Subscribe
Share
Share
Embed
Need to use a SharePoint intranet due to internal policies, company transitions, or legacy systems? When all the available information is overly technical or negative, where do you turn? Enter Fresh Perspectives: Living in the reality of SharePoint intranets. We provide useful, jargon-free insights and real-world examples to help you maximize the benefits of SharePoint intranets and tackle its challenges. Pro-SharePoint but realistic, we debunk misconceptions and share product management insights from Fresh Intranet.
hi, everyone.
Welcome to a new episode
of Fresh Perspectives.
I'm the Jarbas Horst,
senior product manager
for Fresh Intranet.
Today, we have someone
very special with us,
someone you all know, someone
who filled up rooms at events,
who is an inspiration for many.
It's not David, who is also
like an inspirational person.
But David, let's
start with you first.
Hi. Hi. What a welcome.
Dave Bowman, product
director for Fresh Intranet,
continuing the theme
of ownership on this
season, and it seemed
like the right time,
to bring in Chris O'Brien.
Chris, introduce yourself and tell
us who you are and what you do
and what you own.
Thank you both.
I'm Chris O'Brien.
I'm the CTO of
Advania in the UK.
Been in the group for
twelve or fourteen years,
something like that.
I own a big part of our
technology strategy and the
propositions that we
take to our clients,
develop some of
them in particular,
and have spent a lot of my
last few years focused on AI,
which probably is
common to many CIOs and
CTOs.
So developing services around
that, helping us grow in AI,
helping us serve
our clients better.
Also responsible for things
like some of the technical
aspects of our partnerships,
especially with Microsoft
and other key vendors,
how we sort of bring
technology services into that.
And also leading a big part of
our own AI transformation at
Advania.
You know, we're just
like any organisation,
we're modernising,
we're automating more,
we're rewiring parts of
our business with AI,
and a lot of my time and focus
goes into things like that.
And, know, Chris, you're
consulting with customers,
delivering value to the
people that you work with,
what's the thing that you currently
feel most responsible for?
I think it's kind of this
combined inward and outward
focus, but it's all about
kind of transformation and
AI, that's definitely
a big part of it.
Clearly, those aren't the only things
that organisations need to worry about.
They need to focus on building
their business and doing
business well, but
I think everybody is
appreciating that we're in a
sort of technology moment and
there's something to capitalise
on and something that you can
get right and something
that you can get wrong.
Many of my conversations, both
inside the company and with
CIOs and CTOs and boards
and senior leadership
teams, are about, Okay, how
can we do this the right way?
How can we kind
of make the right
decisions in AI, both
on the technology front,
but in terms of how we're
transforming a business?
And that just seems to be
the fulcrum of so many of the
things that I'm involved
in at the moment.
And it's quite interesting.
So I'm fortunate enough to have
a really strong relationship
with Microsoft and be close
to their product strategy.
I've been an MVP for
nineteen years now,
which is strange to think about,
but that's kind of got me
into very interesting rooms in
Microsoft and very
interesting conversations.
So, you know, really good to see
how they approach it.
And, you know, I think we all
have our views on Microsoft,
some things that they
do amazingly well,
some things that, you know,
are newsworthy and maybe
get a different slam.
But, you know, really good to be close
to some of the decision making and some
of the thought process
there that, you know,
try not to have a
bias about anything,
but just to be close to one vendor
at least is really powerful.
Yeah.
And super valuable
given the, you know,
the sort of progress that
they've made on kind of
corporate AI particularly.
Yeah.
You know, you post a
reasonable amount on LinkedIn
and, you know,
I'm I always read your posts
with some level of sort of
seething jealousy
as well, Chris.
And you posted recently about,
you know, the post was kind of,
you know, about being
pro AI but anti misuse,
and to sort of coin a
chat GPT ism, you know,
it stuck with me for a a bit.
And, you know, it was a
really nuanced take, I think,
on the balance of using
AI to create things
and doing it responsibly.
And I guess I'm curious really
about how you arrived at that view.
Is that something about
you? Is that your nature?
Is that something
that you've learned,
a framework that
you've picked up?
Can you talk a
little bit about it?
Yeah.
I mean, I think I
was observing that
AI usage is ballooning
and people in every role,
every function are adopting
it, and that's great.
But I think we're coming into
sort of like a phase of maybe
unintentional misuse of AI.
And the effects of that are really
quite interesting to observe.
And I'm feeling really strongly
that we're kind of making some
big mistakes with AI, and
it needs to be considered maybe
at the company and
the individual level.
