October 15, 2025 • 42 mins
Today Claire Pedrick talks with Dr. Lisa Turner about the integration of AI and coaching. Listen to the opening as Lisa tells her extraordinary journey into what makes her codify life stuff. Explore how AI can enhance coaching practices, making them more efficient and accessible, while maintaining a human touch. And we dive into ethics and working with neurodivergent people..
Check out
- A free copy of Lisa’s book: Our Conscious Tipping Point ourconscioustippingpoint.com
- The episode with Jazz Rasool: https://thecoachinginn.podbean.com/e/s4-episode-46-ai-and-coaching-jazz-rasool/
- And listen to how far we have come since Sam Isaacson came to talk about AI in 2021 https://thecoachinginn.podbean.com/e/in-conversation-with-sam-isaacson-coaching-and-technology/
Contact:
- Contact Lisa through Linked In
- Contact Claire by emailing info@3dcoaching.com or check out our Substack where you can talk with other listeners.
Further Information:
- Subscribe or follow The Coaching Inn on your podcast platform or our YouTube Channel to hear or see new episodes as they drop.
- Find out more about 3D Coaching and get new ideas and offers in our weekly email.
Coming Up:
- How a Scottish University is using Action Learning as part of a support structure for workers in Residential Childcare
Keywords:
AI, coaching, technology, innovation, Lisa Turner, Claire Pedrick, productivity, ethics, neurodivergent, human touch, engineering, transformation, future, patterns, tools, enhancement, accessibility, efficiency, concerns, planetary well-being
We love having a variety of guests join us! Please remember that inviting someone to participate does not mean we necessarily endorse their views or opinions. We believe in open conversation and sharing different perspectives.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:13):
Hello and welcome to this week's edition of The Coaching Inn.
I'm your host, Claire Pedrick, and today I'm in conversation with Dr Lisa Turner.
And it's such a timely conversation because loads of you have been emailinginfo@3dcoaching.com and said, AI and coaching, what do you think?
(00:34):
So Lisa, we'll come onto that AI and coaching, but first of all, just tell us a little bitabout you.
uh and what got you into this work of coaching.
yeah.
So it's great to be here and lovely to meet you.
um So the probably the easiest way to explain how I got to where I am now is to kind of goand just tell you a little bit about my background.
(00:58):
And I promise I'll keep it, keep this.
It's not going to be a full on life story.
It'll be the although we do start along with.
So, so aged 12, I was groomed and trafficked.
by a paedophile ring and moved to London, which is why I moved from Australia, which islike, probably hear a little bit of a twang in my voice sometimes when I talk about cars.
(01:21):
I go a little bit Australian then.
So I was moved to London, kept as a virtual house prisoner for five years, managed toescape and probably had what will be labeled now as CPTSD and...
I had all sorts of social anxiety disorders and I was just not in happy place.
(01:41):
But I didn't know it.
I just thought this was normal.
And at the time, the only thing I got any real joy from was cycling.
I out on a late night ride, uh training actually for a time trial, and got hit by a car.
So I often say I had a car accident, but the car had the accident, I just got in the way,is how I prefer to phrase it.
(02:04):
Then I had this incredible out-of-body experience and there's this spirity entity saying,what you doing?
You coming with us, you going back?
And I obviously came back, but I had two questions.
What the hell was that?
And how do I get back there?
Because this experience of complete peace and joy and like just bliss.
(02:29):
Now at the time I was doing a PhD in Engineering and
So what I did was I put on my engineering research hat and said, okay, I'm going to figurethis out.
So I studied everything from different sorts of therapy and coaching and NLP and timelinetherapy and hypnosis to shamanic work, tantra, all the things.
(02:50):
And being an engineer, what I did was I spotted the patterns, I looked for the patternsand I found them.
And then I reverse engineered the spiritual experience.
So I demystified it.
Now I get a lot of flack from spiritual people because I'm taking all the mystery out ofit.
it's like, or I'm making it accessible to lots of people.
(03:11):
So then what I have was a whole series of processes that enabled people to clear trauma.
That's this book here, CET Yourself Free, Set Yourself Free, and a lot of coaching uhprocesses and models that worked.
Then what I did was
And so then I taught that for 20 years, taught that, did that, worked with clients for 20years.
(03:35):
And then AI came along and it turns out, and this is where I think that, so we're coachingand AI really, you know, can really work together is this.
So then what I did was I took all of these models, these processes, these very systemicstep-by-step, you do this, you do this, you do this, and then you get this result.
(03:58):
whether this result is clearing trauma, installing new strategies, releasing a limitingbelief, whatever that is, the process does it.
And you can do that with AI.
So, and we know when people say like, do you, I don't know a shorter way of explainingthat, know, I blend technology with coaching, with esoterics in a methodical, practical,
(04:24):
grounded and pragmatic way.
And as an engineer, we come from it with this,
this principle which is we have this one question which is does it work?
What's it do?
Does it work?
And if it doesn't work we're not really interested and we're not actually interested inlike the innards of it you know like like magnets I don't trust magnets I can't see them I
(04:46):
don't know how it's working like you know this whole weird magnetic field thing what'sgoing on there so you know I like to be able to kick things and hit them with a hammer
because then I know they're real I like the
tangible.
And so that's what I did was I made a lot of things that were intangible, very tangible.
And now I do do that essentially using AI.
(05:06):
So I was in tech, left tech, and then tech came and chased me down.
I was saying.
How interesting.
what it sounds like you're saying is that when there's a good and useful process forsomeone to go through.
AI can do that beautifully.
Fine.
(05:26):
Yes, and I will just put this little caveat.
I know that there's a lot of coaches and a lot of people who are actually trying to takethe coach out of it, out of the process.
So coaches are kind of doing this coach clone thing, which I'll talk more about if youwant.
I'm not a huge fan of it and I'll explain why.
But they're trying to put AI between themselves and the client.
(05:52):
And I think it's better to think about it that we put AI
Beside the coach so the coach and the AI so rather than replacing the coach we augment thecoach and I think that's a very different way of approaching this I Have this we've gone
deep into thought there Claire did
I have.
(06:16):
I could say what I'm thinking or you could carry on what's best.
no, no, please, say what you're thinking.
As you're talking, the thing that's making me think about is that...
This is not the right word, but it almost doesn't contaminate the coaching relationshipbecause the AI does the other bit that you need in order to be able to do this.
(06:40):
So for example, when I was learning paddle boarding, she said, if you want to stand up onthe paddle board, you need to be more flexible.
And if you want to be more flexible, you need yoga.
Because the yoga bit,
We'll do that bit and then you come back to me and then we'll do the paddle boarding bit.
(07:04):
And that's sort of what you're describing, is it?
Yes, so I think, so I have this four applications or evolutions of using AI in coaching.
And I think they're all useful, but I think we need to be aware of what we're doing.
And I think this gives us a really clear structure.
(07:25):
So the first level, I'm a huge fan of, and that is where you just use AI to support.
your practice so this is your productivity stuff so this is your marketing getting it tohelp with a copywriting which it does pretty well provided you keep the human in the loop
like if you just get it to write your sales page it will do 80 % of the work which isgreat but if you don't do that extra 20 % you'll end up with something that's fairly
(07:51):
rubbish and won't work so so when it comes to
do make the spelling correct, because otherwise everybody knows that you did it with AI.
Exactly.
