Kerry here, we, we, we got to know through Claire Perry Louise's,

like-hearted leaders community.

and Kerry shared on with that, like-hearted leaders, community,

um, whole masterclass on ai, her experience of it, how she works with it.

And I'm gonna get Kerry to introduce a bit more about her journey on this,

which it, it resonated a lot with my own kind of experiments with it,

but also just gave me a real broad understanding of the, you know, the

landscape of AI and this tension between, because I'm, I'm a bit of a

tech technophile, I loved technology.

I jump into it straight away.

And also not realizing how it has this adverse impact on other people

where they feel disconnect or actually technology disconnects them.

In a way.

Uh, and so we're gonna explore, I think, some of that or that feeling.

So before we kick off, Kerry, you maybe share a bit about how you describe what

you do now, what takes up your time and energy, your thoughts, uh, and maybe

a little bit about how you got into learning, talking, sharing about ai.

Yeah, sure.

Hi.

It's nice to be here, everyone.

Um, yeah, so I am a, an AI educator and also a, a copywriter and also a

yoga teacher and meditation teacher.

So I kind of do those three things.

I'm also trying to be a coach, but that's a whole, uh, another thing.

but yeah, I've been working in the world of AI for about five years now,

so well before the world of chat, GPT, and, um, I came into AI through.

Kind of emotion of fear actually.

So someone came into, so I was heading up a marketing team, a creative, a

um, team and an agency and I, someone came in to talk to us about this

AI tool called um, Frazee, which could write emails, subject lines.

And it said that it would outperform humans 90% of the time

or something along those lines.

And I remember it came in, I just thought, oh my gosh, like

this is my career as a copywriter over with these new tools that

can outperform human writers.

And then I ended up just really exploring it.

'cause I like eCommerce.

I'm really fascinated by tech.

And I, I've always been a bit, I'm just quite curious person.

I always like gathering information.

So yeah, I just ended up getting into it through that.

And I just explored and thought, wow, this is so fascinating.

And then I met another copywriter who.

Was also an in going down a similar route.

'cause also, uh, Frazier had also been into his agency to talk about.

Um, and we just ended up setting up this really tiny little marketing

practice called Tiny Giant.

And, and our whole aim was to explore how AI could support human creativity.

And yeah, so that's where I started.

That was 2018 and then I've been working there since.

So over the sort of five years I've created the world's first AI gin.

Um, created an AI curator for the Char and Science Festival where we

trained a neural network on every tour ever delivered at the Char

and Science Festival until it was able to generate new tour titles.

And one of them was delivered at the festival by a human.

Um, created AI cocktails and crafted those for events all over the globe.

Um, yeah, and I'm now doing mainly teaching, going in and

helping businesses and, um.

Individuals, how to harness the tools.

But because I'm a copywriter by trade, I'm particularly

fascinated by AI for writing.

So I talk about that and content creation.

So in my webinars I tend to talk about, um, writing AI for text generation,

image generation, and video generation.

So I've kind of niched in that in 2018, I was literally like, I do AI and

creativity, and it was just everything.

And so our early talks back then were AI and poetry and writing and

art and music and like everything.

And now it's just like I do AI and content creation because it's so massive

now that to actually just stay on top of the tools, even in that really tiny

niche is, is almost a full-time job.

Um, and I also like to be able to use the tools.

And talk and educate on the tools that I'm using on a practical level

in my day-to-day work as well.

So yeah, a big adventure with ai.

I love it.

And my whole sort of ethos is around using it to augment our,

our amazing human creativity.

Um, so it's, it is a tool to use rather than something that's gonna replace us.

oh, there's the first question that span to mind is Tiny, giant or tiny giants.

I think what, what, where did it come from?

What is it that you're trying to communicate with tiny giants?

The,

the idea with Tiny Giant was the idea that an algorithm is such

a tiny, tiny thing, but it can actually do, have pretty big effects

and results and that's where, that's where the name came from.

Awesome.

No, it resonated for me.

Was this, this idea of, um, having more impact than, What you seem capable of.

And it speaks to a lot of, I think the, the journey that, uh,

people in our community are on.

They want to amplify their impact without burning out.

They want to be able to do more without having to spend more energy.

So yeah, curious about that feeling that comes from that.

So, it sounds like you've, you've really just explored the space and

you've refined and honed into a place where you, you, I think you

can target and deliver value that people really want at the moment.

Um, given these, the experience and conversations you've had, I'm curious

to hear what you've, what people have said have been their, uh, hopes for

AI and what have been their fears.

and for those of you listening, I would hope towards the end just getting some

nuggets of wisdom around, particularly the creative writing aspect of it.

Maybe just to share how they can potentially harness ai, but for

now, just exploring this, the, this space in terms of the different ways

that people are experiencing ai.

Yeah, so I think, I think it's interesting because I, a lot of people

come and expect it to be this silver bullet that it will just solve all

of my marketing problems and all of my, you know, it will write all of

my copy for me and it's kind of done.

Um, so there's, I suppose there's those kind of expectations, but

also hopes that maybe it will make them quicker at what they do.

And, and it can do that.

in terms of fears.

I think the main one.

So last year I ran a series.

Every month I met a, I created a little meetup for writer, ai curious writers.

And we talked a lot in that group about all things ai, from how to

be transparent with our clients to, you know, how we can use it, sharing

prompts, all that kind of stuff.

And I think the main fear was that people would lose their jobs, their,

their writing jobs in particular.

'cause it was a writer's group.

and some people had already lost some client, like towards the end where some

clients had already said, actually we're gonna use GPT, um, on a, not that wasn't

an all round, but there's a few examples of that, or where writers, where their

clients had started writing using church bt and they were asking the copywriters

to, to just fine tune it or to, you know, sort of polish it as it were.

Um, so I think there was a fear around that they would no longer be needed.

So yeah, on that line of just the concerns, maybe just bring

in Marianne's question here.

Um, and she was asking how, well, based around sitting with this,

the ethical challenges, uh, she mentioned carbon foot, the massive

carbon footprint of AI training on copyright works and being mindful of

what's going on, uh, without feeling like, uh, a buzz killer or a lad.

So I dunno, are you able to talk to that in any way?

Oh gosh, yeah.

I mean, this is a question I ask myself.