So what I see more
of is people using
AI, which is great, I think
we're all behind that,
but perhaps taking less
ownership of the process and
being accepting of AI
being in the driver's
seat more than the human.
And you see it in pitch decks,
you see it in email
communications,
you see it in vendor
proposals, strategy documents,
anybody doing some analysis,
maybe industry analysis
or competitor analysis.
My inbox is full of kind of
targeted emails from products
and vendors, and you can tell
that they're all AI generated.
And nothing intrinsically
wrong with this.
I think my issue is where
people and teams are using
it and that becomes their
output and the ownership isn't
there, and actually
that's leading a few bad
decisions to be made and
I think a few kind of bad
outcomes for
organisations and teams.
And yeah, you know,
it's this that is on my mind
a little bit at the moment.
You know, AI can be this
massive accelerant but the more
we kind of go through this era,
the more you spot places where
it's kind of not being used in
the right way, I would say.
That post of yours also had
the angle of vibe coding.
That is very relevant for
the space that we operate as
software vendors.
And that means anyone
nowadays can create a software
and bring that to the market,
present that in a demo.
But on the demo,
I think there is a lot of things
that can't be covered, right?
Like how how is the
foundation of the application?
How is the
the re are there reports covering
the quality of the software?
I think what are the questions
that buyers, you know,
software buyers should be asking
today with with Vibe coding in mind?
Yeah.
I mean, I think what vibe
coding has illustrated is that
some bits of software
development get so much easier
and faster and and
the effort collapses,
but it doesn't take away
the end to end of of kind of
creating good software that
can be adopted by a business.
So we're definitely in this
period where it's shining a
light on all of the
challenges that come with
fast AI created software.
How does that run? Is it
performant? Is it scalable?
Is it secure?
Is it going to respond well
to different scenarios?
Has it been tested?
And the answer in
many cases is no,
because Vibe coding isn't
really optimised for that today.
I'm sure we will close that gap.
It will almost move on to
be the next focus area of AI
assisted software development.
But at the same time,
large language models excel at
language and code and syntax
and structure, so no surprise that
we were easily able to solve that.
The kind of aspects of platform
engineering and DevOps and
SecOps, they're a little bit
harder to solve and tend to be
a little bit more
specific to the context.
And we're seeing loads of
places where I'm seeing people
create apps,
they host them somewhere, but
it's not anywhere resilient,
it's not necessarily secure.
The code itself is probably
full of security issues.
It would fail a pen test
immediately, but you know,
is a pen test happening?
Certainly should
be on this stuff.
But it's interesting because
that's a mirror of not just
coding and software
development, but actually I
think how so many
people work today.
We're not just Vibe
coding software,
we are Vibe strategising,
we are Vibe road mapping,
we are Vibe
proposal generating and RFP
responding and it's here
where I think some of the
consequences for businesses,
know, they can shine.
So just to kind of go
on that a bit further,
we've already seen
some examples of this.
I think you might remember
Deloitte were in an
interesting situation where
they did some work for the
Australian government.
Turns out it was full
of AI hallucinations.
They had to give a refund
on a lot of the work.
Same with They issued a
kind of security analysis
of a whole bunch of products.
Turns out that many of the
vulnerabilities that were
listed were not there, they
were effectively hallucinated.
But these are all examples
where the right kind of
validation and fact checking
and analysis on the outputs
didn't really happen.
It goes out with the
organization's name on.
So, you know,
I think we are in this phase
where this becomes a bit more
common.
Yeah, and you know,
there's a big difference
in there between,
AI finishing,
finessing something
that was already really good,
directed by someone
that's got domain
expertise and understands how
to spot the hallucinations
and, you know, can challenge
the AI on is created,
or, you know, versus it being used to
just outsource the thinking entirely
and create something that's,
you know, ultimately thin or,
like you say, of errors.
I think it can be difficult,
particularly for people that
maybe are new to this or
don't necessarily have the
experiences of working in
technology or think about things
with a sort of developer mindset.
How do you know which side
of the line you're on?
Is something that you can feel in
the moment or a hindsight only thing?
I think people that are
very, very deep in AI,
and I think all three
of us probably are,
you become attuned
and customised to
finding the tells of AI,
and then when you know that
something has been AI generated,
it does raise the question of how
closely involved the human was.
And
I think the thing that
comes to my mind is, okay,
the human or the team or the
company that has produced this,
are they fully behind every
last part of the message here
or every last part
of the product?