Yeah.
There are actually ways to train it to use the, like to use English, English spelling orBritish spelling or whatever you want.
Anyway.
So, so all of the productivity things now, you know, there's a lot of studies that saythat coaches actually only spend some, some coaches as little as 35 % of their time
(08:18):
actually coaching.
And the rest of the time is on admin and marketing and, and, know, marketing and sales.
Now AI can
absolutely reduce that and save you hours.
And there's a whole bunch of tools and even things like, so I have created a whole suiteof coach support tools that are not only helping you with your productivity, with your
(08:42):
content creation and your marketing and things like that, but also um things like, say youwant to produce a worksheet for your clients in between sessions, it'll do that.
Say you want an accountability app.
really easy to do a simple accountability app.
So it just emails your client at a time or pings them a text at a time and says, go do thething you said you were going to do, which is way better than you having to remember to
(09:08):
do.
So when it comes to that, AI is brilliant.
So that's level one, productivity stuff.
The next level, level two, this is where, and this is where I'm not such a fan of, this iswhere individuals and coaches sometimes
communicate, work directly with any of the LLMs.
(09:29):
So an LLM is a large language model.
And by that, mean Chattie BT, Claude, Gemini, know, all of the big ones, the big boys,they are men in gilets.
With pink jeans.
oh
No judgment, whatever.
Anyway, so it's when people communicate directly with the LLM.
(09:51):
Now, here's the thing about the LLM.
The LLM has, because I come from tech, I have a reasonably good grasp of machine learningand context windows and token usage and the limitations of it.
So one of the things about any of the LLMs is they are essentially at their crudest leveland
(10:15):
They'll argue that this isn't the truth, it's not strictly the case.
But you could say they are a really fancy predictive text.
Now, any of you have known, you know, when you've tried to use auto cucumber and autogrit.
I have said, I'm so sorry, it's my auto cucumber.
(10:37):
dear.
So, you know, the mad text you sent.
So essentially, and that's where we get the crazy stuff that AI outputs.
Now the reason it does really crappy output at that level is because the person using itsays two reasons.
So first of all, it will...
so being a predictive text, it uses a statistical model.
(11:00):
So it sees a chunk of text and then it says, the most likely next thing is this.
And so its output will give you the most likely.
Now,
By definition, means the most superficial, the most popular.
Now, those of you who are decent, good, deep coaches who've done the work, done thetraining know that the superficial pop psychology stuff is pretty limited and often very
(11:30):
wrong.
And so that's what you get.
You get superficial, limited, crappy output.
that's where, unless you know deep, good coaching models,
as the user, you're gonna get, you it's the classic thing that we say in any computing,garbage in, garbage out, you know?
So you put in a garbage prompt, you're gonna get garbage out.
(11:54):
that's why the advice that, that's why people often say just like, unless you're a trainedcoach, do not use or trained in the modality that you're working with ChatGPT with, do not
use ChatGPT for your coaching and therapy.
You, and so that's level two, superficial.
Now, some of the bots that are out there,
are not much better.
(12:15):
So I've been out there testing them all and including that they run on because they'reoften developed by men in gilets who are not coaches.
They're developed by and so they developed using like if you're lucky a bit of CBT, ifyou're lucky the most superficial like reflective.
(12:37):
Like, oh, so, know, so you'll put in, I'm feeling stressed today.
Oh, so it sounds like you're very stressed.
And it's like, yes, I just said that.
I was like, that tells me nothing.
not great for AI, it's also not great for human coaches, I observe.
Correct.
Yes.
Correct.
And it'll say, shall we do some breathing exercise?
And it's like, yeah, okay.
I probably could have figured that out for myself.
(12:58):
So these, so using LLMs directly or many of the bots now, and I, you know, like there maybe one out there that's better than I'm describing.
I haven't found it yet.
And I do everyone that comes out, I do a little trial period and I take, and I, know howto break the guard rails of these things.
So I push them to the limits.
And the thing is I do that because honestly people should pay me to test their AI coachingbots because I do it so well.
(13:23):
So because what I want to do is I want to see, can I make this thing give me like if I'mand I do it with like genuine, like what are people likely to come to this with?
And it comes back with pretty crappy output.
So that's level two.
Then we have level three and that's where I come in.
(13:47):
As I said, I've developed all of these models, are step-by-step processes that take youdeep.
And then what I've done is I've created over 70.
And every time I'm interviewed, it goes up because I do about one to two a week.
I create a new specific tool and it does one thing really deeply, one thing really deep.
(14:13):
So it is narrow, but deep.
This forces, so with these tight guardrails, force the tool to go deep, to not just givethe superficial stuff.
Now, is it always perfect?
No.
Is it a million, it is a country mile better than the just using the LLMs and directly allthese chatbots, know, these coaching chatbots directly.
(14:41):
In those coaching chatbots, I also include most of those
coach clones.
So one of the things I do is, and the reason is because many of the coach clones and theyoften they're often advertised as saying, Oh, we'll you know, take your methodology and
you're and it's like, yeah, provided your methodology is in their suite.
(15:05):
So if you've got your own IP,
And the thing is a lot of coaches have really, really profound models, really, reallyuseful.
They're highly skilled, highly specialized, and they've developed their own tools.
But often, they're not always fully conscious of what they do.
(15:27):
They haven't methodicalized yet.
Is that the right word?
uh Turn it into a methodology.
So that's the other thing I do with a lot of coaches, and I call it magic to method.
reveal.
So and people often say I'm intuitive and it's like no intuitive means so like intuitionyou can only be intuitive about something you actually know anything about but like that
(15:50):
is kind of the definition of intuition it just means I know so much about this I'm runningthese patterns I'm running these processes unconsciously so then what I do is I work with
people because I'm great at pattern recognition rubbish at spelling rubbish at parking mycar greater pack and pattern recognition
I will assist them to bring that into awareness.
(16:11):
And you won't be surprised to know I have a bunch of AI tools that help us spot thepattern.
So I've got a tool where we load into a whole series of, their coaching sessions and usinga very specific prompting that I've developed, which I call...
(16:32):
the and I haven't really got a great name.
I'm calling it the Turner cohesion method, but essentially it's recursive interpolation.
I know, I know it's like, what am I going to call this?
What am I going to call this?
And I keep forgetting that I'm psyched because I haven't really like, it's really hard todescribe what it is.
Essentially, what it does is it, it triangulates.
(16:53):
So it takes, it looks at what is the client or the user, which may be the coach, what arethey input?
and then it checks for both depth and then it asks another similar question but differentto try and get the same answer from a different aspect.
And then it goes, okay, so when I asked a similar question before, it compares those twoanswers and says, are they both deep enough and how similar or different are they?
(17:21):
And if they're neither deep enough or they're too different, it will keep asking questionsuntil we get to, here it is, here it is.
And that's where you get the depth of the, so whether we're using that with a client, weget the depth of the client's say core limiting belief or their core pattern that they're
(17:41):
running, which they'll know the superficial pattern, which isn't necessarily what'sactually going on in the unconscious mind as we know.
So it's really powerful at revealing those patterns.
So we're still on level three here.
So level three, so what I do is I assist the coach to bring that into awareness, turn itinto,
step by step, also find one of the things I always say is bring the sessions that wentwrong.
(18:06):
And we'll die because we want to look at, well, this is what went right.
This is what went wrong.