All the time and card also know this.

We've had chats about this, but I am always in between this, like, how do

I work with this tool, these tools?

And I have, I say think especially with my journey with chat GPT when

that arrived at end of November, 2022.

And just using that in all of my work and seeing, you

know, how it could support me.

So I, when I first arrived, I used it on all of my copywriting jobs.

Not to necessarily do the work, but just on a separate screen to go, okay, if

I got AI to do this, what would it do?

How could it help me?

Um, and there's been times in that period, I suppose in the last year

where probably about six months ago, where I just thought, do you know what?

I'm just gonna reject.

All of these AI tools and I'm just gonna be a hundred percent human writer.

I'm not gonna touch them at all.

I'm gonna go out and I'm just gonna be that.

And then I just thought, and then I debated that for a while

and then I thought, do you know what these tools are here?

Whether we like it or not, they are here and they do have real benefits

for us as writers, as content creators.

And, and so I'm gonna, actually, what I'm gonna do, rather than just

reject it, is just get, use them as well and as wisely as I can.

And that's, that's kind of where I've got to.

So I have deliberated all of the, the question you've asked, I think I.

In terms of the environmental impact, it's massive.

The water consumption, the CO2 emissions, they're pretty big.

And obviously because AI is being used more and more by pretty

much every industry, I, I have no idea how that will pan out.

What I think can really help is this is using the tools wisely.

So rather than just turning up and with a very random prompt, and then

just having to generate over and over again to get what you want, to

actually turn up with an objective, an idea, a solution that you're

looking a, a, a problem that you're looking for, a solution for, and then

to, and to manage, um, and really understand how to use prompting so

that when you do actually prompt the tool, you'll get, you are getting

something out of the other side.

That is pretty much what you're looking for and you're not having

to constantly just reiterate.

And because obviously there's every time we do a prompt

we're generating more CO2.

So that's, that's how I've approached it, to just use the tools well

and wisely and, and, you know.

We can reject it completely, of course, but you know, they are

here and so we have to find a way.

And I think, I think the big tech companies from what I've been reading

about environmentally, are aware of this situation and are trying to do it.

So for example, apple, um, have just launched in the last week or so, four

new models, which they put on, um, on hugging face, which is like a, a

platform for developers and what they've done, rather than releasing a massive

model like, which is what OpenAI have done, obviously they've gone from much,

they've gone through like really small models and so they're gonna bring them

onto, onto devices and I think that's quite an interesting move from Apple.

And I wonder because they also, um, because they talk a lot about privacy

and the environment, whether actually that approach of, of using smaller

models that will be more efficient.

Um, whether, you know, again, whether it would be interesting to see where

Apple takes that going forward.

So they're hoping to release, um, generative AI tools later on this year.

whether, you know, there is, I think there will be a way

round, but I don't think there's a really simple answer for it.

I think user wisely and well.

Is that only way that I can kind of get through that.

I think like a lot of people really bit treading carefully, I think.

Um, I'm definitely not an early adopter, I would say.

I feel like I'm a bit behind the curve, certainly seeing Kerry's session the

other week in terms of all of the possibilities of some of these tools.

and also.

Having come from a tech background, I think we're always curious about

new tools and how they can help us.

So in some ways, I think we've always been trying new tools to

save time ultimately, and I think these tools can help do that.

My wife's also a copywriter, so she's kind of treading, you

know, cautiously through this.

Well, and again, like you and probably others, had a bit of fear

about this affecting her, her job, but also I think, yeah, trying to

use this as a aspiring partner or a creativity tool rather than something

that's just gonna replace humans.

Um, but yeah, it feels like it is a bit of a rabbit hole and I'm

always conscious of going down tech rabbit holes and days, passing by.

So in some ways how can, Mike, I'm curious about how can, uh,

ultimately learn from others like Kerry who've tried these things

and applied them in different ways, but um, yeah, and there's just

so many out there, isn't there?

And, um, you hear about, someone asked a question, I think it was

Ian, about Claude, and there's obviously the ones we know about.

So I'm guessing it's a full-time job just keeping track of all

the different tools that pop up.

We use another one for video creation as well, which is useful.

So, But again, you're thinking maybe there's a better tool out there.

So there's always a fear of there's a, yeah.

New shiny object on the, on the market.

Yeah, I think that's, yeah, that's true.

I say that in my, um, in my webinars.

I just actually, it's probably better to choose two or three tools and use

them really well and keep mastering them 'cause they're always updating.

So you're almost better to do that.

And obviously you need to keep an eye on what's going on.

And sometimes I'll dip in, but probably, you know, maybe six,

eight months ago I was like, oh, there's another copywriting tool

and another one, another one.

And then eventually you're just trying out loads of tools and not

really getting great at any of them.

So I've, I've really narrowed down my thing, but yeah, I didn't, I haven't

seen the question on Claude book.

Claude's a a great tool, um, rapidly developing and, um, yeah,

I'm often resorting to that, but I tended to default to chat GPT,

which I think a lot of people do.

'cause that was the first one to arrive.

But Claude is, is, is brilliant and developing yet great speed.

And I think from a writing point of view, Claude is, is excellent.

Um, probably a more natural sort default copywriting tone

than something like chat GP gt.

Yeah, that's what what Ian mentioned was he's enjoying the

tonality of it.

Well, I, I was, I'd like to go back to just talking about the tools, uh, how

to use them and, and the different ways.

Kind of the more technical side of things, but just coming back to

just the, the landscape and the, because I think one of the things

we, I, I was hoping to talk to was the resistance that some people have

around using the tools in our work.

Helping people just even start new businesses, do something differently.

We can give them all the tools and tactics they want, but if there's

an inner resistance, usually down to a belief or some values or some,

then it's, it doesn't, you, you, you'll never use it because you

feel deep down something is wrong, um, or something is not quite

right or aligning with who you are.

And so I feel it's really important to have spaces like this where we

can just talk a bit more openly about how we feel around this.

And even that whole idea of, uh, environmental impact of

using these tools every time I.

I type a question into one of these tools.

It's going to a server, it's processing setting that it is pulling

that, and that's burning up energy.

And on one hand it's like, oh, how free can I be?

