October 7, 2025 • 36 mins
AI isn’t magic, it’s a system that works dramatically better guided by your thinking.
In this episode, Colette Mason, founder of The Tech Leadership Method, joins us to cut through the noise around AI and talk about what really matters when it comes to using it responsibly, sustainably, and effectively.
Drawing on her experience helping businesses build real AI systems (not just shiny demos), Colette explains:
Why most AI projects fail before they even start
The difference between AI hype and AI that actually helps your business grow
How to spot “AI slop” and the right questions to ask before hiring anyone claiming to be an AI expert
What leaders should do to ensure long-term, human-centered success with AI
How critical thinking, governance, and system design make or break AI’s value
This conversation dives deep into the mindset shifts needed to embrace AI—not as a replacement for human thinking, but as a tool that enhances it.
“AI isn’t a magic wand—it’s a tool that makes your thinking visible.” — Colette Mason
If you’re a business leader, manager, or strategist navigating the AI landscape, this episode is a must-listen. It’s your guide to staying smart, skeptical, and strategic in an AI-hyped world.
About Colette Mason
Colette is the Anti-Hype AI Advocate who builds human-centred systems that work under real-world pressure.
With 40+ years in IT and 3½ years implementing AI in business, she's the truth-teller cutting through Silicon Valley fantasies to create AI colleagues that enhance human capability instead of replacing it.
Author of "Ditch Rework Build Teamwork," Colette transforms chaos into collaboration using her proven frameworks for sustainable human-AI partnerships.
You can connect with Colette on LinkedIn and cleverclogsai.com.
Mind Blend is all about real experiences and meaningful conversations—through sharing these insights, I hope to inspire and empower you as you navigate your unique journey.
I’d also love to hear from you! What resonated with you today? What topics or guests should we feature next? Let’s keep the conversation going—connect with me on LinkedIn @ karenkchong and @Mind-Blend
Follow Mind Blend to get notified when new episodes are released!
Until next time, stay curious, keep exploring, and let’s continue to blend our minds and discover what’s possible.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Have you noticed how every company these days wants to use
AI? In this episode, I'm mind
blended with Colette Mason, an AI systems expert who is seen
behind the curtain where shiny promises often hide broken
systems and wishful thinking. We dig into what actually makes
AI useful, why it's less about automation and more about
(00:20):
clarity and systems thinking, and how the real danger isn't AI
replacing us, but us forgetting how to think.
If you've ever wondered how to cut through the AI noise and
focus on what truly drives impact, this one's for you.
Welcome to the show Mind Blend. I'm your host, Karen Chong, and
I'll be diving into the minds ofincredible people, each an
(00:44):
expert in their very own way. Together, we'll uncover insights
and share ideas so you can be inspired and empowered to
navigate your own unique journeyin life.
Ready to get curious and discover what's possible?
Let's blend. All right.
Joining me today right here in London is Collette Mason.
(01:06):
Collette is a human centered AI advocate who loves to fix things
that everyone else breaks. So the reason why Colette and I
connected was over human centricity.
Now, as many of you know, I'm a project management long timer
who is passionate about making it work for the human.
Like the process needs to work for the human, the results need
(01:27):
to matter for the human. And Colette is here to make sure
that we are using AI to work forus, like to work with us, not to
replace us. And we're going to have such a
fun conversation today. Hi, Colette.
Hi there. How are you I'm.
Very good. Thank you.
Yes, very, very good. So Colette is someone who has
been in computing for a long time, but let us hear a little
(01:51):
bit more about your journey and how you got to what you're doing
today. OK so I got my first computer on
my 12th birthday and that was written when Adam and the Ants
were in the charts with Ant music, one of the first Adam and
the Ant singles before Duran, Joanne and Rio.
Before that I got my computer and I was an only child and I
(02:14):
found it was good fun. It was entertaining, it was
stretching and I got quite interested in computing.
So much so that by 14 I'd got myfirst job working for a local
business, printing out labels ontheir packaging on a really
noisy dot matrix printer with, you know, it's really, really
awful. And I got even more into it.
(02:34):
I went to university and did economics, which was nowhere me.
Too. Yeah, really wasn't as
interesting as computers. So since I left, I started
working with them. I had my first permanent job
which just lasted for nine months.
Not really. It's out for, you know, 20 years
in the gold watch. So I tried lots of different
(02:55):
types of computing when I was a contractor and I worked for the
government, Metropolitan Police,investment banks and lots of
different. London Underground was another
thing. I used to look after 10,000
e-mail accounts. I did lots of different forms of
corporate IT and latterly I'd say around about 2006 I got in
more into usability, accessibility and what it felt
(03:16):
like to use it. And a few years later I'm a
pretty bad certificate junkie. So I added in being
hypnotherapist and a master NLP practitioner.
Oh wow, you. Have skills and that got me
thinking about, you know, personal development and how you
can phrase things in a particular way will make
somebody feel different. And I saw quite a connection
(03:36):
between usability and how a therapist might work with
somebody and creating a nice environment, a nice space or a
nice flow to get things done andmove on.
And so I thought AI is a really good combination of all the
benefits of tech and it behaves itself.
And. Making people feel good about
using computers because the natural language means they can
(03:57):
talk to it like they couldn't before.
That's how I've ended up here, yeah.
Well, I'm so glad we have Someone Like You who can see
both sides. Nothing is more frustrating when
you get this error message that doesn't mean anything to you at
the human. Yeah.
And so how long have you been inthis field?
(04:18):
I know AI hasn't really been widely available.
I would say about 3 1/2 years using it every day, most of
every day. I didn't feel I could tell I was
doing it to other people becausea, they probably think I was
selling crypto coins the week before.
And also I have to know something works before I could
say to somebody else, give that a go.
(04:40):
So I thought, well, you know, 2,you know, three years at that
point was long enough of a test bed for myself to think I can, I
can talk to other people becauseduring lockdown I started to
write women's saga romance booksbecause I was used to doing
really dull documentation for Fujitsu and people like that.
So and so writing something really dull and boring that I
(05:01):
wasn't interested in wasn't a hurdle for me.
So lots of stories about Victoria Mills were made and
written and that because that was so tedious, I needed the AI
to help me write that. There was cause lots of
challenges with plagiarism and Ididn't really care if it had my
voice or not because I would never say that sort of stuff
anyway. But continuity and errors and
James turning into John halfway through the book or something.
(05:23):
So I had to learn how to make itaccurate and fast.
And I said that was that was theother.
That was my testing ground really.
And then I thought I could turn this to any sort of business
really doesn't have to be fiction or books specifically.
So that's what led me to to there.
So it's mainly been the content creation and collaboration,
plotting that sort of thing rather than creating cinematic
(05:44):
videos. And the more, let's call it
faked end of the spectrum. That wasn't really what I was
interested in and it was really around, you know, three years
ago have the choice. So so that's what I've been,
it's sort of augmenting what I was able to do.
I was like create big documents and I needed to make it more
enjoyable, which which it was did the collaborator.
(06:05):
So you are using it as your assistant and at the same time
testing it out. Yes, because you you see people
now on Amazon, they've uploaded their book and they've obviously
put the chapter through and it says this chapter is stronger
than the last version you gave me.
And that's ended up in the finalbook on Amazon.
So I was like, that's not great.Yeah.
So. So there were lots of things you
(06:25):
had to do to make sure you didn't shoot yourself in the
foot because when in an audienceis hard work and the last thing
you wanted to do was drop the ball with a silly mistake.
And I see the same thing with businesses using it for blogging
and whatever you know, little mistakes creep in and then you
can't have too many of those before your credibility is
undermined. Yeah, it's as bad as just
(06:46):
getting an e-mail that says hi Nick Bracket name.
First name, Yeah, yeah, yeah. So what do you do now?
Like your surface offering is very different now as I
understand. Yes, I've, I've written a book
about AI and how people interactwith it because I thought it's
really more like a team member that sits opposite you the new,
(07:07):
because when I was contracting, there was a constant stream of
new starters. And it was this is the folder
where we keep the things and that's where the canteen is and
lunch is this time. And, and so I, I saw it the same
thing with with AI. So I've worked on a book which
has formulated my take on AI andalso a course to go with that so
that people can then build theirown assistants that dovetail
(07:30):
into their own lives. Not just saying I want you to
write blog posts for me, but a little bit deeper.
So a lot of the people I work with run their own businesses.
So they might like, for instance, if you said to me,
write a WordPress plug in, I'm like great, no problem.
If you say to me, come up with anew brand message and a colour
palette that will resonate with the audience, OK.
(07:51):
And then if it's something really on the copyrighted end of
the spectrum or write a 50 e-mail sequence, I'm like, Oh,
no, I could probably do it, but it's going to be awful.
And so I thought everyone's got the same problem.
So I asked them what they're strong and weak at.