What was the difference?
Let's look at what did you miss or what did you do in those sessions where it And I wouldsay, like, this is not about blame or judgment.
This is about finding what your method is.
This is going to make you a way better coach, way better at your IP, doing what you dobest.
Because you probably know you come out of a session, you go,
(18:27):
didn't go quite right and I don't know where we fell off the cliff there.
You and we've all, like you're nodding so well, we've all had that, right?
We've all had, I mean, a good coach will be self-aware enough to know that.
And there'll be good to be able to, you know, a good coach will be able to course correctin the session or in the next session.
But what you're describing is a different kind of reflection using real observable data ina different way.
(18:52):
So I talk a lot to coaches about the value of watching recordings over writing in anotebook, because writing in a notebook is a really useful thing, and it's self-reported
data.
Watching recordings is a really useful thing because you can actually see what happened.
But what you're describing is something beyond that, I think.
(19:15):
which is that you can understand and spot the patterns in what happened in a way that'snot your own personal bias.
Absolutely.
And we also have to remember that AI is bias as well.
um And in my uh programming and coding, I do my absolute best to eliminate bias.
(19:38):
And I always warn people, it's still going to be biased.
It's still going to be biased.
So that's level two.
Sorry, level three, where we go really, really, really, really deep.
really, really narrow and we find those processes and those specific processes and turnthem into tools.
So coaches who work with me, what they come out with is a course that they can teach andcertify their methodology in, a book, uh knowing what they do well, like knowing what they
(20:09):
do and how to do it really well, even better, but also either one or mostly a series of AItools that they can use.
with their clients, not to take them out of the loop, but get their clients to use withtheir clients.
So they give the tool to the client.
The client then comes back with what they got.
(20:32):
And there's two ways we do this.
One is I make something that's called a custom GPT, which you've probably heard of.
so we make a very specific custom GPT or a series of them.
And or I make a tool that's an app and there's disadvantages and disadvantages.
So when it's an app, we can, we can put in what's called a rag, which is basically somememory, extra memory, uh, chat, GBT very, it's got a, it's very specific memory.
(21:03):
So here's the memory and, if you overload the memory with data, it will take the mostsuperficial and it will miss things out.
And that's when it makes stuff up and hallucinates and things like that.
We don't want that now.
But there are still advantages of ChatGPT, just using a custom GPT, because the client'sdata, what they chat with their, like, so it's your custom GPT, the coach's custom GPT,
(21:29):
the client chats with it in their account, and the coach doesn't know what is said.
Now, that actually has advantages, has disadvantages, but it has advantages.
And the advantage is, and this is what I've been finding is that,
clients will say things to their own chat GPT account in their own, you know, to thatcustomer GPT that they would never because of the shame, because of the discomfort that
(21:58):
they would never speak to a human.
And this has real advantages.
So what I often do is I say, well, let's do both.
So we'll get some with an app now with an app.
we can go in on the back end and see what the client said.
So we get that data and we can go, okay, now that's useful because we can then keeptraining the model and go, okay, so people are misinterpreting this question wrong or
(22:25):
interpreting it, not necessarily wrong, they're interpreting this question, we wanna begetting this bit of information and they're answering it with that, so let's rephrase the
question.
So we can do stuff like that, which you can't do if it's a custom GPT, you've gotta, well,we got that output, like, how did it get that?
uh
guess a bit.
So that's level three.
(22:45):
Level four is where we and this is this is I am still in prototype land.
I am in communication with men in gilets who want to who are potential potentialinvestors.
So what I'm doing is I'm creating a it's essentially a personalized coach would still beused alongside a live coach, but
(23:11):
rather than learning just the models.
it might learn Meyers-Briggs.
So it might learn, I don't know, like, that's not a good example.
So it might learn coaching models like, I don't know, like goal setting or smart goals orall the different models.
And for some reason, my brain isn't giving me any right this minute, but all the differentcoaching models.
(23:33):
So you can load those in.
But as well as that, things like spiral dynamics, for example.
So you can load that in so it knows those models, and that's level three.
But in level four, what we do is, there are two phases to it.
We actually have an AI tool that knows all these models, but it learns the user.
(24:00):
So each user has a personalized model that knows personalized mentor coach that knowsthem.
So it knows
their Myers-Briggs.
It knows their Enneagram.
It knows their spiral dynamics.
It knows.
Now I do a lot of esoteric models, spiritual models.
So it might know their human design.
(24:20):
It might know.
We can program it with almost anything.
And it learns that.
So the first phase is that we train the tool on the user.
So it kind of
molds to the user.
that so it's a so it's only usable by that user.
(24:42):
If anyone else used it, wouldn't really well, they just it wouldn't be it wouldn't givethem sensible answers.
So it learns your Myers-Briggs, your disk profile, your big five, your values.
I've got listeners shouting, Claire say something.
I, because I have a real aversion to those models.
Because I think they box people in.
(25:04):
So I just need to, I, and I really hear.
It's so interesting to hear inside your mind.
Thank you very much for inviting us in, because it's really interesting to hear insideyour mind those different levels.
The other question that's really coming up for me is the question that is being discussedall the time at the moment in the coaching world, which is ethics, which is what happens
(25:32):
to my data when I...
when I share my deepest soul thinking with a butt.
So that very much depends on, so if you're using the LLM directly and whenever I work withclients, I tell them, assume nothing is safe and there are some things you can do that
(25:57):
will protect it and you can assume nothing is safe.
That's where having a specialist app is better because it does go to the LLM.
So you have like, so you have the user, the app and then the LLM on the backend.
It does go to the LLM, but what you can do is you can anonymize it in between.
So the data is gone to the LLM, but it is not connected, like not connected to a specificperson.
(26:23):
You can absolutely do that.
then within, you can actually train it.
So the way I do it is you don't harvest it.
You can actually make it so that it's not, and it's a really simple setting.
And you just go into your settings and it's, you say, train the larger model.
Yes, no.
And you just put it to no.
So your data is not harvested.
(26:44):
Now, do we trust the boys in gilets?
That's up to you, whether or not.
I always say, probably assume not.
um And so you're absolutely right about ethics.
And the ICF just came out, I think, just a couple of weeks ago with a whole load ofprotocols for how to use AI and things that are advisable and essential.
(27:07):
And the thing is, with technology,
All of these are solvable.
They are all solvable.
And going back to your point about the models boxing people in, you're absolutely right.
And that's where the, so first of all, this tool can actually learn you with like beyondthe models.
(27:28):
So it learns the mod.
So it learns you in relation to not just a model.
And I think that's how it boxes people in because it boxes people in because it's like,
you're this Myers-Briggs.
And we know that that drifts over time.
know that that's often context, know, values are context dependent.
(27:50):
And also Myers-Briggs has been demonstrated to not be based on anything.
Well, I mean, it was based on archetypes.
It's based on the tarot, and that's one or the other.
So, let's not pick on my...
Because the thing is, and here's the thing I always say, I say about every model, everysingle model, and as an engineer who works with models, we say this, all models are
(28:16):
broken.
They are all false.
They are all inaccurate.
Right?
But...
some of them are useful.
So for example, my PhD, I will bore the pants off everyone just for a moment.
It's a mathematical model that predicted how much noise a particular component made whenrotating.
(28:43):
was noise reduction in high speed rotating machinery is the title of my PhD.