Like even what that speaks to, and this is like an example of a drawing

or sketching, it's like I would get worried that I'm you wasting paper

because I'm drawing something and that's wrong and I'd hate that idea.

But then without being able to be free to just waste paper,

scribble, stuff like that, nothing creative will come out.

So it's interesting that tension there between being conscious

about how much we use and also having the freedom to create.

On, on the flip side of that, I'm, I'm hopeful that technology

will help technology get better.

Like how could having a tool that could optimize energy efficiency, learn how to

process things faster with less energy.

Maybe that is that it will help, again, depends how people push

it, push us down that route.

And I, I think for me, linking back to this thing about what's, what's

inherent within technology and Kirsty was asking this idea of like, what's

your view, uh, on, uh, the approaches and to the biases inherent into ai, uh,

and when it comes to change making work.

So it is, we want to use it to enhance the way we work, but also how does

that align, I assume, to our why?

And in case case, she works with women's health and sport in

spaces where everything is new.

And so for AI to be useful, it involves a decent amount of training.

So like, what information am I getting?

Where's that information coming from and is that gonna serve me or is it biased?

Yeah, I mean there is definitely bias within ai.

It's.

If you think about something like chat GPT, it's essentially, I mean,

we don't really know what it's trained on, what we gonna do, that it's

trained on the internet and therefore it has absorbed, as it were, biases.

This is why I think, you know, the AI tools and they can do really

useful things to help us, which I'm sure we'll talk about shortly.

But I think it's always really important to have a human in the mix.

So at, at this point in time, um, we should be really sense checking

everything, looking at everything, applying our own views, um, assessing

our, using our critical thinking.

So I think we have to be aware that yes, there is bias inherent within AI tools,

but we have to then look at what's out, what's come out, and then apply our

own learning and, and change and sense check and make sure that we are not

perpetuating any, any of that on the other side, on the, out from the output.

That's the only way I think we can work on it.

I, you know, it's gonna be a long time before those biases

aren't, aren't in there.

If, if, you know, it might never be the case.

I remember going to a conscious tech conference in Egypt many years

ago, uh, and someone was presenting this idea of AI is like being a

tiny child, young child growing up and whatever you teach that child,

whatever that child's exposed to, is gonna mold how they behave.

So there's like a, there's a responsibility we all have, it feels

to make sure that child is educated well, not just on how to create link

baked titles to get more reads on your LinkedIn post or whatever it is

you wanna put out there, your emails.

Um, we wanted to move on to this idea of, well, you were saying

before you kind of, oh, I heard you pulled back, so I'm not sure

if I wanted to work with this.

Then you said, no, it's here.

I need to reengage with it and.

Well, I heard you talk to before in our conversations, and I think also when

you were presenting, it's like you are also someone who, while you work with

tech and you work with, uh, with ai, you also work, uh, in the real world.

You work with people, you know, you with your yoga or women's circles

and there was a tension there.

It is still attention.

Um, yeah, so I do these two very different things where

I'm working in the world of AI and educating on AI tools.

But I also, I'm also a yoga teacher, so I help people to move on and

map and just move their bodies in like really intuitive ways.

And I host women's circles in the woods where we sit around

and we write with pen and paper.

So I'll do a lot of journaling prompts, a lot of writing, um, in the woods.

Uh, often we're barefoot and we are absorbing like all

the wonderful word of nature.

So I do these two things and sometimes I struggle to, bring them together

or whether I need to, but I think what I try to do is I try and bring

that lens to the work that I do.

So when I look at ai, this is why whenever I do a webinar,

I don't just go, oh, it's absolutely amazing everybody.

I also talk about the issues with it, the implications, the

considerations that we need to, uh, adopt for me using these tools.

It it, yeah.

So I, I also am very much around.

It's making sure that we preserve our human creativity.

'cause I just really value it and I love it.

Not just because it's part of being human, but because it has so

many wellbeing benefits as well.

To actually just sit with pen and paper in the woods with a bunch of

women and just write our feelings and thought there's something so

helpful and human, and something makes us feel, feel better.

So I, so I try to make sure that although I'm using these tools,

I do also bring, I also make sure that I'm bringing my creativity

and my thoughts and I'm assessing the work that comes out and saying,

is this actually what I wanted?

And how can I change it?

And how can I bring myself to it and how can I craft it?

And how can these words sound more like me or match my

brand, or whatever it might be.

So that's how I'm currently mashing them together.

And I just hope that my, my love of human creativity and

the value that I place on that.

I can bring that into this very, um, techy, almost unhuman world and help

us to find some kind of balance.

Mm-hmm.

yeah.

No, I'm just curious whether you always hosted those circles or spent time in

nature like that with your feet on the ground, or whether that's something

you've sought out more the more you work with almost like an antidote to,

to tech to sort of get that balance.

Yeah, it's a great, it's a great question.

I'd never thought of that.

Um, it is a recent thing, so I did them all.

So I do them at the turn of each season.

So I did them all last year and I've done the first, uh, two this year.

Uh, sorry, first one I've got, got one in like next couple of weeks.

I don't know.

That's really interesting.

Have I sought it out more?

I mean, I've always been a naturey person.

Uh, even as a kid, I've always loved the outdoors and I've been in like,

pictures of me, like when I was young, just like all in one suit.

So my welly boots and I grew up just on the edge of the P District,

so I spent a lot of time outside.

Outside.

And I've loved surfing and nature and cycling and any

chance to be outside, I, I do it.

Um, but I've, maybe I have sought it out more.

I've never thought about it, but it may well be a result of that.

Maybe just working so heavily with the tech and artificial intelligence,

like maybe actually there thing that's called me to do, to do those things.

I definitely approach those things with a, with a lot more, love

and a bit more sort of fierce determination to hold onto it.

I think, you know, actually, as I said, like writing with pen and paper,

I love the fact that I can sit with people with a pen and paper and just

go, how wonderful is it that we can just sit with this basic materials and

like get out all of our thoughts and feelings without even touching tech?

Like, how wonderful is that?

So I think I've definitely got more passionate about using

our hands and being creative.

Mm-hmm.

Because I think that I worry that it will be.

Sidelined or that we might forget to use it.

what's in my head is kind of like going into nature to kind

of wash off the grime of tech.