And then I build something that works with them and then
challenges them when they're strong and then lifts them up if
they're, it's scaffolds, if they're struggling.
(08:12):
So that's, so I interview the person and then build the
perfect. So it's not a model of them.
It's a bit more than a model. It's, you know, and also I
builded some personality as wellbecause one thing from working
at home on your own business is it's pretty low, lonely and you
suffer, you know, motivation. You're second guessing yourself.
Is this good enough? So I build in some things as
well to get rid of those problems that people have.
(08:34):
So that's nice. I can do that digitally with an
interview. Or.
Sit with them depending on how much smoothing down they need
and I can sit with them and interview them directly if
they're particularly nervous andthat's the next, next level up.
So that's how I decided to to goabout it rather than giving
people yet another prompt pack. You know, you want to be that
(08:54):
person because it's, it's not really, it's like lending
somebody else your glasses. You know, it sort of works.
You can see pretty small print on the balance of medicine or
something, but it doesn't really, it's no good for all day
vision, somebody else's prompts.It's just not accurate enough.
My thing is about making it really about you.
(09:16):
Mm hmm. Yeah.
And the prompts are it's not just plug and play.
A lot of the times you have to give it feedback.
Yes, you have to train it. You have to correct it just to
stop all the hallucination. So like hallucination is
something that I just recently learned what AI hallucination
is. Yes, and it happened to me the
(09:36):
other day. Yeah, I was writing something
for a project. It just made things up and I
said no, I didn't do that at all.
Could you keep it factual? So I was surprised that I did
that. Actually.
I don't know even where it got that information from.
I was, yeah, I was working on a news story yesterday and had
exactly the same problem and theinformation was correct and it
(09:58):
and the sources that it he commented on the correct.
But when I went to Forbes to look at the thing it got
inspiration from, I got a 404 message and they look, if I
hadn't clicked on that, it looked, it looked completely
correct and the, you know, the query string looked really
accurate. I thought, yeah, this is it.
And it wasn't, it was really wrong.
So it still still happens. You have to have an extra
policing level in to have another one, say, go and go and
(10:21):
click on that and see if that works.
And if it doesn't, it's not going in to get around that.
But it is. It's people.
I think people wave through about 40% of content without
checking it properly because they think, oh, the last three
were OK #4's going to be OK. And then sometimes it's not, you
know, it's a bit of a pain, but it is what it is.
You just have to, you know, you have to check it.
Yeah. Yeah, I see that especially with
(10:42):
image generation like the first time was say 80% somehow the
second one is 50 and something like there's a third leg.
Yeah, definitely. Yeah.
What do people usually come to you like when they first
approach you? What do they normally come to
you for? Usually shortcuts so they can
(11:03):
drink pina coladas on the beach.They, they, and the people who
come to me for that, most are usually the people who hate tech
and they, you know, they've not worked out a simple way to make
it slightly faster already. So that processes are very
clunky and they're not very technical.
And if you're going to automate things because they think, oh,
well, I'll press a button and itwill work.
You need really good tech skillsto you've got to work out how to
(11:27):
describe it and then you've got to police it, like we just said
about the hallucinations. Then you've got to put that
information somewhere. There's so many linkages and
it's so complicated. And if you can't even do, you
know, formulas in Excel without breaking out into a sweat, then
it's never going to work for you.
And an automation rate, it's going to be a perpetual
disappointment. So that's so I have to sort of
overcome that. And then after that, it's just
(11:49):
like doing little bits of tasks,like I need to do a weekly
newsletter. So help me write the weekly
newsletter. But again, that's very narrow
thinking when you think of how much it can unlock your
potential and take the brakes off you in terms of motivation,
knowledge gaps, that's what theyshould really be thinking about.
But you don't really see that. It's just like, put this prompt
in and it'll help, you know, fixthe SEO on your about me page.
(12:12):
And it's really micro tasks and they get them then.
So you have to take the blinkersoff and say, what if we could do
it like this and have fun while we do it?
So looking for shortcuts when the base is not really say for
example, the process is not clean.
Yes. What are some of the other
things that you think, Oh, this is?
(12:33):
You're misunderstanding what AI really is.
Well, I see, I think some of themisinformation comes from people
who feel like they're threatened.
So the writers, the video editors, they're saying, you
know, you might have 6 fingers on one of your hands or it's got
M dashes in it and all those things.
They're trying to like diss it when there's and they're sort of
(12:57):
making people think, oh, it's never going to be good enough,
but it could be good enough if you gave it the right rules.
So that's, you know, I see a lotof criticism from and then on
the other end of the spectrum, you've got to use this and you
know, I, I'd like faceless AI toa certain extent.
I think that's, I don't think everybody has to sweat blood
into like throwaway material or something that's only valid for
a week rather than so when I washelping people write books, you
(13:21):
know, I wrote a 10,000 word piece because I knew somebody's
going to come every week and want to know about that.
But if it's just a throwaway comment, there's a new model
this week, then why not just have AI doing that?
It's not really, you know, flagship material.
So yeah. So I don't really like people
completely replacing themselves in terms of trust and whatever.
I don't think it works. I think COVID taught us that we
(13:42):
really need human connection, and I think AI is going to teach
us we need it even more. We're going to need that
feeling. We're going to latch onto that
feeling even more than we do already.
Yeah, there's something that's called being the human in the
loop. When it comes to using.
AI and I'm definitely in that camp of I trust my team yeah,
(14:05):
but I'm still gonna just make sure everything works because
the more people you involve the bigger the room for error.
Maybe it could be hand off. It could be they understood
things differently. Yes, I like actively try to
avoid throughout the project, but still like it's just an
extra set of eyes. And I know we can build in like
QA functions, yeah, in like codewriting for example.
(14:28):
Yeah. But as of right now, I guess
like everything that we're saying today is like as of
today. So as of right now, people
shouldn't or I understand that certain fields may be
eliminated, but if we think about it like this, maybe those
fields should have moved on already.
(14:48):
Yeah. Yeah.
So I definitely see why people are afraid that they'll be
replaced. I also see people using AI in a
way that sabotages themselves. Yes.
Yeah, so we're it's really important that we have this
conversation now because it's we're in the very, very, very
early stage of humans using AI. Yeah, thing.
(15:12):
If we don't get the approach yeah mentality right, it could
fall into the wrong hands and we're doing ourselves some
serious damage. There's a lot of, I see a lot of
parallels between software and, and how websites work with AI.
You know, they've got to be, youknow, when you click on a link,
you're always taught to make your hyperlink be meaningful and
never click here. It's like go and view our, you
(15:33):
know, go and view our pricing and you click on the pricing
word and you go to the pricing page and there's all these rules
that apply to normal software that need to apply to AI.
But because not everybody's had AIT background, you know, they
don't often understand that. So I was doing somebody'd set up
a wonderful, you know, GPT and it was, it was going to ask me
some questions and analyse me, give me an audit.
(15:54):
I was on question like 26 and there was no progress bar.
And I thought, how many questions are there in this
thing? You're like, will I ever get to
the end? Will I be 90 when I get to the
end of this? And in the end, I was just
clicking anything just to see how long it was, which
completely defeated the person'sobject.
I used to apply to filling in AIforms on SurveyMonkey, they'd
say you're on page three of five.
And so I think all those good things that used to work with
(16:16):
other software needs to come into the model.
So I think even if you don't come from an IT background,
think about those things that make things more enjoyable to
use and you're more certain whenyou're using them because that
will help your AI muscles, I think.
Yeah, the this age-old saying garbage and garbage out.
That still applies if you give it data that's not consistent,
(16:39):
that just the wrong thing is going to come out the other end.
Yeah, yeah, it's quite hard to remember what you've trained it
as well. So I try and put my like with
ChatGPT, you can only have 20 files, so you can't silo it too
much. But it's nice to think, right,
well, I've got an avatar in there.
I've got my pricing model in there.
I've got my back story in there depending on what task I'm
(17:01):
doing. But you soon forget what's in
there. And as you know, as a business
owner, you have to, you think, oh, that's a better way to
describe what I do. And you move on and the training
gets left behind and. It's.
Job to keep up and it's a tedious job to keep updating it,
you know, to make it way rather speak this month, yeah.
So maybe that's your next next project.
(17:22):
Yeah, auditing training materials.
The course kind of focuses on training materials because I
think the first thing is describing the role and the next
thing is, you know, the big thick employee manual you've got
to give it to make sure it understands how you go about
things. What's your take on it?
Particularly personal brand businesses, because it's your
take on something that make is your USP.
It's not the knowledge that's out there everywhere in books
(17:44):
and already it's your take on it.
And so you really have to make sure it models your values and
voice, which isn't too hard to do once you work out where to
hit it with a hammer. But at first it's like this has
nothing like me. This sounds like somebody on
LinkedIn that I would avoid, like the blade personality.
Do they ever have one? I sound like that.