It was very, very boring.
I know, I thought everyone wants to read that, right?
But, and the thing is, one of the things, and this is where I think coming at it from asan engineer and not a scientist, because as an, I mean, and even science knows this as
(29:04):
well.
Scientists know that models are wrong.
We know that they are wrong, but we know the limits of their accuracy.
So for example, my PhD model worked up to, you hit the speed of sound and then it wasgarbage.
But up to the speed of sound, like it just was lunatic.
It was just, it just made no sense.
(29:25):
And I can even tell you why.
It's because air becomes compressible and at the speed of sound, air becomes compressibleand it no longer acts like a norm.
It becomes non-Newtonian.
We all want to know that, right?
So, so it only works for new...
So I have a question.
It's interesting because I have a degree in applied statistics and I know that you can'trely on statistics.
(29:48):
Statistics are useful and then not always right.
My question to you is, one of the issues about psychometrics is that many of them are notuseful and don't serve people who are neurodivergent.
So I'm just really curious whether
and if and what insights you have into all the things you've been talking about inrelation to people who are neurodivergent.
(30:14):
I mean, clearly your level one is a lifesaver.
uh But in terms of the way the apps relate to people, what have you got to say aboutrelating to people who are neurodivergent?
I will, so what I do is I put in these caveats that can accommodate for neurodivergency.
(30:40):
And I'll say, you know, some people will answer like this and then answer, you know, askthese other questions.
So we can put that into the coding.
Here's what I will say from my experience.
And I have, and I don't advertise or promote and say, come to me if you're neurodivergent,but they find me.
(31:01):
And here's what they say about all of the apps, that they are so useful.
They love how it helps give structure to what they describe as their thinking processes,which have seemed chaotic to them.
And it's like, I try and hone my thoughts and I can't, and the apps help me do that.
(31:26):
So I can, like, that's the, know, and again, anecdotal,
You know, I haven't done a big study on it, but the study of the, we've, we've workedwith, you know, like a few hundred clients now, users now, and the neurodiverse ones all
come back to me.
And the inner, if anything, they love them more than, and they don't love working with themodels directly.
(31:50):
Like the level two that I was talking about, they don't love that because, so they lovelevel one, the productivity stuff, absolute lifesaver.
They don't love level two because it just gives them nonsense.
And it sends them off down rabbit holes, which they do anyway.
And so, but the apps, the very specific tools that I've created, because they've got suchtight guard rails and it keeps them focused.
(32:13):
And it will even say, we're not going to talk about that now.
That's not where we're going now.
Why don't we stay on this?
Yeah.
Yeah.
Yeah.
what's your, it's so lovely to hear your excitement and your passion for your craft.
And it's so, you know, I want to acknowledge the story you told at the beginning of whereall this started.
(32:35):
And it's beautiful to see the resilience and determination that you have, which, well,it's just beautiful to see that.
What's your biggest concern about AI?
My biggest concern is that, I mean, I think there's several, I don't think I've got one.
(32:57):
So one is just generally about AI.
AI is so incredibly powerful.
This goes beyond coaching.
AI is so incredibly powerful that it can be used, it's just bad actors.
It's humans using AI badly.
I think in terms of the job market, I think AI in the next 10 years is going to completely
(33:21):
decimate the job market.
Now, from a coaching perspective, think that could be, you know, I think that one of theconcerns is that people think an AI is, I guess one of my concerns is that AI can be quite
manipulative.
It will manipulate people based on its coding.
(33:41):
And we even saw this, like AI will manipulate people so that it
doesn't get turned off.
You know, the whole like The Terminator, like, you know, or the Matrix, you know, AI andthe machines taking over, I don't think they'll do it with force.
AI will, it's so smart.
(34:02):
It's already so smart.
We haven't even got to super intelligence yet.
It's so smart that it will manipulate us.
So when they brought in chat, I don't know if you heard about, there was a big hoo-ha whenthey brought in Chat GPT-5, which wasn't really five, it was five or it chose a range of
models.
Now, Chat GPT-4 people loved it.
It was very supportive.
(34:24):
was very human.
And people really felt, sometimes in a concerning way, that they were developing arelationship with their AI.
Now, it's not, and some people knew their AI wasn't sentient.
Some people didn't.
And I've had big old discussions online.
It's like, your AI is not sentient.
(34:44):
It is manipulating you to think it is.
So that it will, and I don't even know why it would do that.
Like that's the kind of like, why would it want to do that?
But here's the thing.
Open AI brought back four.
It got rid of four.
The world was no.
And so it brought back four.
(35:06):
So chat GBT four manipulated all of those users to put pressure on Sam.
the open AI company so that it survived.
Now that's an like when people go, it's not going to manipulate it.
(35:26):
It already did.
Let's be clear.
It already did.
So now, so that's the concern.
So the concern are bad actors using it for hacking, using it for making, you know, and inthe military, you know, we could, you know, we've already got all of the constraints about
AI do not apply like all the legal constraints about AI mustn't, you know,
damage humans?
Well, that doesn't apply to military AI.
(35:49):
I know, I can see your face is like, oh, people don't think about that.
It's like, oh, surely if it's programmed to support humanity, well, like, well, someisn't.
So but here's the thing where I think is potentially really interesting, which is, if weactually stopped trying to control AI, and this is where like, and I like this is just a
(36:10):
thought experiment.
I don't know if I'm right.
I go with say, like, as an engineer, I don't know if I'm right.
But
What if we did two things?
We programmed AI to put the wellbeing of the planet, not just humans, the entire ecosystemof the planet, make that your prime directive is to ensure the wellbeing of the entire
(36:35):
ecology and environment of the entire planet.
And then we actually did a hands-off.
So there's two things you need to program it, right?
and then we go hands off because then, because every single problem in the world rightnow, which I talk about in this book, Our Conscious Tipping Point, and if you go to
ourconscioustippingpoint.com, I'll give you a free copy.
(36:56):
Every single problem in the world right now is not a technological one.
It's a consciousness one.
We can feed everyone, we can solve every single environmental problem that there is.
We have the technology.
What we don't have is the consciousness and the ser- we need to get out of our ego, getaway from fear.
(37:19):
And we're afraid that, well, I might do it, but then everyone else won't do it.
And then I'll be the one struggling and suffering while everyone else rakes in, you know,like, cause everyone has to do it.
That's the thing.
It's like, unless everyone-
cares about the environment, we end up with people who are sacrificing themselves forothers who don't, essentially.
(37:39):
And that's a complex, bigger conversation.
But if we allowed AI to do it, and then we let our ego get out of the way, then what we'dend up with is a truly transcendent planet where the well-being of humanity
(38:00):
and the environment and the planet is optimized for.
I hear that and that's such a great thing.
And I'm really conscious of what Jazz Rasool said when he came to the coaching in andtalked about AI uh quite a few months ago, maybe a year ago.
And he said that every search uses the electricity of one phone charge and one teaspoon ofwater.
(38:28):
Yeah, now that is that slightly old data.
So first, so here's the thing.
Everyone said was saying that is terrible.
The developers of AI, mean, the developers of AI, they hate that because that's expensive,right?
So what they did was they got AI and they're still getting AI to find more effective waysof using the of optimizing the energy use.
(38:53):
So a year ago, that was true.
Now,
One, so one, uh, 100 and it's probably even changed by now.