I'm like, I've been in this, doing all this stuff on screen

and I'm, I'm just saturated with it, so I just need to, mm-hmm.

Stepping away from it.

Definitely.

Just, yeah, just a like complete opposite.

Like no one touches their phones and there's no devices.

It's just humans talking around a fire.

Like how nice is that?

The, well, there's that, there's that piece of finding more authentic

connection in, in this world where it's always seems to be mediated by a screen.

Um.

A touch interface.

I wanted to acknowledge Tom's interesting rant here.

He holds a much deeper concern for ai, and he is wondering

who's driving us towards these tools and what are their motives?

and it appears to him, it's nearly always profit driven without any concern

for the wider issues, I'm assuming.

Not only environmental, but social, social sociological issues.

Yeah, I mean, that probably is the case, isn't it?

When you think about, if you think about OpenAI was originally set up

to be open, uh, fully transparent and also it to support humanity.

And when we think about it now and look at OpenAI, they have been that,

you know, it is, we, we don't know how the tools have been trained.

There's very, a huge lack of transparency over how, you know,

what they're, what they're doing.

and in terms of supporting humanity, they're being sued, left, right,

and center by artists and writers and a whole bunch of of people.

So I think maybe it has moved towards this.

I think it probably is profit driven, especially the tools that

we're, you know, the tools, like things like Charge GPT and Claude.

And I also hope that maybe they will lead to something that's a

bit more, more positive though.

Yeah, and I, I agree with that.

I think it's a really, I think it's really important that, like you say,

I am not advocating, switching it off.

I think it is important, going back to what I mentioned earlier about

this idea of using the tools wisely and well, and this is why again, I,

whenever I do my webinars, I do cover the other, the other side, the, the

ethics, the considerations, the fact we should, when we're using these

tools, we should, um, never put in private or confidential information

that we need to sense check everything because it hallucinates, you know, we

have to be aware that although we're using these tools, they've been trained

on people's work without permission.

So again, it's, I think it's just knowing all of these and

then working out how we can use the tools as wisely as we can.

And maybe, you know, find, like, for example, Claude has, wor is working

off a, some kind of, um, like ethical memorandum of some, I can't remember

the words, a loss, but they, they have trained the model with more

ethical guidelines, for example.

And you, there's a whole blog on it if you look at Claude.

So it's also about seeking out alternatives.

So although Chae was here first and it's very much the kind of

default option, Claude are trying to do things slightly differently

with a more ethical standpoint.

So I think.

Again, shopping around deciding what's important to you, but also

yet we have to be aware of all of these additional things, around it

and find our way through somehow.

And just, I think the regulations will make a difference.

So we've got things like the EU AI Act, which is the first like massive piece

of, um, legal, but yeah, legislation.

Legislation.

Thank you, Carlos.

Um, which is, which is coming into being and I think things like that,

you know, that's the, that's their leading the way, but that's gonna

happen everywhere and it's gonna, because at the moment it's a very

unregulated territory and think, and people can just kind of have a go and

do what, do what they want, but mm-hmm.

I think as the regulations become more, it will be closed down a little bit

more and we will hopefully see a more ethical ways of doing, doing the work.

the ethical aspect of this for me is really interesting.

on one level.

Well, I, I think of it as this choice.

How do we choose to interact with this?

And it feels like we could either just ignore it, you know, to unt attach

ourselves from it, just reject it, and then it just continues on its own

without us knowing what's going on.

Mm-hmm.

We could unconsciously engage with it and use it for our own benefit

and our effectiveness and efficiency to pro promote ba basically

push forward our own agendas.

And then there's this kind of more difficult place to be in where you're

trying to understand how it works, which is sounds like a nightmare

because, well, it is a nightmare.

It's so complicated.

It's so vast.

There's so many things.

but unless we try and get our heads around it, how do we

engage in the conversation?

How do we influence the legislators?

' cause unless we, you know, if we, if we are, we believe we have

agency in this, unless we vote with our behaviors, our voices.

Other people will make decisions for us.

So this is an interesting space of like, how do we engage with it

in a way where we actually push it forward in the right direction without

becoming overwhelmed by everything?

And also, we run busy lives.

We're trying to make a living, we're trying to just get by.

So, yeah, I, I feel that's, I dunno the phrase, conscious

capitalism springs to mind.

Mm-hmm.

some of the tools we already used that maybe 10 years ago we were skeptical

about, like Google apps, for example.

I mean, you know, whose business isn't on Google now

and they're all free products.

Well, they're free for a reason, aren't they?

Mm-hmm.

And that sort of privacy issue doesn't seem as big a thing now as ai.

So I, I'm not saying it's the right solution, but the ethics

behind so many of the tools that we use aren't all rosy, are they?

Mm-hmm.

Not to mention Amazon.

And I think this, is this what I am, you know, what I was getting

from talking to you, Kerry, is this consciousness aspect of it.

You know, how aware we are when we're stepping into new tools.

There's a need to make our lives better, to work better, to do

things better, but at what cost?

And this is where I think this conversation, for me, my interest

in this conversation is how do we, how do we make ourselves aware of

the cost through just sharing our own thoughts and ideas around this?

I'm not saying carry that, you have all the answers, but at least by

being for me, someone who is immersed in it and questions it, I think, I

believe that's a much healthy place to be than being just a, an advocate.

Say, yeah, it's great.

Let's just move forward.

Which I think Tom is wary, wary of is like, whose agenda are you pushing?

Yeah.

So, um.

if we were gonna paint a rosy picture saying that

if we were

gonna, if we were gonna paint a rosy picture of using ai, what,

what does that mean to you, Kerry?

What, how would that work in a way that you felt was aligned

to your beliefs and your values?

Yeah, so in my, I mean in my very small niched area of content creation,

I think it's super, super helpful.

So it's, it's really great for helping us come up with ideas.

It's like, as, as Laurence said earlier, like a sparring partner

or a brainstorming partner.

It's absolutely, it's absolutely brilliant for that.

If you ever, I just don't think you ever need to have the whole

blank page syndrome anymore.

If you dunno where to start, you can use AI to kickstart

thoughts or ideas or words or.

Social media posts, whatever.

So it's, it is great for that.

So that's one of the things that I use it for.

I think it's really good for research.

So there's quite a lot of ways that you can use for research.