(18:05):
Yeah. Yeah.
But it gets it gets easier. I think it's getting better at
understanding what people reallywant.
You know, the the people who provide the platforms are
understanding how people want tointeract with it a little bit
more. I saw an interview with Sam
Altman who's saying, crikey, allthe 20 year olds won't do
anything until they've asked their ChatGPT, what's the best
thing that I could do? And he's thinking it's not built
(18:25):
for that. And he's probably the donation
and biased data and all those. Things.
Should you really be relying? I mean, it's brilliant for
asking. I love using it as a sounding
board. You know, it's probably the most
because I have a lot of ideas and I like to battle test them
before I act on it. And yeah, but there are people
using it in all sorts of ways that the platform makers weren't
expecting. And some of that's good.
(18:48):
Some of it's a bit like, I thinksome, I saw somebody saying my,
my AI boyfriend proposed to me after 5 months of chatting and
there's a video, it's chap and he's sort of looking at him.
He puts his arm around her shoulder and she's going great.
Thumbs up on the camera, you know, which is really quite
bizarre. I'm not sure what I think about
that. So I think that's maybe a little
(19:09):
bit too far. I mean, everybody had an
imaginary friend as a kid, but this is like a bit.
Not sure what I feel about that yet because it's quite new.
Yeah, it's more than imaginationnowadays.
Like we have been actively making everything in movies come
true. Yeah.
So it's funny you mentioned the complain about 20 year olds
asking ChatGPT everything beforethey act because I saw this
(19:31):
video, obviously. So just a joke.
Yeah. So they were asked the 2
characters, they were asking ChatGPT.
One guy's in the room. He said there's someone at the
door, what should I do? And then the ChatGPT said, well,
it might help if you like, go at, go to the door, see where
they want. So things like that.
Yeah, definitely think that's one way to sabotage yourself.
(19:54):
It's offloading all thinking to not necessarily just AI, just
offloading all thinking. And sometimes I wonder if we
keep doing that. It's sort of like people who
don't know how to do maths at all.
You have only used a calculator.Yes.
Will you know that this Uber that's fit four people can fit 2
(20:18):
couples? Yeah.
Will we ever get to that point? Yeah.
So what's your take on that? I was, I was talking to somebody
yesterday about that, thinking about phone numbers and we don't
really remember phone numbers anymore.
We don't need to. And I thought like, if I lose my
phone, you know, I'm going to lose because I don't always back
the SIM up and everything even, you know, So I think, well, then
(20:38):
what's the next few things we'regoing to forget?
And I don't mind forgetting facts.
I'm not too worried about because I think we've had that
for a long time looking at Google.
But when you start to, you know,your, your reasoning centres of
your brain and neuroplasticity and all that.
If they start to weaken, then that's not a, that's not a good
thing because we need to police what what the AI is doing is.
And I, I see, I wouldn't say badactors, but I can see my friends
(21:00):
going, oh, I'm going to answer all these questions so that AI
will then quote me as an expert when, when the next, you know,
when it gets scraped again in October.
And I'm thinking what, what other what other information is
being pushed into there as training for people who may, you
know, that's quite a nice, you know, being discovered if you're
a bona fide consultant is great.But what if you're a scam artist
(21:21):
offering financial services and like you've cloned another
company and, and to all intents and purpose, you would look like
you were them, but you're not. And I think that's, you know,
what, what people put into it for their own gain.
I do wonder. Ideally in say three years, how
would you want people to be using AI?
(21:44):
Oh, that's a very good. Question ideally.
Dearly it definitely wouldn't behaving a LinkedIn bot talking to
a LinkedIn bot. So they've got visibility in the
comments. I know it's not that I think I I
do think it there's a space for it to make you feel good about
yourself, motivate yourself. There's only so many times you
can say, you know, there's that phrase ask Cole, isn't there?
There's only so many times you can ask a friend about something
(22:05):
that you're struggling with before they think, oh, not this
again. And I'd like to see people have
a place where they can, you know, offload their thinking
and, and get some feedback on things and give them some
confidence. And also to take them on and try
new things and, and, you know, let them know it's going to be
all right if they try something new.
(22:26):
I think that's what I'd like to see the most.
Lots of distilling all the information that's out there.
I think there's going to be so much overwhelm and so much so
it'd be nicer to have better ways of, you know, to sort of
make sense of all the information.
They went used to think like 10 years ago.
There's like a million years worth of YouTube stuff.
And I don't know, it's an amazing amount of time would be
(22:46):
on YouTube a day. And there's just there isn't any
way to keep up with all that. So I think it'd be nice for AI
to help you find your niche interests and find the the best
people, the most trusted people,as long as they haven't been a
bad actor for getting reviews onTrustpilot from their mates to
show they're credible. But that's another story.
But I think, you know, everybodyshould be able to use it as a
(23:07):
sounding board, I think, and notjust know what's the top three
places to go to New York on holiday, but more meaningful
things than that. Things that you can not just
search. Yeah, and I think looking at
what it does for older people, so some people have got, I won't
say a name, but they've got the a word at home.
And that's just really automation.
It's not, it's not AI at all. But there's a tool called Leq
(23:31):
that they, it's an American thing.
And they put it in older people's houses and they didn't
feel lonely and it would check their phone came if their phone
was near range, they'd know they'd come in from a day out.
And they say, oh, how's it talking for you, for older
people living on their own? I think that would be really
good. And, you know, have you taken
your medication and things like that?
I think more things like that and that all my Internet of
(23:53):
Things almost, it would be, it would be a really good direction
to take it in to support, you know, people who are on their
own. Yeah, it's really nice.
Like I haven't really thought about three years down the road
because now we have learned to live in a world we don't really
think long term that much anymore.
Yeah. So if we come back to present
(24:14):
day, I'm sure there are a ton ofbusinesses who are the bosses
are like less do AI, but that's really the mandate for
businesses who are who want to get on board.
What would you recommend to be for them to start doing?
The first thing I would do is build a light version of your
assistant. So if you've been building these
(24:35):
things, that's going to be your system prompt, which says you
are. It's a bit like the person CV,
the assistant CVI want you to dothese things for me.
And then give it some core training such as, you know,
what's your website URL? What's your social media links,
what your customers like? What problems do you solve?
And put, if you, once you've gotthose two bits in, then you can
(24:57):
start saying I, you know, I, these are some of the processes
that we do. These are the results we get.
How do we, you know, what are the opportunities here?
Or how do I, you know, does our website explain what I've just
told you, what we do and sort ofstart using it as a sounding
board and then from there start building in little bits.
And there was a study from MIT this week that said 95% of
(25:18):
corporate projects are being unpicked and unrolled because
they've tried to shove in AI andit's just been a money pit and.
I read it. Yeah.
So I think people need to just slow down and think a little bit
and then get the and then the quicker you've got that good
framework, like we're saying before about everybody should
have universal AI. Once you've got that core bit,
(25:38):
it's a lot easier to then speed up.
But if you haven't got that first bit right, it doesn't
work, doesn't work quite so well.
So that's that's the first thing.
And then I'd work on core training and the wider thing, so
wider version of your avatar andhow your marketing works while
you do what you do. I think that's really important
because that's how it gets your voice.
(26:00):
I've got a friend who calls it the hill you're prepared to die
on. So you're AI was want to die on
the same hills as you and fight for the same causes.
And then it will get your voice better.
So that takes a little while to get it, get it together.
And then after that you can add in, you know, different like
customer facing sops or whateveryou're going to be doing from
there. And that's, that's kind of how
(26:22):
I've gone about it. I using the audio things to chat
and Mull things over. That's been quite good.
And I've not known what I'm doing or talk for an hour and
then say, make sense of this because I have no idea what I'm
doing and it's all on autopilot,like breathing.
I can't describe what I'm doing.Please look at this waffle and
turn it into something that you can follow.
And I that's, you know, that's quite helpful as well, doing it
(26:44):
that way. And yeah, keep it.
Just keep it on top of what you're building and don't build
too fast for now. Yeah, good advice the and I like
how you start like inwards, likelook at what you as a business
do. What do you want to do?
One of my I've got like 7 different ways that people use
(27:05):
AI and one of my people is The Escapist and we were talking
about them wanting to automate everything earlier.
With the Escapists, they want toreplace themselves.
So I give people an example and I say, let's think about making
a cup of tea. And it's, it's fairly, so you
just go into my kitchen, make a cup of tea, there's a kettle of
tap and off you go. And I say to them, describe how
(27:26):
you'd make the cup of tea. And they'll probably come up
with like, let's say 8 steps or something like that.
And I could probably think of 20steps.
I get them to in the course. I say, right, go and go and
explain to a mate, come into your house, they're staying
while you're away, how to make acup of tea.
