1_04-28-2026_103009: ai com.

bernie_1_04-28-2026_113009:
you're really underselling it.

It is, it is way more
comprehensive than some.

And when you say recipes,
I just think of food.

But we'll,

carlos_1_04-28-2026_103009: Oh

bernie_1_04-28-2026_113009:
we'll put a link.

carlos_1_04-28-2026_103009:
more like, yeah, go.

bernie_1_04-28-2026_113009:
After you Carlos

carlos_1_04-28-2026_103009: Yeah.

More than recipes.

Like, yeah, some step by step cases
that people can apply to their own.

And day-today.

Yeah.

bernie_1_04-28-2026_113009: Ace.

We are gonna put a link
to that in the show notes.

Also, folks, there's links to European
Co-working day, which will be somewhere

around where we publish this podcast.

There is the LinkedIn
co-working group, which is.

Hundreds of you engaging around coworking
stuff in there and wherever you are

in the world, be careful out there.

It is a jungle

carlos_1_04-28-2026_103009: Thanks
so much, Bernie, for having me.

bernie_1_04-28-2026_113009: Joe.

carlos_1_04-28-2026_103009: Bye.