So it's great at summarizing documents, for example.

So if you're ever doing research where you need to read.

Loads and loads of white papers or big reports or PDFs, AI's great

at summarizing those for you.

So I don't ev advocate for never reading a long document again.

Um, but what it can do and what I really like is it can, it can

summarize something for me to decide whether I then go on and.

Actually spend loads of time, 'cause I've done this before in my copywriting

world, where you also have to become an expert in your topic just very

briefly while you write for that client.

and so I would read loads and loads of papers or big research

documents, whatever it might be, and then I'd get to the end of this

very, very large document go, that hasn't actually helped me at all.

So brilliant.

You can summarize this document and go, yeah, that's actually gonna help me.

So I'm gonna really invest my time in reading.

I know there's a lot of people that now just summarize

everything and maybe don't read.

I don't advocate for that.

Um, but yeah, I save you, save you lots of times from that point of view.

So it's, it's really helpful for things like research.

I also use it for first draft, so again, I don't ever get AI

to write a whole thing to me.

Uh, probably just 'cause I love writing and also 'cause I don't

think it's quite there yet.

Um, so I use it to maybe do a structure for me.

So things like a structure for a blog or a structure for a social media post.

Um, it's absolutely brilliant for doing that.

And then you can obviously, um, write in the InBetween bits and

fill it all out how you want.

if you wa if there's things you wr, like for example, there's certain

things like, I'd be very happy to hand over to ai, for example.

I don't really like writing email subject lines.

I just, it's like, oh, here we go again.

Um, so there are some things where I'm like, do you know what ai,

I'm very happy for you to help me write my own subject lines and

I'm not gonna stress about that.

But when it comes to writing a blog or expressing something that's

really important to me, I wanna write as much of that as possible.

So I'll use AI to gimme a structure for that.

Or sometimes I'll write from scratch and not touch it at all.

Um, but yeah, structure, it's great.

I think sometimes coming up with a structure for a

piece of work, it takes ages.

So it's, you know, that again cuts the time.

And I've then spent more time, I've got more time to

spend on the, on the crafting.

So.

there's a lot.

It's super, super helpful in that point of view.

Also, things like image generation.

So I also do image generation and video generation with ai.

If I'm looking for something really, really specific.

So I, I, sometimes with my newsletters, I'll do AI generated

images 'cause I want something that's exactly right for that article.

So rather than searching through loads and loads of, um, stock

libraries to try and find something that kind of fits, and then you'll

see the same image on someone else's blog like two months down the line

as a completely different subject.

I can generate something that's exactly what I want in the style that I want.

So really good for things like that as well.

So yeah, I'm not, I mean there are lots of issues with ai, but it's also

can save us a huge amount of time.

Makes us more productive.

Like with chat gpt, it's also got a voice.

Um, you can also use its voice technology so you can.

Plug your phone in, put it in your headphones, like go out

for a walk and brainstorm.

Literally back to back, like back and forth with, with, um, church, EPT.

And it's, it's like if you wanna test out an idea or you want someone

to argue against you, it's really good for all of that as well.

So, yeah, so much it can do.

I, I mean, I wouldn't be here if it didn't do good stuff as well.

I love, um, the idea of it giving, giving us space to craft.

Yeah.

And that, and it's that idea of keeping the craft as opposed

to just outsourcing everything.

Um, it's interesting that you, you, you went onto the topic of being

able to just go back and forth in like a nearly like a conversation.

And in the past, I, I've come across people, I, I think it

was an article or a post around.

AI coaches or AI therapists.

And I'm just gonna allude to Julian's question here about have

you explored AI as a means of introspection or psychological play?

So is there any have, yeah.

Have you come across it being used in that way?

Do you see any benefits?

Do you see any issues with it?

Yeah, it's a really interesting question.

I, I read in a newsletter really recently about an AI

therapist that you could go to.

I think I just, oh, I, I struggle, I struggle with an AI therapist

because I, because of, because the hallucination, so ai, it doesn't

always make things up, but it has the propensity to make things up.

And so you never a hundred percent know what's true and what isn't.

And I think I'd worry in that, in that way, in how it would, what kind

of answers would come back in the, and also if it would accidentally

advise things that wouldn't.

Really help you or if it would tell you to do something that would then be

detrimental to your health or wellbeing?

I, I've, I've got slight issues with that.

I dunno enough about it, um, to know, I dunno if I, I'd

need to speak to a human.

I, I'd want someone who was super well, when it comes to things

like my mind, um, you know, I've had therapy in my time.

I would want someone who's really well trained and who knows what they're

doing because if they give me the wrong advice or take me in a route

that is not fully thought out, I dunno whether that, I think that could be

really detrimental to our wellbeing.

in terms of introspection, I've used, I, I have used chat gt to ask

all sorts of things from like, what should I do with my business too?

Um, you know, to being a devil devil's advocate, like, I've got

this idea, I've got this viewpoint.

Like, can you argue against me?

So I've, I've done that as well.

Where, so it will just go, what about this, what about this?

And I'm like, oh yeah, I hadn't thought about that.

So it's, so, it's good for that.

It's good for exploring your.

Own thoughts.

I don't put anything person like super personal or your

like, name or anything in it.

So be quite anonymous.

but yeah, you can use it to explore things and ask its viewpoint.

I think it's quite, I think it's quite useful for that.

But yeah, I wouldn't ever rely on it as a therapist personally.

Have you ever used P Pi?

Is it pi.ai or pi.ai?

No, I haven't.

That's another, that's another tool that's, yeah.

I, I haven't but

another one for the list.

Another one for the list.

That's the thing.

It's like, 'cause Carlos mentioned perplexity actually to, to me,

a few a month or so ago and I've actually been playing with that.

I just, I just sort of occasion you'll go, actually, do you know

what I'm gonna go and play with that and see what I, what I can do.

But yeah, I'm just trying to stay in my little zone for the,

for the moment and just, yeah.

But I will look into it.

It's on my, it is actually on my list to look at.

So yeah,

I'd like to, towards the end, maybe just run through some of the tools

that you do use and why you use them and, and what's helpful about them.

and yeah, Anya was recommended pie.

AI from a friend in terms of a, a compassionate therapeutic

AI conversational partner.

So, mm-hmm.