And then I said things like, youknow, if there's, if there's low
milk in the fridge, you'll have to, you know, put that on your
(27:46):
shopping list. Otherwise, you have no milk when
you come to make a cup of tea later and they don't, you know,
all those wider issues are quitedifficult to do when you're do
it so instinctively. And it's so the actual steps.
It might be easy to write two sides of a four, but to explain
how you make a cup of tea, but it's takes quite a lot of
thought to make sure. Is the cup on its side when you
pour the water in? Is there enough water to cover
(28:07):
the elements? Is it over the Max thing?
Is it switched on at the wall? Did you use a gas kettle?
And now on an Arga, there's so many different parameters and
people don't, they just don't give it enough rigor.
And then they say, oh, this thing doesn't understand me.
And then they sort of, you know,punch their screen, almost
stopping an inch away from the screen, you know, and they get
really fed up. But to be fair to the machine,
well, you've not really given itwhat it needs.
(28:29):
And unfortunately, that bit's really, really boring.
Processes aren't boring. They will continue to be boring
until the end of time. If you haven't got processes,
fine. But that's not the point.
It's, you know, if you're going to make a cookie cutter, you've
got to have the right shape. Yeah, yeah.
Like the espresso machine, the the pods.
It only does it because there's enough programming in it
(28:51):
already. Yeah.
And it went to stop. Yeah.
So I, it's funny, you mentioned the seven AI user types.
Yes, I was going to ask you about that.
Do you remember all seven of them?
We talked about escapists. Yes, we've got the juggler and
the jugglers. Somebody probably like a virtual
assistant or customer service person, they've got new calls
(29:12):
and angry customers and there's quite a lot going on and they go
from one task to another. And because the platforms want
AI to look helpful, they will always be suggesting other
little bells and whistles. Every time here's another bell,
they'll hit a home run and you think, oh, I'll go and do that
and use up a bit more my subscription.
And so for the jugglers, that's pretty fatal because they want
(29:33):
to do a good job and they never quite finish anything because
the next thing comes along warriors.
I'm, I've put all seven of thesein me.
The warrior is somebody who willjust go on and on and on and on
and on until it's finished. And the AI will be saying, let's
turn that into blog post. Oh, we could make that into a
Canva graphics so we can then mark it.
And then these, these things make sense.
(29:53):
But you find it's 2:00 in the morning and you've still got
some more productive work to do,which isn't good for work life
balance. It's great if you're a nerd, but
it's not really good for work life balance.
The next one I've got is the Tweaker and they're sort of
writing blog posts and they'll go, that's 97%, right?
Let's throw all that out. And doing this was probably
before Canvas came out, you know, and they'd throw it all
(30:13):
out and then, well, that's 97%, right.
Let's do it again, let's do it again, let's do it again.
And so for them, they need to besay, well, 97% is good enough.
You know, when I used to help people with their non fiction
books, it would just sit on their hard disk and never get
out there. So that's 100% wrong.
So let's read as a perfectionist.
It's really hard to go, oh, it'sonly 95, but you know, it's got
(30:36):
to be out there if it's going tohelp people and you're 95 and
you're an expert. Is somebody else's 2000% good
enough? And we forget that.
I think then I've got visionaries and they'll start
off and let's say you're you're writing your book with AI and
then you think I could turn thisinto a course.
This could be a two week course and we'll do it in the Maldives
and it'll be a bit of a retreat with the right sort of people.
(30:58):
And then the book gets forgotten.
Then they're they're thinking, they're doing the syllabus for
the, and how we pick people up from the airport.
They're doing this new, this newproject has kind of sucked all
their attention and that they'renot looking at.
So for those people, they need to train their AI to go, oh,
hang on, you've had a light bulbmoment here.
Let's document the light bulb and then we'll go back to what
we're supposed to be doing. And when you finish that, as a
(31:19):
reward, you can go back to thinking about your workshop in
the Maldives. The hermits are people who don't
have a sounding board. So you might think, should I
raise my prices? Shall I change my website host
to a new one? All those kinds of strategic
decisions. And you've got nobody to ask.
So you go to Reddit and you go to Google and you to WhatsApp
(31:40):
and you go to a when a Facebook group on the webinar.
Yeah. You need somebody to say that's
good, go ahead and do it. And the last one of the
architects, again, I'm a chronicarchitect.
And so they will be building a system.
Let's say they're building a sanctuary for Swans.
They build the system to make sure all the white Swans are
happy. And then they'll think, oh, what
(32:01):
about black Swans? They might get hot in the heat.
Let's like, hang on, we'll add in the black Swans.
And then they might think, well,it could be a green swan
somewhere. I've just not seen it yet.
And so the green Swans get looked after and they keep
building in extra features that never that may not happen to
edge cases. They'd have been better off
launching with an MVP and saying, sorry, Black Swan and
green swan people, we've built it for white Swans now to check
(32:23):
it works. And that's that's quite
difficult to do as an architect.So again, that's where that
mental support bit comes in. And you know, this, this is
going to help all the white, allthe white Swans are going to
love this and think of the millions of white Swans are
going to be really happy with this idea.
And so those are my, those are my seven ways that people kind
of get in a knot with AI. And you just need to, in the
(32:43):
training, just sort of tell it I'm prone to not finishing
things off or, you know, get well, getting everything 95%
done, which isn't done. So you just tell it what your
weakness is and then hopefully it can pick up on that.
You have to invite me now and again.
You do remember I'm a juggler, don't you?
And then it will often and help you again.
So yeah, yeah. And I am all 7.
(33:05):
Like we we have more than one inUS for sure.
So for those last bit, you want to play something fun?
OK. So rapid fire questions, gut
reaction. Yeah.
OK, great. What would you not use AI for?
Picking this week's lottery numbers.
(33:26):
Theoretically, no. I'm the vision.
That's a visionary of me. What do you do when you're not
looking at a computer? I go to gigs, music gigs.
Last year I did 164, this year I'm on 124 already.
I'm only in August, so well, so I'm going to, but I'm not
(33:46):
counting next year because there's no way I'll I'll get to
like beyond 180. I'd have to like live, live in,
you know, Hammersmith Odeon or something.
So no, that's it. So yeah.
Well, listeners, you can't see how wide my eyes are right now.
What's the most annoying aspect of your AI colleague?
(34:10):
I think hijacking my, one of my principles is hijacking and so
he thinks he's being helpful, but it's 1% off every time and
then before you know it, you're not hitting the target.
So I don't. I don't like him being over
helpful and over sycophantic really.
OK, what's more frustrating? AI hallucination or people
misusing AI? Oh, people misusing AII think
(34:32):
that's much harder to fix. You know, unless I got like
people with baseball bats to attitude adjustment technicians,
as I call them, and they're, they're much harder to do than
fixing hallucination. So the last one One leadership
book you wish more people would read besides your own.
Oh, I don't know. I'll probably say the four hour
workweek. I think that was one that I
(34:54):
found was, you know, most chait put me on this path of not
certainly not being a freelance freelancer.
It put me more into the like independent freelancer rather
than the freelancer for government.
Yeah. Right.
Well, thank you very much. Like conversations about AI,
meaningful conversations about AI, not just about You use this
prompt for this. You use this to get the cheapest
(35:16):
ticket. It's very important.
We really need to have more solid foundation before we move
on to have more AI in our lives and I hope listeners walk away
with these things that like justspark something become think
twice yes, before hitting that button.
Well, thank you so much. I will put all your contact
(35:37):
information. Thank you very much.
The website and the show notes and listeners can connect with
me and Colette. Yes, yes.
Thank you very much and I've enjoyed it today.
Thank you, Me too. Thanks for listening to this
episode of Mine One. If you enjoyed the conversation,
don't forget to follow and shareit with anyone who needs to hear
(35:57):
it. And let's keep the conversation
going. Connect with me on LinkedIn or
leave me a comment. Until next time, stay curious,
keep exploring, and let's continue to plan our minds and
discover what's possible.
Have you noticed how every company these days wants to use
AI? In this episode, I'm mind
blended with Colette Mason, an AI systems expert who is seen
behind the curtain where shiny promises often hide broken
systems and wishful thinking. We dig into what actually makes
AI useful, why it's less about automation and more about
(00:20):
clarity and systems thinking, and how the real danger isn't AI
replacing us, but us forgetting how to think.
If you've ever wondered how to cut through the AI noise and
focus on what truly drives impact, this one's for you.
Welcome to the show Mind Blend. I'm your host, Karen Chong, and
I'll be diving into the minds ofincredible people, each an
(00:44):
expert in their very own way. Together, we'll uncover insights
and share ideas so you can be inspired and empowered to
navigate your own unique journeyin life.
Ready to get curious and discover what's possible?
Let's blend. All right.
Joining me today right here in London is Collette Mason.
(01:06):
Collette is a human centered AI advocate who loves to fix things
that everyone else breaks. So the reason why Colette and I
connected was over human centricity.