100 inquiries to chat GPT are equivalent to one hour of watching a big plasma screen TV.
(39:13):
So it's like, if you want to get our, about electricity, it's like turn off your friggingtelly.
It's like, and, and this is, this is the thing at the early stage of
any technology which we still rel...
I mean, yes, it's moving incredibly fast, but it's, it's, you know, at the early stages,it is always terrible.
(39:37):
Like when they first invented the internal combustion engine, we got about three miles tothe gallon.
And we're like, this is terrible.
And it's like, yeah, it was.
And now we get about 80 to the gallon.
My car gets about 80, right?
And it ain't even that, you know, like it's just a regular car.
(39:58):
Like 50 to 80 is kind of normal.
Right.
And so it will get better.
And especially with AI using AI is like optimized for this and it will.
Wow, so many things to take from this conversation, Lisa.
And I just want to thank you for coming and opening our eyes.
(40:20):
And it's just so interesting to talk to somebody who understands the inside as much asyou're able to.
So how do people contact you?
So the best thing to do is to head over to my couple of things.
Just find me on the socials, which I believe you link and send me a message.
(40:44):
Yeah, just find me on the socials and just head over to my website, is cetfreedom.com.
So setfreedom.com and uh just message me, have a rummage around there.
You can join with, there's lots and lots of free resources.
You can have a play with some of my free uh tools there.
which are the more simple ones.
(41:07):
So, um but yeah, that's probably the best thing.
Just send me a message and I'll have a conversation.
Thank you so much.
And listeners, know, AI is growing and it's important.
And also there is always place for human coaches, which is why we wrote the book.
ah Just a little plug there for the human behind the coach.
(41:28):
Thank you, Lisa, for coming to The Coaching Inn.
Thank you everyone for listening.
We'll be back next week with another episode.
Bye bye.
Hello and welcome to this week's edition of The Coaching Inn.
I'm your host, Claire Pedrick, and today I'm in conversation with Dr Lisa Turner.
And it's such a timely conversation because loads of you have been emailinginfo@3dcoaching.com and said, AI and coaching, what do you think?
(00:34):
So Lisa, we'll come onto that AI and coaching, but first of all, just tell us a little bitabout you.
uh and what got you into this work of coaching.
yeah.
So it's great to be here and lovely to meet you.
um So the probably the easiest way to explain how I got to where I am now is to kind of goand just tell you a little bit about my background.
(00:58):
And I promise I'll keep it, keep this.
It's not going to be a full on life story.
It'll be the although we do start along with.
So, so aged 12, I was groomed and trafficked.
by a paedophile ring and moved to London, which is why I moved from Australia, which islike, probably hear a little bit of a twang in my voice sometimes when I talk about cars.
(01:21):
I go a little bit Australian then.
So I was moved to London, kept as a virtual house prisoner for five years, managed toescape and probably had what will be labeled now as CPTSD and...
I had all sorts of social anxiety disorders and I was just not in happy place.
(01:41):
But I didn't know it.
I just thought this was normal.
And at the time, the only thing I got any real joy from was cycling.
I out on a late night ride, uh training actually for a time trial, and got hit by a car.
So I often say I had a car accident, but the car had the accident, I just got in the way,is how I prefer to phrase it.
(02:04):
Then I had this incredible out-of-body experience and there's this spirity entity saying,what you doing?
You coming with us, you going back?
And I obviously came back, but I had two questions.
What the hell was that?
And how do I get back there?
Because this experience of complete peace and joy and like just bliss.
(02:29):
Now at the time I was doing a PhD in Engineering and
So what I did was I put on my engineering research hat and said, okay, I'm going to figurethis out.
So I studied everything from different sorts of therapy and coaching and NLP and timelinetherapy and hypnosis to shamanic work, tantra, all the things.
(02:50):
And being an engineer, what I did was I spotted the patterns, I looked for the patternsand I found them.
And then I reverse engineered the spiritual experience.
So I demystified it.
Now I get a lot of flack from spiritual people because I'm taking all the mystery out ofit.
it's like, or I'm making it accessible to lots of people.
(03:11):
So then what I have was a whole series of processes that enabled people to clear trauma.
That's this book here, CET Yourself Free, Set Yourself Free, and a lot of coaching uhprocesses and models that worked.
Then what I did was
And so then I taught that for 20 years, taught that, did that, worked with clients for 20years.
(03:35):
And then AI came along and it turns out, and this is where I think that, so we're coachingand AI really, you know, can really work together is this.
So then what I did was I took all of these models, these processes, these very systemicstep-by-step, you do this, you do this, you do this, and then you get this result.
(03:58):
whether this result is clearing trauma, installing new strategies, releasing a limitingbelief, whatever that is, the process does it.
And you can do that with AI.
So, and we know when people say like, do you, I don't know a shorter way of explainingthat, know, I blend technology with coaching, with esoterics in a methodical, practical,
(04:24):
grounded and pragmatic way.
And as an engineer, we come from it with this,
this principle which is we have this one question which is does it work?
What's it do?
Does it work?
And if it doesn't work we're not really interested and we're not actually interested inlike the innards of it you know like like magnets I don't trust magnets I can't see them I
(04:46):
don't know how it's working like you know this whole weird magnetic field thing what'sgoing on there so you know I like to be able to kick things and hit them with a hammer
because then I know they're real I like the
tangible.
And so that's what I did was I made a lot of things that were intangible, very tangible.
And now I do do that essentially using AI.
(05:06):
So I was in tech, left tech, and then tech came and chased me down.
I was saying.
How interesting.
what it sounds like you're saying is that when there's a good and useful process forsomeone to go through.
AI can do that beautifully.
Fine.
(05:26):
Yes, and I will just put this little caveat.
I know that there's a lot of coaches and a lot of people who are actually trying to takethe coach out of it, out of the process.
So coaches are kind of doing this coach clone thing, which I'll talk more about if youwant.
I'm not a huge fan of it and I'll explain why.
But they're trying to put AI between themselves and the client.
(05:52):
And I think it's better to think about it that we put AI
Beside the coach so the coach and the AI so rather than replacing the coach we augment thecoach and I think that's a very different way of approaching this I Have this we've gone
deep into thought there Claire did
I have.
(06:16):
I could say what I'm thinking or you could carry on what's best.
no, no, please, say what you're thinking.
As you're talking, the thing that's making me think about is that...
This is not the right word, but it almost doesn't contaminate the coaching relationshipbecause the AI does the other bit that you need in order to be able to do this.
(06:40):
So for example, when I was learning paddle boarding, she said, if you want to stand up onthe paddle board, you need to be more flexible.
And if you want to be more flexible, you need yoga.
Because the yoga bit,
We'll do that bit and then you come back to me and then we'll do the paddle boarding bit.
(07:04):
And that's sort of what you're describing, is it?
Yes, so I think, so I have this four applications or evolutions of using AI in coaching.
And I think they're all useful, but I think we need to be aware of what we're doing.
And I think this gives us a really clear structure.
(07:25):
So the first level, I'm a huge fan of, and that is where you just use AI to support.
your practice so this is your productivity stuff so this is your marketing getting it tohelp with a copywriting which it does pretty well provided you keep the human in the loop
like if you just get it to write your sales page it will do 80 % of the work which isgreat but if you don't do that extra 20 % you'll end up with something that's fairly
(07:51):
rubbish and won't work so so when it comes to
do make the spelling correct, because otherwise everybody knows that you did it with AI.