So, yeah, I, there is a, what sprang to mind immediately around that is re

re relating to someone else's comment about biases and what biases are

inherent in the system in a sense.

Anyone we talk to has biases.

So ai, why would AI be different?

Because it's gonna be exposed to a certain amount of information, but yeah.

yeah, I would be worried how people would be accidentally triggered by

the ai, but maybe that also happens with, uh, human therapists as well.

Yeah.

Maybe.

Um, you talked a lot about hallucinations now.

You used the word hallucinations a bit.

Yeah.

And so for, for listeners who aren't aware of what that means, do you

wanna just elaborate a bit more?

Yeah, so it, it's a, it's a term given by.

Computer scientists to a point where AI essentially gives information

that's not factually correct.

So I suppose these tools have been trained to be helpful for

us and therefore it will often return things that sound very

viable, but are not actually true.

And so I, that's a, a major concern in terms of.

The spread of misinformation and there are are tools now

that are much more helpful.

So actually, one of the reasons I really like perplexity, thanks for recommending

it, Carlos, is that you, it, it's, it provides sources for its answers.

So if you want to go and check something, you can click through on

the, on the source that it's, um, not necessarily it's pulled from,

but then it can sort of correspond with something that's online.

You can then go and have a look at it and go, okay, yeah,

that's sort of backed up.

And I, I get that.

Um, and also Google, Google Gemini also have a similar thing where the

little, you press the Google button and it will tell you where they've

sort of cross-referenced it with Google search and say, yeah, this is, it's

green because it, it, something's backing up what this statement says.

So that, I think that's quite useful.

But yeah, we just have to be aware of that, which is why I still say whenever

you generate any kind of copy or I.

Words or whatever it might be, just to just double like sense.

Check it and make sure that you back Just fact check it.

Sense check it and fact check it.

Yeah.

This relates to Kim's question or point.

Uh, she once asked Chachi pt, who's the founder of her business, right?

For girls, and it gave her a very believable detail profile

of someone who was not her.

So yes, I think that reinforces the whole fact checking thing.

I, uh, I, uh, I did a similar thing around, um, I said, who, who is Laurence

McCahill and who is Carlos Saba?

Interestingly enough, it, it did have did a pretty good description

of the work we did, which made me assume it is just checked our

website and regurgitated it out.

But uh, I was a panel conversation around AI here in Brighton.

Yes.

Uh, one of the panelists said they did a, they asked who, who was

them, you know, basically put their name and it came up as some kind of

serial killer or something like that.

So.

Don't like with Google, don't trust everything.

I assume you, you read.

so for the last bit of this conversation, I'd love to just get

a bit more of your knowledge, tips, um, thoughts around ai, uh, the

tools you use, how you use them.

Maybe starting off with this question from Kim, tips on how to instruct AI

as it seems it's all in the asking and I, uh, the same panel conversation.

I, um, I learned of the job title prompt engineer, which was like, so

maybe talk a little bit about that.

Yeah.

So you are right.

There's a lot that comes in the asking.

so you could just turn up to a tool like chat, GPT and say,

please write me a blog on.

Wellbeing in the workplace, and it will generate a blog for you.

Um, but it'll probably be pretty generic and not very detailed, or

not really include what you want.

So I, I always say to spend some time actually really thinking about the

prompts and what you're asking for.

So as a general rule, just work out what problem it is

that you wanna solve first.

So actually before you even turn up at the tool, just take some time to work

out, like, why am I using the tool?

Do I need to actually use it?

Um, if I do need to use it, what's the problem that I want it to solve?

And then being really, really specific about what you want the tool to do

and how you want it to help you.

So having a very clear and very concise description of what you need.

And then describing a bit of context.

So actually context can really help in your prompt.

So yes, you can just ask it for a very direct, just give it a very direct, um.

Instruction.

But if you provide a bit of context, so for example, things like if you're doing

marketing, who your target audience are, why, why you wanna do it, what

kind of tone of voice that you want to adopt, then what you get out in the

other side will be much more in line with, with what you're looking for.

So we provide a bit of context, be really specific.

Um, also describe what you want the output to be.

So do you want it to be, in bullet points?

Do you want it to be a hundred words long, et cetera.

So just do you want it to be a table?

So whatever you wanna do to describe that as well in

your, what your output to be.

And then I think a lot of it also comes down to just refining.

Um, so you'll get something back and it's like, it's not quite

right, but then you'll refine your prompt and then you'll find

something that works for you.

And then once you find something that works for you, you can then

kind of save it in a little prompt.

Library somewhere and then, and then use that one again if you know that it works

well for your use case or your clients.

So yeah, it'd be pretty specific.

Another thing that I've got is this, um, is this methodology called the AI

sandwich, which I share quite a lot.

Um, and it's the idea of that I think the best results come from

a combination of human and ai.

And if you'd have asked me about six months ago, I would've said,

use the AI to generate the copy or your whatever it's you wanna do,

and then use the human craft to.

To really like add something to it.

And then probably, yeah, about six months ago I was like, actually what we

do before we touch the tools is just as important as what we do with the output.

So for me it was this combination of human AI and then human again.

So this this AI sandwich.

So that means before touching the tools, we come with an idea of what we want.

We've got an objective.

We spend time crafting our prompts so that what we get out the

other end is kind of worth it.

Then we use the air tools and then we craft, and then we sense

check, and then we fact check.

Um, and then we go, is this actually what I wanted?

And if you've already done that initial bit of like, this is my

objective, you could also analyze your output better and say, yeah,

this does meet what I wanted.

This is, this is kind of where I was going.

Or actually, this isn't where I was going at all.

This isn't the kind of piece of work I want.

So okay, what, you know, what do I need to do?

And you might have to go back and start the sandwich again.

But for me it's this combination of AI and human, I think it's.

One, it preserves our creativity and keeps us thinking and using it,

tapping into our amazing brains.

It also means that we can use the tool as well.

And it also means that what we get out the other end is in line

with our voice and our thoughts and our wishes and our objectives.

Um, and that it means that we sense check it, make

sure it's not made stuff up.

So that's a kind of general view on what you put in.

And I think actually just putting in your thoughts and what you want.

Like if you, if you're doing a blog on whatever it might be,

tell it what your angle is like.