Now, as many of you know, I'm a project management long timer
who is passionate about making it work for the human.
Like the process needs to work for the human, the results need
(01:27):
to matter for the human. And Colette is here to make sure
that we are using AI to work forus, like to work with us, not to
replace us. And we're going to have such a
fun conversation today. Hi, Colette.
Hi there. How are you I'm.
Very good. Thank you.
Yes, very, very good. So Colette is someone who has
been in computing for a long time, but let us hear a little
(01:51):
bit more about your journey and how you got to what you're doing
today. OK so I got my first computer on
my 12th birthday and that was written when Adam and the Ants
were in the charts with Ant music, one of the first Adam and
the Ant singles before Duran, Joanne and Rio.
Before that I got my computer and I was an only child and I
(02:14):
found it was good fun. It was entertaining, it was
stretching and I got quite interested in computing.
So much so that by 14 I'd got myfirst job working for a local
business, printing out labels ontheir packaging on a really
noisy dot matrix printer with, you know, it's really, really
awful. And I got even more into it.
(02:34):
I went to university and did economics, which was nowhere me.
Too. Yeah, really wasn't as
interesting as computers. So since I left, I started
working with them. I had my first permanent job
which just lasted for nine months.
Not really. It's out for, you know, 20 years
in the gold watch. So I tried lots of different
(02:55):
types of computing when I was a contractor and I worked for the
government, Metropolitan Police,investment banks and lots of
different. London Underground was another
thing. I used to look after 10,000
e-mail accounts. I did lots of different forms of
corporate IT and latterly I'd say around about 2006 I got in
more into usability, accessibility and what it felt
(03:16):
like to use it. And a few years later I'm a
pretty bad certificate junkie. So I added in being
hypnotherapist and a master NLP practitioner.
Oh wow, you. Have skills and that got me
thinking about, you know, personal development and how you
can phrase things in a particular way will make
somebody feel different. And I saw quite a connection
(03:36):
between usability and how a therapist might work with
somebody and creating a nice environment, a nice space or a
nice flow to get things done andmove on.
And so I thought AI is a really good combination of all the
benefits of tech and it behaves itself.
And. Making people feel good about
using computers because the natural language means they can
(03:57):
talk to it like they couldn't before.
That's how I've ended up here, yeah.
Well, I'm so glad we have Someone Like You who can see
both sides. Nothing is more frustrating when
you get this error message that doesn't mean anything to you at
the human. Yeah.
And so how long have you been inthis field?
(04:18):
I know AI hasn't really been widely available.
I would say about 3 1/2 years using it every day, most of
every day. I didn't feel I could tell I was
doing it to other people becausea, they probably think I was
selling crypto coins the week before.
And also I have to know something works before I could
say to somebody else, give that a go.
(04:40):
So I thought, well, you know, 2,you know, three years at that
point was long enough of a test bed for myself to think I can, I
can talk to other people becauseduring lockdown I started to
write women's saga romance booksbecause I was used to doing
really dull documentation for Fujitsu and people like that.
So and so writing something really dull and boring that I
(05:01):
wasn't interested in wasn't a hurdle for me.
So lots of stories about Victoria Mills were made and
written and that because that was so tedious, I needed the AI
to help me write that. There was cause lots of
challenges with plagiarism and Ididn't really care if it had my
voice or not because I would never say that sort of stuff
anyway. But continuity and errors and
James turning into John halfway through the book or something.
(05:23):
So I had to learn how to make itaccurate and fast.
And I said that was that was theother.
That was my testing ground really.
And then I thought I could turn this to any sort of business
really doesn't have to be fiction or books specifically.
So that's what led me to to there.
So it's mainly been the content creation and collaboration,
plotting that sort of thing rather than creating cinematic
(05:44):
videos. And the more, let's call it
faked end of the spectrum. That wasn't really what I was
interested in and it was really around, you know, three years
ago have the choice. So so that's what I've been,
it's sort of augmenting what I was able to do.
I was like create big documents and I needed to make it more
enjoyable, which which it was did the collaborator.
(06:05):
So you are using it as your assistant and at the same time
testing it out. Yes, because you you see people
now on Amazon, they've uploaded their book and they've obviously
put the chapter through and it says this chapter is stronger
than the last version you gave me.
And that's ended up in the finalbook on Amazon.
So I was like, that's not great.Yeah.
So. So there were lots of things you
(06:25):
had to do to make sure you didn't shoot yourself in the
foot because when in an audienceis hard work and the last thing
you wanted to do was drop the ball with a silly mistake.
And I see the same thing with businesses using it for blogging
and whatever you know, little mistakes creep in and then you
can't have too many of those before your credibility is
undermined. Yeah, it's as bad as just
(06:46):
getting an e-mail that says hi Nick Bracket name.
First name, Yeah, yeah, yeah. So what do you do now?
Like your surface offering is very different now as I
understand. Yes, I've, I've written a book
about AI and how people interactwith it because I thought it's
really more like a team member that sits opposite you the new,
(07:07):
because when I was contracting, there was a constant stream of
new starters. And it was this is the folder
where we keep the things and that's where the canteen is and
lunch is this time. And, and so I, I saw it the same
thing with with AI. So I've worked on a book which
has formulated my take on AI andalso a course to go with that so
that people can then build theirown assistants that dovetail
(07:30):
into their own lives. Not just saying I want you to
write blog posts for me, but a little bit deeper.
So a lot of the people I work with run their own businesses.
So they might like, for instance, if you said to me,
write a WordPress plug in, I'm like great, no problem.
If you say to me, come up with anew brand message and a colour
palette that will resonate with the audience, OK.
(07:51):
And then if it's something really on the copyrighted end of
the spectrum or write a 50 e-mail sequence, I'm like, Oh,
no, I could probably do it, but it's going to be awful.
And so I thought everyone's got the same problem.
So I asked them what they're strong and weak at.
And then I build something that works with them and then
challenges them when they're strong and then lifts them up if
they're, it's scaffolds, if they're struggling.
(08:12):
So that's, so I interview the person and then build the
perfect. So it's not a model of them.
It's a bit more than a model. It's, you know, and also I
builded some personality as wellbecause one thing from working
at home on your own business is it's pretty low, lonely and you
suffer, you know, motivation. You're second guessing yourself.
Is this good enough? So I build in some things as
well to get rid of those problems that people have.
(08:34):
So that's nice. I can do that digitally with an
interview. Or.
Sit with them depending on how much smoothing down they need
and I can sit with them and interview them directly if
they're particularly nervous andthat's the next, next level up.
So that's how I decided to to goabout it rather than giving
people yet another prompt pack. You know, you want to be that
(08:54):
person because it's, it's not really, it's like lending
somebody else your glasses. You know, it sort of works.
You can see pretty small print on the balance of medicine or
something, but it doesn't really, it's no good for all day
vision, somebody else's prompts.It's just not accurate enough.
My thing is about making it really about you.
(09:16):
Mm hmm. Yeah.
And the prompts are it's not just plug and play.
A lot of the times you have to give it feedback.
Yes, you have to train it. You have to correct it just to
stop all the hallucination. So like hallucination is
something that I just recently learned what AI hallucination
is. Yes, and it happened to me the
(09:36):
other day. Yeah, I was writing something
for a project. It just made things up and I
said no, I didn't do that at all.
Could you keep it factual? So I was surprised that I did
that. Actually.
I don't know even where it got that information from.
I was, yeah, I was working on a news story yesterday and had
exactly the same problem and theinformation was correct and it
(09:58):
and the sources that it he commented on the correct.
But when I went to Forbes to look at the thing it got
inspiration from, I got a 404 message and they look, if I
hadn't clicked on that, it looked, it looked completely
correct and the, you know, the query string looked really
accurate. I thought, yeah, this is it.
And it wasn't, it was really wrong.
So it still still happens. You have to have an extra
policing level in to have another one, say, go and go and
(10:21):
click on that and see if that works.
And if it doesn't, it's not going in to get around that.
But it is. It's people.
I think people wave through about 40% of content without
checking it properly because they think, oh, the last three
were OK #4's going to be OK. And then sometimes it's not, you
know, it's a bit of a pain, but it is what it is.
You just have to, you know, you have to check it.
Yeah. Yeah, I see that especially with
(10:42):
image generation like the first time was say 80% somehow the
second one is 50 and something like there's a third leg.
Yeah, definitely. Yeah.
What do people usually come to you like when they first
approach you? What do they normally come to
you for? Usually shortcuts so they can
(11:03):
drink pina coladas on the beach.They, they, and the people who
come to me for that, most are usually the people who hate tech
and they, you know, they've not worked out a simple way to make
it slightly faster already. So that processes are very
clunky and they're not very technical.
And if you're going to automate things because they think, oh,
well, I'll press a button and itwill work.