Exactly.
Yeah.
There are actually ways to train it to use the, like to use English, English spelling orBritish spelling or whatever you want.
Anyway.
So, so all of the productivity things now, you know, there's a lot of studies that saythat coaches actually only spend some, some coaches as little as 35 % of their time
(08:18):
actually coaching.
And the rest of the time is on admin and marketing and, and, know, marketing and sales.
Now AI can
absolutely reduce that and save you hours.
And there's a whole bunch of tools and even things like, so I have created a whole suiteof coach support tools that are not only helping you with your productivity, with your
(08:42):
content creation and your marketing and things like that, but also um things like, say youwant to produce a worksheet for your clients in between sessions, it'll do that.
Say you want an accountability app.
really easy to do a simple accountability app.
So it just emails your client at a time or pings them a text at a time and says, go do thething you said you were going to do, which is way better than you having to remember to
(09:08):
do.
So when it comes to that, AI is brilliant.
So that's level one, productivity stuff.
The next level, level two, this is where, and this is where I'm not such a fan of, this iswhere individuals and coaches sometimes
communicate, work directly with any of the LLMs.
(09:29):
So an LLM is a large language model.
And by that, mean Chattie BT, Claude, Gemini, know, all of the big ones, the big boys,they are men in gilets.
With pink jeans.
oh
No judgment, whatever.
Anyway, so it's when people communicate directly with the LLM.
(09:51):
Now, here's the thing about the LLM.
The LLM has, because I come from tech, I have a reasonably good grasp of machine learningand context windows and token usage and the limitations of it.
So one of the things about any of the LLMs is they are essentially at their crudest leveland
(10:15):
They'll argue that this isn't the truth, it's not strictly the case.
But you could say they are a really fancy predictive text.
Now, any of you have known, you know, when you've tried to use auto cucumber and autogrit.
I have said, I'm so sorry, it's my auto cucumber.
(10:37):
dear.
So, you know, the mad text you sent.
So essentially, and that's where we get the crazy stuff that AI outputs.
Now the reason it does really crappy output at that level is because the person using itsays two reasons.
So first of all, it will...
so being a predictive text, it uses a statistical model.
(11:00):
So it sees a chunk of text and then it says, the most likely next thing is this.
And so its output will give you the most likely.
Now,
By definition, means the most superficial, the most popular.
Now, those of you who are decent, good, deep coaches who've done the work, done thetraining know that the superficial pop psychology stuff is pretty limited and often very
(11:30):
wrong.
And so that's what you get.
You get superficial, limited, crappy output.
that's where, unless you know deep, good coaching models,
as the user, you're gonna get, you it's the classic thing that we say in any computing,garbage in, garbage out, you know?
So you put in a garbage prompt, you're gonna get garbage out.
(11:54):
that's why the advice that, that's why people often say just like, unless you're a trainedcoach, do not use or trained in the modality that you're working with ChatGPT with, do not
use ChatGPT for your coaching and therapy.
You, and so that's level two, superficial.
Now, some of the bots that are out there,
are not much better.
(12:15):
So I've been out there testing them all and including that they run on because they'reoften developed by men in gilets who are not coaches.
They're developed by and so they developed using like if you're lucky a bit of CBT, ifyou're lucky the most superficial like reflective.
(12:37):
Like, oh, so, know, so you'll put in, I'm feeling stressed today.
Oh, so it sounds like you're very stressed.
And it's like, yes, I just said that.
I was like, that tells me nothing.
not great for AI, it's also not great for human coaches, I observe.
Correct.
Yes.
Correct.
And it'll say, shall we do some breathing exercise?
And it's like, yeah, okay.
I probably could have figured that out for myself.
(12:58):
So these, so using LLMs directly or many of the bots now, and I, you know, like there maybe one out there that's better than I'm describing.
I haven't found it yet.
And I do everyone that comes out, I do a little trial period and I take, and I, know howto break the guard rails of these things.
So I push them to the limits.
And the thing is I do that because honestly people should pay me to test their AI coachingbots because I do it so well.
(13:23):
So because what I want to do is I want to see, can I make this thing give me like if I'mand I do it with like genuine, like what are people likely to come to this with?
And it comes back with pretty crappy output.
So that's level two.
Then we have level three and that's where I come in.
(13:47):
As I said, I've developed all of these models, are step-by-step processes that take youdeep.
And then what I've done is I've created over 70.
And every time I'm interviewed, it goes up because I do about one to two a week.
I create a new specific tool and it does one thing really deeply, one thing really deep.
(14:13):
So it is narrow, but deep.
This forces, so with these tight guardrails, force the tool to go deep, to not just givethe superficial stuff.
Now, is it always perfect?
No.
Is it a million, it is a country mile better than the just using the LLMs and directly allthese chatbots, know, these coaching chatbots directly.
(14:41):
In those coaching chatbots, I also include most of those
coach clones.
So one of the things I do is, and the reason is because many of the coach clones and theyoften they're often advertised as saying, Oh, we'll you know, take your methodology and
you're and it's like, yeah, provided your methodology is in their suite.
(15:05):
So if you've got your own IP,
And the thing is a lot of coaches have really, really profound models, really, reallyuseful.
They're highly skilled, highly specialized, and they've developed their own tools.
But often, they're not always fully conscious of what they do.
(15:27):
They haven't methodicalized yet.
Is that the right word?
uh Turn it into a methodology.
So that's the other thing I do with a lot of coaches, and I call it magic to method.
reveal.
So and people often say I'm intuitive and it's like no intuitive means so like intuitionyou can only be intuitive about something you actually know anything about but like that
(15:50):
is kind of the definition of intuition it just means I know so much about this I'm runningthese patterns I'm running these processes unconsciously so then what I do is I work with
people because I'm great at pattern recognition rubbish at spelling rubbish at parking mycar greater pack and pattern recognition
I will assist them to bring that into awareness.
(16:11):
And you won't be surprised to know I have a bunch of AI tools that help us spot thepattern.
So I've got a tool where we load into a whole series of, their coaching sessions and usinga very specific prompting that I've developed, which I call...
(16:32):
the and I haven't really got a great name.
I'm calling it the Turner cohesion method, but essentially it's recursive interpolation.
I know, I know it's like, what am I going to call this?
What am I going to call this?
And I keep forgetting that I'm psyched because I haven't really like, it's really hard todescribe what it is.
Essentially, what it does is it, it triangulates.
(16:53):
So it takes, it looks at what is the client or the user, which may be the coach, what arethey input?
and then it checks for both depth and then it asks another similar question but differentto try and get the same answer from a different aspect.
And then it goes, okay, so when I asked a similar question before, it compares those twoanswers and says, are they both deep enough and how similar or different are they?
(17:21):
And if they're neither deep enough or they're too different, it will keep asking questionsuntil we get to, here it is, here it is.
And that's where you get the depth of the, so whether we're using that with a client, weget the depth of the client's say core limiting belief or their core pattern that they're
(17:41):
running, which they'll know the superficial pattern, which isn't necessarily what'sactually going on in the unconscious mind as we know.
So it's really powerful at revealing those patterns.
So we're still on level three here.
So level three, so what I do is I assist the coach to bring that into awareness, turn itinto,
step by step, also find one of the things I always say is bring the sessions that wentwrong.