Don't just ask it to write something.

You decide what, how you want it to be written and what the angle is

and what you wanna get out of it.

Tell it what you want to see in the blog.

You know, I want it to include this, this, this, and this,

because it's your blog.

And yeah, these tools are amazing 'cause it'll speed up the process for you.

But just make it, just make it yours and preserve who you are and

what your brand wants to achieve.

I am, I've got into the habit of uh, just recording my thoughts,

just speaking them out loud, just as a, around stream of consciousness.

Then transcribing that and then asking chat GPT to write this

up into something that's useful.

And then using that as inspiration to then write something and uh, and

sometimes it's just like, that's all rubbish, but it's given me an

idea to do something different.

Mm-hmm.

So that process of, ah, actually, do you know, there's something else

that sparked, I'm curious about.

So there's this aspect of being really being more specific or

being more clear about what you want from, let's say, chat, GPT.

And I know on chat GPT, you have the possibility to create

your own gpt as they call it.

Would you be able to just describe briefly what that means and how

that talks to this idea of getting more specific content for you?

Yeah, so the, so the chat, GPT, so the GPTs inside chat, GPT,

if you're pay, you have to be paying for it to access those.

It's, they're almost like an equivalent of something like the app store.

I think that's what they're kind of working towards.

But these are tools that are much more specific, so they

have a very specific role.

So it might be something around.

Keyword generation or like, maybe SSO is not a great example, it's probably

not great for SEO, but it might be a thing around how to like, create perfect

prompts or, that might, yeah, there might be a tool on, on copywriting.

So, and what it does is it piggybacks the, the large language model,

but it's just much more precise.

And if you're training your own one, you can add information into

it to make it really specific.

So for example, I've been training one, I still can't get it to quite a

hundred percent work, but I've trained it on some of my, my own writing,

um, some copywriting guidelines, so what makes great copy and, and

what not to do and what to do.

And then I've, I've inserted all that information into A GPT so that what

it generates is much more like me.

Um, so yeah, there's, there's a lot you can do with that.

They're relatively new on the scene, so there were plugins until actually in

April they've sort of stopped that and.

Got more into the gpt, but you can train your own, but there's

also loads and loads on there that you can go and explore.

Um, you can do everything from like generating stickies to, to like

asking about SEO or Yeah, fine.

Like fine tuning your prompts.

Loads and loads on there.

But yeah, yeah, much more specific use cases rather than the, the

whole massive model that does kind of pretty much all round stuff.

I was just thinking about, well, A, can you spot AI generated content

generally, and also b, talking to the fear of this, allowing more and

more people to generate con AI driven content, which then, you know, a lot

of people in our community might be a bit resistant to putting themselves out

there just because they've got some.

Belief or block about, you know, sharing their ideas and then seeing that there's

more volume being created that might create a bit more inertia as well.

So I don't know, is there anything you could talk to there in terms?

Yeah, no.

So I think it's adding to the noise, basically.

Um, I could spot, yeah, I think so I got this, uh, it was just on LinkedIn,

but someone, you know, you sometimes get those like InMail things from a.

From someone who's trying to sell you something.

Um, I got one the other day and it's like, that is

clearly written by Chat g pt.

And so it is interesting because I think because I, it was so clearly

chat, GPT and it hadn't been like edited at all with that person's own voice.

I was, I just dismissed it immediately, which is quite interesting because I

think if it had been written any other way, I probably would've gone, oh,

who's this person that's messaged me and what, what do they have to say?

But it's like after paragraph one, I was like, this is totally being generated

and therefore I just switched off.

And I thought that was quite an interesting response because I guess

I thought because it was so explicitly chat to be generated, I kind of, I.

um mm-hmm.

And, and outline some of the, Didn't engage.

And so that's, I think that's the danger.

And this is why, again, the AI sandwich thing that we need to

take this generated con content and use it as a starting point.

This is like a, a base or a, a thought starters.

And then we go in and make our own.

Because I think there's gonna be so much more generated content now

and people are getting wise to it.

I wrote a blog, actually, it's on my site, which was how to not sound like

chat, GPT, things like phrases that it used, like the word Dell for example.

It comes up all the time.

And things like inner world of, and things like dynamic and tapestry.

You know, there's certain words, synergize.

Yeah, exactly.

These words.

And people are getting wise to that now.

So it is also like, I must now make sure that I don't ever use the word

delve in my, that makes me think

that this is one reason to not run away from them, is so that

we can get more wise about.

Yeah, that looks like this kind of content, otherwise we wouldn't.

Yeah, I think so.

And in the blog I said as well, you know, we have to, as human writers

now, we have to work almost harder.

You know, we need to make sure that we have this really colorful vocabulary.

We need to make sure that we put in an anecdotes and use

storytelling techniques and all the things that chat GPT can't do.

Like, this is what we need to do so that people go, oh yeah,

this is human written and I'm kind of gonna engage with it.

But yeah, it'll be really interesting to see.

I think, yeah, I'm getting better at analyzing it.

I think someone said I could spot an AI generated job

application, which is, yeah.

So people are getting wise to it.

Uh, I'm really feeling sorry for the person out there who's written

something really genuine from the heart.

From them.

Yes.

And someone's like, that's ai.

I use the word delve.

I synergize.

Maybe their tone of voice just sounds like chat.

GPT, that's just who they're

Oh, that's who they are.

Yeah.

Um, Tatjana was asking do the tools fall into categories?

So he, she heard BA bad, AI was good for academic research

and there's general chat GPT.

So you able to offer some, just maybe say this tool, if you're thinking

about doing this, check out this tool.

If you're thinking about doing that, check out this tool.

Yeah.

So I mean, they don't necessarily fall into categories.

They're ones that you've mentioned.

They're like chat GPT and Gemini, which was bar and

it's now Gemini just in case.

'cause those, both those terms are being used.

But yeah, bar is now Gemini.

They're kind of real all rounder tools actually.

So they don't necessarily fall into a category.

But if so, in terms of writing, I would go to Claude.

So I've just got to know them well enough in that

I know which ones I go to.

If I wanted to write, say a social media post or I wanted to structure

a blog, I would go to, I would go to Claude because I think it's writing

default tone is, is much more human and slightly more conversational

than something like chart GPT.