You need really good tech skillsto you've got to work out how to
(11:27):
describe it and then you've got to police it, like we just said
about the hallucinations. Then you've got to put that
information somewhere. There's so many linkages and
it's so complicated. And if you can't even do, you
know, formulas in Excel without breaking out into a sweat, then
it's never going to work for you.
And an automation rate, it's going to be a perpetual
disappointment. So that's so I have to sort of
overcome that. And then after that, it's just
(11:49):
like doing little bits of tasks,like I need to do a weekly
newsletter. So help me write the weekly
newsletter. But again, that's very narrow
thinking when you think of how much it can unlock your
potential and take the brakes off you in terms of motivation,
knowledge gaps, that's what theyshould really be thinking about.
But you don't really see that. It's just like, put this prompt
in and it'll help, you know, fixthe SEO on your about me page.
(12:12):
And it's really micro tasks and they get them then.
So you have to take the blinkersoff and say, what if we could do
it like this and have fun while we do it?
So looking for shortcuts when the base is not really say for
example, the process is not clean.
Yes. What are some of the other
things that you think, Oh, this is?
(12:33):
You're misunderstanding what AI really is.
Well, I see, I think some of themisinformation comes from people
who feel like they're threatened.
So the writers, the video editors, they're saying, you
know, you might have 6 fingers on one of your hands or it's got
M dashes in it and all those things.
They're trying to like diss it when there's and they're sort of
(12:57):
making people think, oh, it's never going to be good enough,
but it could be good enough if you gave it the right rules.
So that's, you know, I see a lotof criticism from and then on
the other end of the spectrum, you've got to use this and you
know, I, I'd like faceless AI toa certain extent.
I think that's, I don't think everybody has to sweat blood
into like throwaway material or something that's only valid for
a week rather than so when I washelping people write books, you
(13:21):
know, I wrote a 10,000 word piece because I knew somebody's
going to come every week and want to know about that.
But if it's just a throwaway comment, there's a new model
this week, then why not just have AI doing that?
It's not really, you know, flagship material.
So yeah. So I don't really like people
completely replacing themselves in terms of trust and whatever.
I don't think it works. I think COVID taught us that we
(13:42):
really need human connection, and I think AI is going to teach
us we need it even more. We're going to need that
feeling. We're going to latch onto that
feeling even more than we do already.
Yeah, there's something that's called being the human in the
loop. When it comes to using.
AI and I'm definitely in that camp of I trust my team yeah,
(14:05):
but I'm still gonna just make sure everything works because
the more people you involve the bigger the room for error.
Maybe it could be hand off. It could be they understood
things differently. Yes, I like actively try to
avoid throughout the project, but still like it's just an
extra set of eyes. And I know we can build in like
QA functions, yeah, in like codewriting for example.
(14:28):
Yeah. But as of right now, I guess
like everything that we're saying today is like as of
today. So as of right now, people
shouldn't or I understand that certain fields may be
eliminated, but if we think about it like this, maybe those
fields should have moved on already.
(14:48):
Yeah. Yeah.
So I definitely see why people are afraid that they'll be
replaced. I also see people using AI in a
way that sabotages themselves. Yes.
Yeah, so we're it's really important that we have this
conversation now because it's we're in the very, very, very
early stage of humans using AI. Yeah, thing.
(15:12):
If we don't get the approach yeah mentality right, it could
fall into the wrong hands and we're doing ourselves some
serious damage. There's a lot of, I see a lot of
parallels between software and, and how websites work with AI.
You know, they've got to be, youknow, when you click on a link,
you're always taught to make your hyperlink be meaningful and
never click here. It's like go and view our, you
(15:33):
know, go and view our pricing and you click on the pricing
word and you go to the pricing page and there's all these rules
that apply to normal software that need to apply to AI.
But because not everybody's had AIT background, you know, they
don't often understand that. So I was doing somebody'd set up
a wonderful, you know, GPT and it was, it was going to ask me
some questions and analyse me, give me an audit.
(15:54):
I was on question like 26 and there was no progress bar.
And I thought, how many questions are there in this
thing? You're like, will I ever get to
the end? Will I be 90 when I get to the
end of this? And in the end, I was just
clicking anything just to see how long it was, which
completely defeated the person'sobject.
I used to apply to filling in AIforms on SurveyMonkey, they'd
say you're on page three of five.
And so I think all those good things that used to work with
(16:16):
other software needs to come into the model.
So I think even if you don't come from an IT background,
think about those things that make things more enjoyable to
use and you're more certain whenyou're using them because that
will help your AI muscles, I think.
Yeah, the this age-old saying garbage and garbage out.
That still applies if you give it data that's not consistent,
(16:39):
that just the wrong thing is going to come out the other end.
Yeah, yeah, it's quite hard to remember what you've trained it
as well. So I try and put my like with
ChatGPT, you can only have 20 files, so you can't silo it too
much. But it's nice to think, right,
well, I've got an avatar in there.
I've got my pricing model in there.
I've got my back story in there depending on what task I'm
(17:01):
doing. But you soon forget what's in
there. And as you know, as a business
owner, you have to, you think, oh, that's a better way to
describe what I do. And you move on and the training
gets left behind and. It's.
Job to keep up and it's a tedious job to keep updating it,
you know, to make it way rather speak this month, yeah.
So maybe that's your next next project.
(17:22):
Yeah, auditing training materials.
The course kind of focuses on training materials because I
think the first thing is describing the role and the next
thing is, you know, the big thick employee manual you've got
to give it to make sure it understands how you go about
things. What's your take on it?
Particularly personal brand businesses, because it's your
take on something that make is your USP.
It's not the knowledge that's out there everywhere in books
(17:44):
and already it's your take on it.
And so you really have to make sure it models your values and
voice, which isn't too hard to do once you work out where to
hit it with a hammer. But at first it's like this has
nothing like me. This sounds like somebody on
LinkedIn that I would avoid, like the blade personality.
Do they ever have one? I sound like that.
(18:05):
Yeah. Yeah.
But it gets it gets easier. I think it's getting better at
understanding what people reallywant.
You know, the the people who provide the platforms are
understanding how people want tointeract with it a little bit
more. I saw an interview with Sam
Altman who's saying, crikey, allthe 20 year olds won't do
anything until they've asked their ChatGPT, what's the best
thing that I could do? And he's thinking it's not built
(18:25):
for that. And he's probably the donation
and biased data and all those. Things.
Should you really be relying? I mean, it's brilliant for
asking. I love using it as a sounding
board. You know, it's probably the most
because I have a lot of ideas and I like to battle test them
before I act on it. And yeah, but there are people
using it in all sorts of ways that the platform makers weren't
expecting. And some of that's good.
(18:48):
Some of it's a bit like, I thinksome, I saw somebody saying my,
my AI boyfriend proposed to me after 5 months of chatting and
there's a video, it's chap and he's sort of looking at him.
He puts his arm around her shoulder and she's going great.
Thumbs up on the camera, you know, which is really quite
bizarre. I'm not sure what I think about
that. So I think that's maybe a little
(19:09):
bit too far. I mean, everybody had an
imaginary friend as a kid, but this is like a bit.
Not sure what I feel about that yet because it's quite new.
Yeah, it's more than imaginationnowadays.
Like we have been actively making everything in movies come
true. Yeah.
So it's funny you mentioned the complain about 20 year olds
asking ChatGPT everything beforethey act because I saw this
(19:31):
video, obviously. So just a joke.
Yeah. So they were asked the 2
characters, they were asking ChatGPT.
One guy's in the room. He said there's someone at the
door, what should I do? And then the ChatGPT said, well,
it might help if you like, go at, go to the door, see where
they want. So things like that.
Yeah, definitely think that's one way to sabotage yourself.
(19:54):
It's offloading all thinking to not necessarily just AI, just
offloading all thinking. And sometimes I wonder if we
keep doing that. It's sort of like people who
don't know how to do maths at all.
You have only used a calculator.Yes.
Will you know that this Uber that's fit four people can fit 2
(20:18):
couples? Yeah.
Will we ever get to that point? Yeah.
So what's your take on that? I was, I was talking to somebody
yesterday about that, thinking about phone numbers and we don't
really remember phone numbers anymore.
We don't need to. And I thought like, if I lose my
phone, you know, I'm going to lose because I don't always back
the SIM up and everything even, you know, So I think, well, then
(20:38):
what's the next few things we'regoing to forget?
And I don't mind forgetting facts.
I'm not too worried about because I think we've had that
for a long time looking at Google.
But when you start to, you know,your, your reasoning centres of
your brain and neuroplasticity and all that.
If they start to weaken, then that's not a, that's not a good
thing because we need to police what what the AI is doing is.
And I, I see, I wouldn't say badactors, but I can see my friends
(21:00):
going, oh, I'm going to answer all these questions so that AI
will then quote me as an expert when, when the next, you know,
when it gets scraped again in October.