And I think, you know,
the way that I use AI,
I kind of want to be
very much in control,
try not to outsource
the whole task,
be very extensive and
directive in sort of how I'm
getting AI's help,
very clearly explain data that
I'm giving it or artifacts that
I'm giving it and how
they should be used,
not just throw a load of kind of files
in there and ask for something back.
But then I think it's
also really spending time going
through the outputs
with a really fine comb.
And for me personally, I
feel that AI helps me a
lot, It accelerates me.
And if I'm producing a
PowerPoint deck for a strategy
document, for example,
then I get more from AI.
That's what AI is really good at
compared to me generating the deck.
But that's also where
the danger is, I think.
Because what I get is
probably more slides,
more density on the slide,
more points being expressed,
and then, you know,
for me personally,
I think I need to own every
single last bullet point
and tagline and framing
message and and kind of sub
banner down here, the AI is
really good at producing slides.
I need to own every last
one of those messages.
And so I find myself taking
away as much as anything else.
But that's how I think about it.
And then I know that I can
truly stand behind the output
if I'm scrutinised, if I'm
kind of up in front of a board,
it's defensible.
But that's what I think others
are maybe not doing, and I see
c level executives,
senior leaders,
budget holders being
convinced by a really
shiny deck, and actually it's mainly
AI generated from a single prompt.
I really like that idea of
using subtraction as a test.
Right?
Because if you
genuinely understand
your domain or what the output
that the thing has created,
you should be able to be able
to say, I don't need that.
That's a you know,
there's there's fluff,
extra points being made,
not backing up the narrative
or the story, get rid of it.
And I think that that feels like
a really good test in there.
And I think that's that's like you
using AI to augment your capabilities.
Right? So that's a
big differentiator.
But I think regardless
of how you use AI, right,
it's augmenting your skills,
it's doing the entire work
for you, you own the outcome.
When this thing goes out, when
it's presented to someone,
when it's published somewhere,
you are responsible for what
you are putting out there.
I think you should take
the credit whether it's good or
bad, because sometimes
people put something,
publish something,
and they will say, oh,
and that lends well,
and they will say, oh,
that's AI generated.
That wasn't me, that was AI.
In the end it's you,
because if it goes badly,
you are responsible for
that outcome as well.
Yeah, I mean, there's an interesting
kind of organizational dynamic to that.
You know, in some ways I feel
it should almost be
a firing offence,
almost like a job losing
situation when somebody's using
the excuse of, oh that was
AI, AI generated that bit.
Because if can't stand
behind every part of it,
then it shouldn't be there.
So as we kind of navigate this
stage and the relationship
between humans and AI
is ever changing, yeah,
this accountability and
this ownership and this
almost scrutiny of the results,
I think that's that's where people
are getting the best results from AI.
You know, you're getting
a higher quality out output.
You're probably moving
faster for sure.
But hopefully, AI is additive.
It's finding more information,
it's finding more data points,
it's finding ways of
strengthening the argument.
And because it's this sort
of almost like finely tuned
combination of
human and machine,
that's where the optimum is.
But I just feel that lots of
people are probably not in that
place yet.
It's fooling people because
people think that's how it has
been created when it's not.
And I think if you're in the
space and it's your domain,
you can really spot those places
where it's not happening like that.
I
think I've been on
with it as well.
Remember doing presentations
about the Pope coat many years
ago in the very
early days of AI.
And, you know, there's this sort of obvious
content that's been generated with AI.
Like, I'm a big fan of AI
generated cat videos, for example.
I spend a lot more time looking
at them than I should do.
And, you know, I know
it's AI generated,
and I know it's frivolous,
and I know it's stupid,
but it's still there's like a
level of appeal of it for me.
And I think I've begun
to interpret kind of written
content as well as, like,
I know that I'm reading
an AI generated post,
but it doesn't hide the fact
that there's interesting
thought behind it.
And, you know, the fact that someone has
got a good point that they're making,
and they are a smart person,
and they've used AI to create
a piece of content, you know,
like I'm accepting that it's AI
and I am accepting that it's good.
I think it feels possible to
live in that place, isn't it?
Yeah, I think so.
You know, these things
aren't mutually exclusive.
But then I think the
counterpoint to that is you see
quite a lot where
the human touch isn't there.
So so just to keep on that
point for a while, you know,
I was looking at a thread
this morning on LinkedIn,
Microsoft announced their
quality results in the two
days, and so Satya
does does a post.