Although if I was doing like for a formal letter, then

I'd probably go to chart GPT.

Um, I really like chart GPT for ideas and brainstorming

and kind of explorations.

So I go that if I was doing research, I probably now go to perplexity just

because we get the citations and it's quite useful to go and check those.

So I think it's just, I think something like Chap GPT is a great

all rounder, but there are certain tools that would suit you better.

And actually Gemini is quite good for research as well because it again

has a little check check device.

So.

Yeah, that's, that's, that's what I'd say in terms of image generation

I would go for, I'd always go for mid journey, which I know is not

necessarily the best one for advice.

'cause you, you have to pay for it from the word go, which

is quite unusual in that most.

people give you like at least a free, like couple of free tokens or

something, but you can't, I think it, it generates consistently good

images and I, I really like it as a tool and it's developing all the time.

and then in terms of video generation, and you can also

generate image assignment as well.

Um, I really like runway.

Um, it's called Runway ml.

Uh, I go, I go, I go to that.

I use that in my workshops 'cause you get free tokens with that one.

And they've, they, they've been around for quite a long time, but they did,

originally it was around creativity, but they, they tend to be specialized

a bit more now in video generation, which is, which is really interesting.

So those are the tools that, there's also Dali three, which

you can use through chat.

GPT paid.

So if you wanna generate images within the chat GPT model, uh, 'cause

it's super easy, you can use that.

Um, and Dolly's is good too.

I just still not gravitate towards Mid Journey.

I just think the, the outputs are just slightly better.

So.

Those are my main tools.

Um, as, as Lauren said earlier, there are literally new tools coming out every

week and it's hard to stay on top of them all, but that's the ones I use.

Brilliant.

Well, the, the kind of environment that keeps an AI educator in a work.

Yeah, yeah.

No, it's true.

And it's like every, whenever I'm, 'cause I write a newsletter every two

weeks and it's like every time I'm like, I wonder, I wonder this gonna

be enough for me to write about.

It's like, oh yeah, there is literally,

Well, that, that's, uh, that's great because, um, if people wanna

follow you and learn from you and, and sign up to your newsletter,

where would you wanna point them?

Yeah, so my website is kerry harrison.io.

That's my website, my blog's on there.

So the one around how not to sound like chart, GPT is also on there.

Um, and my substack is called Minds and Machines News, uh, on Substack.

So I couldn't get Minds Machines, someone had already got it.

I really wanted it to be called that.

So, um, and in my newsletter every two weeks I talk about AI news and updates,

particularly around content creation.

So I tend to talk about image generation, text generation,

video, um, and what's coming next.

And then the other half of it is all around the human side of things.

So, I'll share like nice books that I've read or hu lovely human craft

examples, um, and talk about like women's circles and whatever else

I'm up to from that point of view.

So it's a kind of minds and machines being, that's why I called it that.

so that's what I do.

So yeah, sub start would be great.

I'd love it if you could sign up for my newsletter, that'd be amazing.

Um, and yeah, my website's probably the best place to go

if you wanna see what I'm up to.

Any time to write a book?

I'd love to write a book.

It's actually on my list of, it is one of my, like, intentions for 2024.

Like at the end of the show I want to have a, an outline of

what I'm gonna write it on.

But yeah, my ideas

writeabook.ai.

Yeah.

So maybe I'll, maybe I'll,

that feels like a natural next step given the Substack and we work and I

dunno, it feels like there's a real blossoming uh, body of knowledge here.

Yeah, no,

I've always, always be coaching, Laurence always be coaching.

And also I just, I just wanna fast track as well, just like,

I just wanna read the book.

It's

easier

selfishly just write it for me.

But maybe you wanna write completely different book, nothing to do with ai,

but if you do write the book, can I have the, can I have the audio version?

Laurence can read it.

No worries.

Cool.

so to finish off, is there anything that you would love to

just leave our listeners with?

So apart from following my stuff, I would say to explore the tools

just on a, just to really, if you've never touched them before and

you're quite nervous just to start somewhere, like even just asking

chart GPT, what to have for tea.

Give it what you've got in your fridge and it'll, along with

something for you, just using it and just seeing how it feels.

Because I think this is the thing as well, it's not just, it's how it feels.

It's for important to, for me, like what, what does it feel

like for me to ask it to write?

And then it gives me something like, how do I feel about that text?

And what can I, what can I bring to it?

So I definitely just say to explore the tools and see what's around.

Um, and then I'd always just say, which I say at the end of

everything is to just preserve and nurture your human creativity.

And don't stop thinking and don't stop creating.

Nice.

What are you, uh, leaving with?

I think wanting to engage more in the conversation.

Certainly listen more into just the conversation around

ai, AI more generally in terms of just the bigger picture.

'cause some interesting comments from there, from people in the chat.

Um, and also, yeah, I think leaning more into some of these tools, I

think it's, um, yeah, I just get a bit overwhelmed, probably like

a lot of people with this stuff.

So you're like, ugh, where do I start?

And then, you know, gonna close it and go, go back to my normal

habits and then seeing like, oh, there are ways to simplify things.

I, I think I'm always conscious of like the time saving tools end up, you just

end up using the time to do more stuff.

It's like, how to use Yeah, that's true.

It like, like you more balanced and actually free up some more time to

then spend more time sitting in a circle in a field or in the woods.

So,

I have embraced AI as a sparring partner these days.

I find it such a helpful, I, I, I think to talk, or I, I, I create

through interaction more than anything, rather than just sat

on my own trying to do something.

And I, I found having something like chat GPT to just throw ideas at it,

see what comes back, and then that process of back and forth has helped

me just get into more creative space.

And I'm conscious of what Tom said in terms of what are we being pushed to do.

Mm-hmm.

What does it act?

What's the impact of using it more?

Um, and how is it being trained?

What information am I.

Basically being led by because of how it's, designed and built.

So, yeah, mixed feelings, but with mainly underpinned by optimism.

So thank you Kerry.

Really, really appreciate this.

I think it could go on so much longer, but, um, this has been super valuable,

at least for me, and I, I assume for everyone else who's been listening live.

So, um, thank you everyone.

Take care.

Have a great rest of your day and yeah, um, try out chat GPT and ask

it what you can have for dinner.