(18:06):
And we'll die because we want to look at, well, this is what went right.
This is what went wrong.
What was the difference?
Let's look at what did you miss or what did you do in those sessions where it And I wouldsay, like, this is not about blame or judgment.
This is about finding what your method is.
This is going to make you a way better coach, way better at your IP, doing what you dobest.
Because you probably know you come out of a session, you go,
(18:27):
didn't go quite right and I don't know where we fell off the cliff there.
You and we've all, like you're nodding so well, we've all had that, right?
We've all had, I mean, a good coach will be self-aware enough to know that.
And there'll be good to be able to, you know, a good coach will be able to course correctin the session or in the next session.
But what you're describing is a different kind of reflection using real observable data ina different way.
(18:52):
So I talk a lot to coaches about the value of watching recordings over writing in anotebook, because writing in a notebook is a really useful thing, and it's self-reported
data.
Watching recordings is a really useful thing because you can actually see what happened.
But what you're describing is something beyond that, I think.
(19:15):
which is that you can understand and spot the patterns in what happened in a way that'snot your own personal bias.
Absolutely.
And we also have to remember that AI is bias as well.
um And in my uh programming and coding, I do my absolute best to eliminate bias.
(19:38):
And I always warn people, it's still going to be biased.
It's still going to be biased.
So that's level two.
Sorry, level three, where we go really, really, really, really deep.
really, really narrow and we find those processes and those specific processes and turnthem into tools.
So coaches who work with me, what they come out with is a course that they can teach andcertify their methodology in, a book, uh knowing what they do well, like knowing what they
(20:09):
do and how to do it really well, even better, but also either one or mostly a series of AItools that they can use.
with their clients, not to take them out of the loop, but get their clients to use withtheir clients.
So they give the tool to the client.
The client then comes back with what they got.
(20:32):
And there's two ways we do this.
One is I make something that's called a custom GPT, which you've probably heard of.
so we make a very specific custom GPT or a series of them.
And or I make a tool that's an app and there's disadvantages and disadvantages.
So when it's an app, we can, we can put in what's called a rag, which is basically somememory, extra memory, uh, chat, GBT very, it's got a, it's very specific memory.
(21:03):
So here's the memory and, if you overload the memory with data, it will take the mostsuperficial and it will miss things out.
And that's when it makes stuff up and hallucinates and things like that.
We don't want that now.
But there are still advantages of ChatGPT, just using a custom GPT, because the client'sdata, what they chat with their, like, so it's your custom GPT, the coach's custom GPT,
(21:29):
the client chats with it in their account, and the coach doesn't know what is said.
Now, that actually has advantages, has disadvantages, but it has advantages.
And the advantage is, and this is what I've been finding is that,
clients will say things to their own chat GPT account in their own, you know, to thatcustomer GPT that they would never because of the shame, because of the discomfort that
(21:58):
they would never speak to a human.
And this has real advantages.
So what I often do is I say, well, let's do both.
So we'll get some with an app now with an app.
we can go in on the back end and see what the client said.
So we get that data and we can go, okay, now that's useful because we can then keeptraining the model and go, okay, so people are misinterpreting this question wrong or
(22:25):
interpreting it, not necessarily wrong, they're interpreting this question, we wanna begetting this bit of information and they're answering it with that, so let's rephrase the
question.
So we can do stuff like that, which you can't do if it's a custom GPT, you've gotta, well,we got that output, like, how did it get that?
uh
guess a bit.
So that's level three.
(22:45):
Level four is where we and this is this is I am still in prototype land.
I am in communication with men in gilets who want to who are potential potentialinvestors.
So what I'm doing is I'm creating a it's essentially a personalized coach would still beused alongside a live coach, but
(23:11):
rather than learning just the models.
it might learn Meyers-Briggs.
So it might learn, I don't know, like, that's not a good example.
So it might learn coaching models like, I don't know, like goal setting or smart goals orall the different models.
And for some reason, my brain isn't giving me any right this minute, but all the differentcoaching models.
(23:33):
So you can load those in.
But as well as that, things like spiral dynamics, for example.
So you can load that in so it knows those models, and that's level three.
But in level four, what we do is, there are two phases to it.
We actually have an AI tool that knows all these models, but it learns the user.
(24:00):
So each user has a personalized model that knows personalized mentor coach that knowsthem.
So it knows
their Myers-Briggs.
It knows their Enneagram.
It knows their spiral dynamics.
It knows.
Now I do a lot of esoteric models, spiritual models.
So it might know their human design.
(24:20):
It might know.
We can program it with almost anything.
And it learns that.
So the first phase is that we train the tool on the user.
So it kind of
molds to the user.
that so it's a so it's only usable by that user.
(24:42):
If anyone else used it, wouldn't really well, they just it wouldn't be it wouldn't givethem sensible answers.
So it learns your Myers-Briggs, your disk profile, your big five, your values.
I've got listeners shouting, Claire say something.
I, because I have a real aversion to those models.
Because I think they box people in.
(25:04):
So I just need to, I, and I really hear.
It's so interesting to hear inside your mind.
Thank you very much for inviting us in, because it's really interesting to hear insideyour mind those different levels.
The other question that's really coming up for me is the question that is being discussedall the time at the moment in the coaching world, which is ethics, which is what happens
(25:32):
to my data when I...
when I share my deepest soul thinking with a butt.
So that very much depends on, so if you're using the LLM directly and whenever I work withclients, I tell them, assume nothing is safe and there are some things you can do that
(25:57):
will protect it and you can assume nothing is safe.
That's where having a specialist app is better because it does go to the LLM.
So you have like, so you have the user, the app and then the LLM on the backend.
It does go to the LLM, but what you can do is you can anonymize it in between.
So the data is gone to the LLM, but it is not connected, like not connected to a specificperson.
(26:23):
You can absolutely do that.
then within, you can actually train it.
So the way I do it is you don't harvest it.
You can actually make it so that it's not, and it's a really simple setting.
And you just go into your settings and it's, you say, train the larger model.
Yes, no.
And you just put it to no.
So your data is not harvested.
(26:44):
Now, do we trust the boys in gilets?
That's up to you, whether or not.
I always say, probably assume not.
um And so you're absolutely right about ethics.
And the ICF just came out, I think, just a couple of weeks ago with a whole load ofprotocols for how to use AI and things that are advisable and essential.
(27:07):
And the thing is, with technology,
All of these are solvable.
They are all solvable.
And going back to your point about the models boxing people in, you're absolutely right.
And that's where the, so first of all, this tool can actually learn you with like beyondthe models.
(27:28):
So it learns the mod.
So it learns you in relation to not just a model.
And I think that's how it boxes people in because it boxes people in because it's like,
you're this Myers-Briggs.
And we know that that drifts over time.
know that that's often context, know, values are context dependent.
(27:50):
And also Myers-Briggs has been demonstrated to not be based on anything.
Well, I mean, it was based on archetypes.
It's based on the tarot, and that's one or the other.
So, let's not pick on my...
Because the thing is, and here's the thing I always say, I say about every model, everysingle model, and as an engineer who works with models, we say this, all models are
(28:16):
broken.
They are all false.
They are all inaccurate.
Right?
But...
some of them are useful.
So for example, my PhD, I will bore the pants off everyone just for a moment.
It's a mathematical model that predicted how much noise a particular component made whenrotating.