And oh my gosh, reading
that thread is yeah,
it it's kind of reading those
types of threads is super
interesting these days you've
probably got a few hundred replies,
Sixty percent are very bland,
very generic AI generated
where, you know,
somebody's just got a bot
and it's basically feeding in
Satya's post and, you know,
generate me an appropriate
response to this.
That's about two paragraphs and
probably comes from my angle.
But that's hundreds
of the comments.
And what's interesting is you
see humans responding to some
of those comments who
haven't probably realised
they're bot generated.
But also you see, in many cases,
the entire thread is sort of sixty
or eighty percent of those replies.
And what's interesting is
there isn't really anything
additive to the discussion or the
debate or the context in most of that.
So in some cases, yes,
there can be and the AI has
landed on something that is a
powerful observation that
is worth thinking about.
But so many of these slightly
hyperbolic replies
where, you know,
oh this is a pivotal moment
and what stands out about this?
Compounding.
Yeah, this is industry defining.
And in many cases it is not,
but it's just the typical
constructs that AI uses.
So I think there's a balance.
Where is high quality stuff and
the AI has maybe been guided
well enough to produce
some insight that's worthy,
then that's great.
I don't think we need to
care too much about where the line
between human and
AI generated is.
But it's the fluff and the
noise that I think would be
great for us all to evolve
to have less of this somehow.
Hear about the dead
internet theory now.
So much of the internet
is now AI generated,
and that's then going to
be used to train the next
generation of language models.
Where do we get to for an
internet in ten years' time?
Yeah, and is that something
we that we need to just
accept it, live with
the AI generated constructs
that we have there now?
One thing that I saw
today on LinkedIn,
someone posted that there
is a new functionality,
it's called Seems
like AI is locked.
So it's a button,
you can click on that and it
reports basically the content.
Are we going to see, think,
you think more of that
or maybe AI free spaces,
if that's even possible?
I think, yeah, we'll
always see more of these,
I think, kind of constructs and
aspects of AI generated content.
Look, to be clear,
I think there are some that are
just entirely harmless because
the types of expressions and
phrases and constructs that you
see in AI that
can be irritating,
I don't think there's anything
harmful in that even if it's
just a little bit mildly
irritating to read.
You know, the it's
not x, it's y.
It's not an x, it's
not a y, it's a zed.
Or this is load bearing,
this compounds over time.
You know, what nobody tells
you or what nobody's talking about.
All of that stuff, I don't
think is too harmful.
Going back to some of the
examples in the business world,
what I think is more
that we need to pay attention
to is these cases where AI has
made a leap perhaps
that is founded
something that isn't robust.
We get a lot of AI where
it's opinion stated as fact.
And that's the thing that is
harder to spot and that I think
executives and boards are
kind of not attuned to.
You almost need to be
like a critical journalist
or a McKinsey level kind
of partner or analyst that
can almost x-ray a piece of content
and just look for these things.
Yeah.
So some of the things that I
kind of think are a bit more
insidious and harmful.
So yeah, opinion status
fact or weak attribution,
weak data points,
but AI is presenting something
ultra convincing, ultra polished,
and you get fooled by that
because it's so substantive and
well articulated that the
human brain just intrinsically
thinks that it's robust
from bottom to top.
Also, where maybe the
AI is citing something,
there is a citation of sorts,
but if you go and follow that,
it's not an
authoritative source.
It's almost like the equivalent
of pulling a citation from the
Daily Mail rather than
the New York Times or the
Washington Post.
But hey, there was a citation and
AI presented it to look strong.
Or a smart marketing team that
have got a good handle on their
Reddit strategy or
LinkedIn strategy,
and they're following
the playbooks
that are created by the
AI tools themselves that
essentially are just saying
the more authority signals that
you've got out there,
the more likely it is that
people will find your product.
So it does feel a little bit
self fulfilling in that regard.
Yeah, that is a great
example, you know,
where AI has generated
something and it's kind of
evidential proof.
It's really a product on
the internet somewhere that
produced some research to stimulate
their own market and product sales.
It's not anything authoritative,
it's not peer reviewed,
it's not anything that somebody else
would take as being a real
indicative source
of information,
but packaged up into the
really shiny research output
or deck that AI has produced,
it's really easy to miss these
things unless you've almost
tuned yourself to look for them.
Yeah.
I think the it's
sort of version of em
dashes and load bearing
statements that is creeping its
way into business case
building and decision making in
organizations, that that feels
like that's where the danger is.
Yeah, it's going to be really
interesting over the next few years.
That idea of kind of big
organisational decisions,
a big business case,
a big investment,
maybe open up in a new country
or start a new product line.