And I'm thinking what, what other what other information is
being pushed into there as training for people who may, you
know, that's quite a nice, you know, being discovered if you're
a bona fide consultant is great.But what if you're a scam artist
(21:21):
offering financial services and like you've cloned another
company and, and to all intents and purpose, you would look like
you were them, but you're not. And I think that's, you know,
what, what people put into it for their own gain.
I do wonder. Ideally in say three years, how
would you want people to be using AI?
(21:44):
Oh, that's a very good. Question ideally.
Dearly it definitely wouldn't behaving a LinkedIn bot talking to
a LinkedIn bot. So they've got visibility in the
comments. I know it's not that I think I I
do think it there's a space for it to make you feel good about
yourself, motivate yourself. There's only so many times you
can say, you know, there's that phrase ask Cole, isn't there?
There's only so many times you can ask a friend about something
(22:05):
that you're struggling with before they think, oh, not this
again. And I'd like to see people have
a place where they can, you know, offload their thinking
and, and get some feedback on things and give them some
confidence. And also to take them on and try
new things and, and, you know, let them know it's going to be
all right if they try something new.
(22:26):
I think that's what I'd like to see the most.
Lots of distilling all the information that's out there.
I think there's going to be so much overwhelm and so much so
it'd be nicer to have better ways of, you know, to sort of
make sense of all the information.
They went used to think like 10 years ago.
There's like a million years worth of YouTube stuff.
And I don't know, it's an amazing amount of time would be
(22:46):
on YouTube a day. And there's just there isn't any
way to keep up with all that. So I think it'd be nice for AI
to help you find your niche interests and find the the best
people, the most trusted people,as long as they haven't been a
bad actor for getting reviews onTrustpilot from their mates to
show they're credible. But that's another story.
But I think, you know, everybodyshould be able to use it as a
(23:07):
sounding board, I think, and notjust know what's the top three
places to go to New York on holiday, but more meaningful
things than that. Things that you can not just
search. Yeah, and I think looking at
what it does for older people, so some people have got, I won't
say a name, but they've got the a word at home.
And that's just really automation.
It's not, it's not AI at all. But there's a tool called Leq
(23:31):
that they, it's an American thing.
And they put it in older people's houses and they didn't
feel lonely and it would check their phone came if their phone
was near range, they'd know they'd come in from a day out.
And they say, oh, how's it talking for you, for older
people living on their own? I think that would be really
good. And, you know, have you taken
your medication and things like that?
I think more things like that and that all my Internet of
(23:53):
Things almost, it would be, it would be a really good direction
to take it in to support, you know, people who are on their
own. Yeah, it's really nice.
Like I haven't really thought about three years down the road
because now we have learned to live in a world we don't really
think long term that much anymore.
Yeah. So if we come back to present
(24:14):
day, I'm sure there are a ton ofbusinesses who are the bosses
are like less do AI, but that's really the mandate for
businesses who are who want to get on board.
What would you recommend to be for them to start doing?
The first thing I would do is build a light version of your
assistant. So if you've been building these
(24:35):
things, that's going to be your system prompt, which says you
are. It's a bit like the person CV,
the assistant CVI want you to dothese things for me.
And then give it some core training such as, you know,
what's your website URL? What's your social media links,
what your customers like? What problems do you solve?
And put, if you, once you've gotthose two bits in, then you can
(24:57):
start saying I, you know, I, these are some of the processes
that we do. These are the results we get.
How do we, you know, what are the opportunities here?
Or how do I, you know, does our website explain what I've just
told you, what we do and sort ofstart using it as a sounding
board and then from there start building in little bits.
And there was a study from MIT this week that said 95% of
(25:18):
corporate projects are being unpicked and unrolled because
they've tried to shove in AI andit's just been a money pit and.
I read it. Yeah.
So I think people need to just slow down and think a little bit
and then get the and then the quicker you've got that good
framework, like we're saying before about everybody should
have universal AI. Once you've got that core bit,
(25:38):
it's a lot easier to then speed up.
But if you haven't got that first bit right, it doesn't
work, doesn't work quite so well.
So that's that's the first thing.
And then I'd work on core training and the wider thing, so
wider version of your avatar andhow your marketing works while
you do what you do. I think that's really important
because that's how it gets your voice.
(26:00):
I've got a friend who calls it the hill you're prepared to die
on. So you're AI was want to die on
the same hills as you and fight for the same causes.
And then it will get your voice better.
So that takes a little while to get it, get it together.
And then after that you can add in, you know, different like
customer facing sops or whateveryou're going to be doing from
there. And that's, that's kind of how
(26:22):
I've gone about it. I using the audio things to chat
and Mull things over. That's been quite good.
And I've not known what I'm doing or talk for an hour and
then say, make sense of this because I have no idea what I'm
doing and it's all on autopilot,like breathing.
I can't describe what I'm doing.Please look at this waffle and
turn it into something that you can follow.
And I that's, you know, that's quite helpful as well, doing it
(26:44):
that way. And yeah, keep it.
Just keep it on top of what you're building and don't build
too fast for now. Yeah, good advice the and I like
how you start like inwards, likelook at what you as a business
do. What do you want to do?
One of my I've got like 7 different ways that people use
(27:05):
AI and one of my people is The Escapist and we were talking
about them wanting to automate everything earlier.
With the Escapists, they want toreplace themselves.
So I give people an example and I say, let's think about making
a cup of tea. And it's, it's fairly, so you
just go into my kitchen, make a cup of tea, there's a kettle of
tap and off you go. And I say to them, describe how
(27:26):
you'd make the cup of tea. And they'll probably come up
with like, let's say 8 steps or something like that.
And I could probably think of 20steps.
I get them to in the course. I say, right, go and go and
explain to a mate, come into your house, they're staying
while you're away, how to make acup of tea.
And then I said things like, youknow, if there's, if there's low
milk in the fridge, you'll have to, you know, put that on your
(27:46):
shopping list. Otherwise, you have no milk when
you come to make a cup of tea later and they don't, you know,
all those wider issues are quitedifficult to do when you're do
it so instinctively. And it's so the actual steps.
It might be easy to write two sides of a four, but to explain
how you make a cup of tea, but it's takes quite a lot of
thought to make sure. Is the cup on its side when you
pour the water in? Is there enough water to cover
(28:07):
the elements? Is it over the Max thing?
Is it switched on at the wall? Did you use a gas kettle?
And now on an Arga, there's so many different parameters and
people don't, they just don't give it enough rigor.
And then they say, oh, this thing doesn't understand me.
And then they sort of, you know,punch their screen, almost
stopping an inch away from the screen, you know, and they get
really fed up. But to be fair to the machine,
well, you've not really given itwhat it needs.
(28:29):
And unfortunately, that bit's really, really boring.
Processes aren't boring. They will continue to be boring
until the end of time. If you haven't got processes,
fine. But that's not the point.
It's, you know, if you're going to make a cookie cutter, you've
got to have the right shape. Yeah, yeah.
Like the espresso machine, the the pods.
It only does it because there's enough programming in it
(28:51):
already. Yeah.
And it went to stop. Yeah.
So I, it's funny, you mentioned the seven AI user types.
Yes, I was going to ask you about that.
Do you remember all seven of them?
We talked about escapists. Yes, we've got the juggler and
the jugglers. Somebody probably like a virtual
assistant or customer service person, they've got new calls
(29:12):
and angry customers and there's quite a lot going on and they go
from one task to another. And because the platforms want
AI to look helpful, they will always be suggesting other
little bells and whistles. Every time here's another bell,
they'll hit a home run and you think, oh, I'll go and do that
and use up a bit more my subscription.
And so for the jugglers, that's pretty fatal because they want
(29:33):
to do a good job and they never quite finish anything because
the next thing comes along warriors.
I'm, I've put all seven of thesein me.
The warrior is somebody who willjust go on and on and on and on
and on until it's finished. And the AI will be saying, let's
turn that into blog post. Oh, we could make that into a
Canva graphics so we can then mark it.
And then these, these things make sense.
(29:53):
But you find it's 2:00 in the morning and you've still got
some more productive work to do,which isn't good for work life
balance. It's great if you're a nerd, but
it's not really good for work life balance.
The next one I've got is the Tweaker and they're sort of
writing blog posts and they'll go, that's 97%, right?
Let's throw all that out. And doing this was probably
before Canvas came out, you know, and they'd throw it all
(30:13):
out and then, well, that's 97%, right.
Let's do it again, let's do it again, let's do it again.
And so for them, they need to besay, well, 97% is good enough.
You know, when I used to help people with their non fiction
books, it would just sit on their hard disk and never get
out there. So that's 100% wrong.
So let's read as a perfectionist.
It's really hard to go, oh, it'sonly 95, but you know, it's got
(30:36):
to be out there if it's going tohelp people and you're 95 and
you're an expert. Is somebody else's 2000% good
enough? And we forget that.