AI will be at the heart of
all of the thinking in this.
And it will be really interesting
to look back in ten years'
time and try to do some
analysis of whether the quality
of this work and the decision
making was increased or whether
there was a decrease.
It'd be really hard to measure,
but I think if humans aren't
almost taking the right steps
and looking for the
right things here,
I could definitely believe
there might be a decrease in
the, almost like the
validity, the robustness,
the accuracy of those decisions.
And that could have really big
impacts on organizations and
industries and who knows,
maybe even economies.
So I just think it's just an
interesting moment where we all
need to pay attention to
how to get the best from AI.
Yeah.
Well, and speaking of things
getting stupider then,
organisations starting to
track things like prompt volume
or this idea of token maxing,
encouraging employees
to just use AI, right?
That there's no sort of desired
outcome other than just we need
to get people using AI and it will
inevitably lead to good things.
That feels like a really
flawed basis to make a case for
whatever it is that
they're making.
Yeah, been really
interesting to observe that.
And you know, on the surface
it makes sense, doesn't it?
You know, we're all trying
to build our AI muscle,
we're all trying to get
more leverage from AI,
therefore if we put some
organisational measures in
place that encourage
and incentivise this,
surely there'll be an
organisational gain.
You
know, I think we're kind of off
on the other side of that now,
but it's interesting.
The thing that made people
change course was more kind of
the changing economics of AI and
the fact that if you are
encouraging whole spread use,
just throw AI at everything and
don't worry about
the consumption.
That was okay in the
previous era of AI costs and
economics when so much of it
was just per seat and you could
have predictable costs.
But now all of the major AI
vendors have switched to some
form of consumption model everybody's
token costs have ballooned,
then you know it's almost been that
that's sort of led to a rethink.
But I
think there was an immaturity
there in many organisations.
And I've seen it in clients
that we work with where
the measures that they have
looked for have just been kind
of so off base compared to
the organisation strategy.
A great example is,
I've talked to many clients
where they need to report on
how many AI agents are
running in the business.
But that is not a metric
that is necessarily
going help you understand
whether AI is helping the
organisation work better.
You could just spin up a
thousand inconsequential AI
agents in ten minutes, but
hey, you've gamed that number.
So it's another place
where, I don't know,
I think it's easy
to get it wrong,
but surely the answer is,
don't you just need to almost
like hold dear the fundamentals
of the business?
What makes this particular
business in this particular
industry kind of go faster and and kind
of be more efficient and work better?
You know, probably you need to
get beyond revenue and EBITDA and
financial measures,
but every organisation should be
able to assess its own progress.
And that would be a far more
effective way of doing things
than how many
tokens you're using.
Yeah.
There is the because
everyone is just
focused on adopting
AI right now,
so they want to see the success
in it and measuring the AI
success score.
As you've said there, right,
so we business have had
the APIs already in place.
So it's focusing again on those
numbers and making sure that
it's evolving in the direction
it should be evolving.
I think as we continue
to go through the era,
I think the level of
maturity will increase.
I know, you know, I also
think a couple of things.
Like, I actually think there
is a positive aspect to this.
So it's often the
sort of less technical
c suite, board, investor,
something like that
that's putting these
metrics in place.
And I think there's an element
that is natural if you're not
super close to the technology and
you're not thinking about the mechanics.
And so I always think it's
natural that some of this
happens and nobody should be blaming
folks for thinking like this.
I actually think the underlying
kind of push and attention and
desire to seize the
opportunity that AI presents in
rewiring a business, in kind
of developing something that
wasn't possible before,
I actually think that underlying
drive is really positive.
And sometimes it's good to
sort of push back on individual
parts of it, but I think the
underlying intent and what it's
trying to achieve
and the fact that
boards and CEOs,
they're ultimately trying to
agitate their people and their
organisation to do better.
I actually think there's
something really positive in
that, even if it doesn't always
translate to a metric or a
measure or a kind of
tactic that makes sense,
but this should be a dialogue.
I kind of was on stage recently
talking about the fact that we
need to converge as an industry
and a sort of set of people.
Technical people are too
far removed from what makes
their organisation
work effectively.
They don't really understand
any of the boardroom conversation
and that's entirely
natural because
they're not in
that conversation.
But how can they direct their
efforts and what they focus on
and what they're trying to
solve for in a complete absence
of some understanding of kind
of the organisational drivers.
And at the same time,
we've got business
people and leaders and
managers that would really
benefit from just building an
AI agent or two. You know,
get your hands dirty.