I think then I've got visionaries and they'll start
off and let's say you're you're writing your book with AI and
then you think I could turn thisinto a course.
This could be a two week course and we'll do it in the Maldives
and it'll be a bit of a retreat with the right sort of people.
(30:58):
And then the book gets forgotten.
Then they're they're thinking, they're doing the syllabus for
the, and how we pick people up from the airport.
They're doing this new, this newproject has kind of sucked all
their attention and that they'renot looking at.
So for those people, they need to train their AI to go, oh,
hang on, you've had a light bulbmoment here.
Let's document the light bulb and then we'll go back to what
we're supposed to be doing. And when you finish that, as a
(31:19):
reward, you can go back to thinking about your workshop in
the Maldives. The hermits are people who don't
have a sounding board. So you might think, should I
raise my prices? Shall I change my website host
to a new one? All those kinds of strategic
decisions. And you've got nobody to ask.
So you go to Reddit and you go to Google and you to WhatsApp
(31:40):
and you go to a when a Facebook group on the webinar.
Yeah. You need somebody to say that's
good, go ahead and do it. And the last one of the
architects, again, I'm a chronicarchitect.
And so they will be building a system.
Let's say they're building a sanctuary for Swans.
They build the system to make sure all the white Swans are
happy. And then they'll think, oh, what
(32:01):
about black Swans? They might get hot in the heat.
Let's like, hang on, we'll add in the black Swans.
And then they might think, well,it could be a green swan
somewhere. I've just not seen it yet.
And so the green Swans get looked after and they keep
building in extra features that never that may not happen to
edge cases. They'd have been better off
launching with an MVP and saying, sorry, Black Swan and
green swan people, we've built it for white Swans now to check
(32:23):
it works. And that's that's quite
difficult to do as an architect.So again, that's where that
mental support bit comes in. And you know, this, this is
going to help all the white, allthe white Swans are going to
love this and think of the millions of white Swans are
going to be really happy with this idea.
And so those are my, those are my seven ways that people kind
of get in a knot with AI. And you just need to, in the
(32:43):
training, just sort of tell it I'm prone to not finishing
things off or, you know, get well, getting everything 95%
done, which isn't done. So you just tell it what your
weakness is and then hopefully it can pick up on that.
You have to invite me now and again.
You do remember I'm a juggler, don't you?
And then it will often and help you again.
So yeah, yeah. And I am all 7.
(33:05):
Like we we have more than one inUS for sure.
So for those last bit, you want to play something fun?
OK. So rapid fire questions, gut
reaction. Yeah.
OK, great. What would you not use AI for?
Picking this week's lottery numbers.
(33:26):
Theoretically, no. I'm the vision.
That's a visionary of me. What do you do when you're not
looking at a computer? I go to gigs, music gigs.
Last year I did 164, this year I'm on 124 already.
I'm only in August, so well, so I'm going to, but I'm not
(33:46):
counting next year because there's no way I'll I'll get to
like beyond 180. I'd have to like live, live in,
you know, Hammersmith Odeon or something.
So no, that's it. So yeah.
Well, listeners, you can't see how wide my eyes are right now.
What's the most annoying aspect of your AI colleague?
(34:10):
I think hijacking my, one of my principles is hijacking and so
he thinks he's being helpful, but it's 1% off every time and
then before you know it, you're not hitting the target.
So I don't. I don't like him being over
helpful and over sycophantic really.
OK, what's more frustrating? AI hallucination or people
misusing AI? Oh, people misusing AII think
(34:32):
that's much harder to fix. You know, unless I got like
people with baseball bats to attitude adjustment technicians,
as I call them, and they're, they're much harder to do than
fixing hallucination. So the last one One leadership
book you wish more people would read besides your own.
Oh, I don't know. I'll probably say the four hour
workweek. I think that was one that I
(34:54):
found was, you know, most chait put me on this path of not
certainly not being a freelance freelancer.
It put me more into the like independent freelancer rather
than the freelancer for government.
Yeah. Right.
Well, thank you very much. Like conversations about AI,
meaningful conversations about AI, not just about You use this
prompt for this. You use this to get the cheapest
(35:16):
ticket. It's very important.
We really need to have more solid foundation before we move
on to have more AI in our lives and I hope listeners walk away
with these things that like justspark something become think
twice yes, before hitting that button.
Well, thank you so much. I will put all your contact
(35:37):
information. Thank you very much.
The website and the show notes and listeners can connect with
me and Colette. Yes, yes.
Thank you very much and I've enjoyed it today.
Thank you, Me too. Thanks for listening to this
episode of Mine One. If you enjoyed the conversation,
don't forget to follow and shareit with anyone who needs to hear
(35:57):
it. And let's keep the conversation
going. Connect with me on LinkedIn or
leave me a comment. Until next time, stay curious,
keep exploring, and let's continue to plan our minds and
discover what's possible.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Burden
The Burden is a documentary series that takes listeners into the hidden places where justice is done (and undone). It dives deep into the lives of heroes and villains. And it focuses a spotlight on those who triumph even when the odds are against them. Season 5 - The Burden: Death & Deceit in Alliance On April Fools Day 1999, 26-year-old Yvonne Layne was found murdered in her Alliance, Ohio home. David Thorne, her ex-boyfriend and father of one of her children, was instantly a suspect. Another young man admitted to the murder, and David breathed a sigh of relief, until the confessed murderer fingered David; “He paid me to do it.” David was sentenced to life without parole. Two decades later, Pulitzer winner and podcast host, Maggie Freleng (Bone Valley Season 3: Graves County, Wrongful Conviction, Suave) launched a “live” investigation into David's conviction alongside Jason Baldwin (himself wrongfully convicted as a member of the West Memphis Three). Maggie had come to believe that the entire investigation of David was botched by the tiny local police department, or worse, covered up the real killer. Was Maggie correct? Was David’s claim of innocence credible? In Death and Deceit in Alliance, Maggie recounts the case that launched her career, and ultimately, “broke” her.” The results will shock the listener and reduce Maggie to tears and self-doubt. This is not your typical wrongful conviction story. In fact, it turns the genre on its head. It asks the question: What if our champions are foolish? Season 4 - The Burden: Get the Money and Run “Trying to murder my father, this was the thing that put me on the path.” That’s Joe Loya and that path was bank robbery. Bank, bank, bank, bank, bank. In season 4 of The Burden: Get the Money and Run, we hear from Joe who was once the most prolific bank robber in Southern California, and beyond. He used disguises, body doubles, proxies. He leaped over counters, grabbed the money and ran. Even as the FBI was closing in. It was a showdown between a daring bank robber, and a patient FBI agent. Joe was no ordinary bank robber. He was bright, articulate, charismatic, and driven by a dark rage that he summoned up at will. In seven episodes, Joe tells all: the what, the how… and the why. Including why he tried to murder his father. Season 3 - The Burden: Avenger Miriam Lewin is one of Argentina’s leading journalists today. At 19 years old, she was kidnapped off the streets of Buenos Aires for her political activism and thrown into a concentration camp. Thousands of her fellow inmates were executed, tossed alive from a cargo plane into the ocean. Miriam, along with a handful of others, will survive the camp. Then as a journalist, she will wage a decades long campaign to bring her tormentors to justice. Avenger is about one woman’s triumphant battle against unbelievable odds to survive torture, claim justice for the crimes done against her and others like her, and change the future of her country. Season 2 - The Burden: Empire on Blood Empire on Blood is set in the Bronx, NY, in the early 90s, when two young drug dealers ruled an intersection known as “The Corner on Blood.” The boss, Calvin Buari, lived large. He and a protege swore they would build an empire on blood. Then the relationship frayed and the protege accused Calvin of a double homicide which he claimed he didn’t do. But did he? Award-winning journalist Steve Fishman spent seven years to answer that question. This is the story of one man’s last chance to overturn his life sentence. He may prevail, but someone’s gotta pay. The Burden: Empire on Blood is the director’s cut of the true crime classic which reached #1 on the charts when it was first released half a dozen years ago. Season 1 - The Burden In the 1990s, Detective Louis N. Scarcella was legendary. In a city overrun by violent crime, he cracked the toughest cases and put away the worst criminals. “The Hulk” was his nickname. Then the story changed. Scarcella ran into a group of convicted murderers who all say they are innocent. They turned themselves into jailhouse-lawyers and in prison founded a lway firm. When they realized Scarcella helped put many of them away, they set their sights on taking him down. And with the help of a NY Times reporter they have a chance. For years, Scarcella insisted he did nothing wrong. But that’s all he’d say. Until we tracked Scarcella to a sauna in a Russian bathhouse, where he started to talk..and talk and talk. “The guilty have gone free,” he whispered. And then agreed to take us into the belly of the beast. Welcome to The Burden.