And I think there's this kind
of idea of crossing the streams
more, spending time
on the other side,
carving out time to, you know,
if you're on the business side,
be a bit more technical.
Try and build an agent,
understand the practicalities,
understand what's easy, what's hard
and, you know, what's possible.
But then also on
the technical side,
understand how AI can
create value in your
organization and try
and contribute to that.
Find things to automate that
are in that line rather than
the the real kind of
hobby project that you'd love to
automate in your team,
but is not gonna move
the dial on anything.
I think some level of
convergence of of these role
types in terms of
how time gets spent
would would be
helpful all around.
I wonder if that's
the reason why sorry,
why Satya posted on LinkedIn
and on social media that he is
now coding, using GitHub
Copilot for coding and or
whether that's just
marketing, maybe.
Probably just marketing.
Silly with him.
His agent posted.
Yeah.
You know, I think there's it it
still feels like we're in a moment
where and and look, Chris,
I really I have appreciated your
optimism around these things.
It's always always
been valuable to me,
particularly someone that
isn't particularly optimistic.
I feel like we're in a point
still though where you can be
maybe the CISO
in the room in these
conversations and be feeling
like you're the only person
that has identified this
initiative isn't going to work
or that there is a critical
flaw or failing in the message.
You know, do you agree and do
you think that it potentially is,
we're currently in a moment
where it's career limiting to
be one of the only
voices in the room going,
we shouldn't do this.
I think that's a really
interesting question.
You know, people that are
somehow at the the kind of
meeting point of technology
and how it works, but also,
organisational
value and strategy.
It is an interesting
place to be,
you can sometimes spot
things that others may miss.
And look, think anybody
in that position,
it's incumbent on the role
to be pointing out where there's
danger, where there's risk,
where there's something
that is being missed.
But at the same time, look,
I think if there's something
that I've learned in the last
decade of becoming
less technical and
coming into more
leadership oriented roles,
it's that leaders often
get there for a reason.
They don't get every
decision right,
they're not perfect
strategists, they're not
perfect people, but a
lot of their gut feel
and expertise and
experience and sort of
intrinsic understanding
of how things can work,
they do often think at a
higher level of elevation.
So it's some meeting
in the middle.
I think where you see something
that is clearly wrong and is
not going to work out and it's
another AI strategy that's
been founded on a completely
non robust hypothesis
that isn't founded in the
right kind of data or evidence,
you've got to point that out and
make sure that pitfall is avoided.
I think there's also a
kind of you know,
needing to take a step back and
recognise where even with that
perspective, even with that
ability to see some things,
your perspective also
may not be right too.
It's finding the balancing
act between what maybe needs a
pushback or an intervention
or a sort of flag versus
what may be,
yeah, you you may
not agree with it,
you may see some flaws in it,
but the underlying intent
and sort of, in some cases,
it's a bet that's being made.
And you may not have
made the bet yourself,
but often it can be right
for the organization or the board
or the leadership.
Thank you, Chris.
It's been really good
talking to you today.
Final question from us then.
What's one decision in your
work that currently isn't yours
to make that you wish it were?
Okay.
In terms of things that
I'm not in control of,
I think the one thing
I keep coming back to
is almost where UK
organisations are
in terms of their usage
and adoption of AI.
Look, think we're in this
period where everybody is saying
they're doing AI, everybody
has their own internal story,
and they can genuinely
point to things.
But if you were to to sort
of plot this on some kind of
chart, the level of sort of
ambition and bullishness and
and willingness to take some
bets by UK organizations,
actually, it's not there.
You know, people are
being ultra conservative.
People are tinkering
around the edges,
not necessarily
wanting to invest.
And look, I've obviously got a bias
here because we're a services provider.
We We make money
from serving our
clients and providing solutions.
But I think so many
people are holding back.
And it's right to be
circumstantial and to want
robust evidence that there's
going to be ROI and to not
just go for AI for AI's sake
or go for technology for
technology's sake.
But I would really
like to see more
bullishness, more
ambition, more bet taking,
because AI is having a really
dramatic impact in places today.
So many places you can
get it wrong for sure,
but so many places you can get
it right if you're willing to
stand behind it and take a bet.
And I'd like to
see more of that,
both from the organisations
that we work with and I think
you know I'd like to think
that we're actually doing a good job
internally, but net net I'd
like to see more of that.
Chris, great talking to you
today, thanks for joining us.
Thank you. Thanks
for the time, thanks.