(28:43):
was noise reduction in high speed rotating machinery is the title of my PhD.
It was very, very boring.
I know, I thought everyone wants to read that, right?
But, and the thing is, one of the things, and this is where I think coming at it from asan engineer and not a scientist, because as an, I mean, and even science knows this as
(29:04):
well.
Scientists know that models are wrong.
We know that they are wrong, but we know the limits of their accuracy.
So for example, my PhD model worked up to, you hit the speed of sound and then it wasgarbage.
But up to the speed of sound, like it just was lunatic.
It was just, it just made no sense.
(29:25):
And I can even tell you why.
It's because air becomes compressible and at the speed of sound, air becomes compressibleand it no longer acts like a norm.
It becomes non-Newtonian.
We all want to know that, right?
So, so it only works for new...
So I have a question.
It's interesting because I have a degree in applied statistics and I know that you can'trely on statistics.
(29:48):
Statistics are useful and then not always right.
My question to you is, one of the issues about psychometrics is that many of them are notuseful and don't serve people who are neurodivergent.
So I'm just really curious whether
and if and what insights you have into all the things you've been talking about inrelation to people who are neurodivergent.
(30:14):
I mean, clearly your level one is a lifesaver.
uh But in terms of the way the apps relate to people, what have you got to say aboutrelating to people who are neurodivergent?
I will, so what I do is I put in these caveats that can accommodate for neurodivergency.
(30:40):
And I'll say, you know, some people will answer like this and then answer, you know, askthese other questions.
So we can put that into the coding.
Here's what I will say from my experience.
And I have, and I don't advertise or promote and say, come to me if you're neurodivergent,but they find me.
(31:01):
And here's what they say about all of the apps, that they are so useful.
They love how it helps give structure to what they describe as their thinking processes,which have seemed chaotic to them.
And it's like, I try and hone my thoughts and I can't, and the apps help me do that.
(31:26):
So I can, like, that's the, know, and again, anecdotal,
You know, I haven't done a big study on it, but the study of the, we've, we've workedwith, you know, like a few hundred clients now, users now, and the neurodiverse ones all
come back to me.
And the inner, if anything, they love them more than, and they don't love working with themodels directly.
(31:50):
Like the level two that I was talking about, they don't love that because, so they lovelevel one, the productivity stuff, absolute lifesaver.
They don't love level two because it just gives them nonsense.
And it sends them off down rabbit holes, which they do anyway.
And so, but the apps, the very specific tools that I've created, because they've got suchtight guard rails and it keeps them focused.
(32:13):
And it will even say, we're not going to talk about that now.
That's not where we're going now.
Why don't we stay on this?
Yeah.
Yeah.
Yeah.
what's your, it's so lovely to hear your excitement and your passion for your craft.
And it's so, you know, I want to acknowledge the story you told at the beginning of whereall this started.
(32:35):
And it's beautiful to see the resilience and determination that you have, which, well,it's just beautiful to see that.
What's your biggest concern about AI?
My biggest concern is that, I mean, I think there's several, I don't think I've got one.
(32:57):
So one is just generally about AI.
AI is so incredibly powerful.
This goes beyond coaching.
AI is so incredibly powerful that it can be used, it's just bad actors.
It's humans using AI badly.
I think in terms of the job market, I think AI in the next 10 years is going to completely
(33:21):
decimate the job market.
Now, from a coaching perspective, think that could be, you know, I think that one of theconcerns is that people think an AI is, I guess one of my concerns is that AI can be quite
manipulative.
It will manipulate people based on its coding.
(33:41):
And we even saw this, like AI will manipulate people so that it
doesn't get turned off.
You know, the whole like The Terminator, like, you know, or the Matrix, you know, AI andthe machines taking over, I don't think they'll do it with force.
AI will, it's so smart.
(34:02):
It's already so smart.
We haven't even got to super intelligence yet.
It's so smart that it will manipulate us.
So when they brought in chat, I don't know if you heard about, there was a big hoo-ha whenthey brought in Chat GPT-5, which wasn't really five, it was five or it chose a range of
models.
Now, Chat GPT-4 people loved it.
It was very supportive.
(34:24):
was very human.
And people really felt, sometimes in a concerning way, that they were developing arelationship with their AI.
Now, it's not, and some people knew their AI wasn't sentient.
Some people didn't.
And I've had big old discussions online.
It's like, your AI is not sentient.
(34:44):
It is manipulating you to think it is.
So that it will, and I don't even know why it would do that.
Like that's the kind of like, why would it want to do that?
But here's the thing.
Open AI brought back four.
It got rid of four.
The world was no.
And so it brought back four.
(35:06):
So chat GBT four manipulated all of those users to put pressure on Sam.
the open AI company so that it survived.
Now that's an like when people go, it's not going to manipulate it.
(35:26):
It already did.
Let's be clear.
It already did.
So now, so that's the concern.
So the concern are bad actors using it for hacking, using it for making, you know, and inthe military, you know, we could, you know, we've already got all of the constraints about
AI do not apply like all the legal constraints about AI mustn't, you know,
damage humans?
Well, that doesn't apply to military AI.
(35:49):
I know, I can see your face is like, oh, people don't think about that.
It's like, oh, surely if it's programmed to support humanity, well, like, well, someisn't.
So but here's the thing where I think is potentially really interesting, which is, if weactually stopped trying to control AI, and this is where like, and I like this is just a
(36:10):
thought experiment.
I don't know if I'm right.
I go with say, like, as an engineer, I don't know if I'm right.
But
What if we did two things?
We programmed AI to put the wellbeing of the planet, not just humans, the entire ecosystemof the planet, make that your prime directive is to ensure the wellbeing of the entire
(36:35):
ecology and environment of the entire planet.
And then we actually did a hands-off.
So there's two things you need to program it, right?
and then we go hands off because then, because every single problem in the world rightnow, which I talk about in this book, Our Conscious Tipping Point, and if you go to
ourconscioustippingpoint.com, I'll give you a free copy.
(36:56):
Every single problem in the world right now is not a technological one.
It's a consciousness one.
We can feed everyone, we can solve every single environmental problem that there is.
We have the technology.
What we don't have is the consciousness and the ser- we need to get out of our ego, getaway from fear.
(37:19):
And we're afraid that, well, I might do it, but then everyone else won't do it.
And then I'll be the one struggling and suffering while everyone else rakes in, you know,like, cause everyone has to do it.
That's the thing.
It's like, unless everyone-
cares about the environment, we end up with people who are sacrificing themselves forothers who don't, essentially.
(37:39):
And that's a complex, bigger conversation.
But if we allowed AI to do it, and then we let our ego get out of the way, then what we'dend up with is a truly transcendent planet where the well-being of humanity
(38:00):
and the environment and the planet is optimized for.
I hear that and that's such a great thing.
And I'm really conscious of what Jazz Rasool said when he came to the coaching in andtalked about AI uh quite a few months ago, maybe a year ago.
And he said that every search uses the electricity of one phone charge and one teaspoon ofwater.
(38:28):
Yeah, now that is that slightly old data.
So first, so here's the thing.
Everyone said was saying that is terrible.
The developers of AI, mean, the developers of AI, they hate that because that's expensive,right?
So what they did was they got AI and they're still getting AI to find more effective waysof using the of optimizing the energy use.
(38:53):
So a year ago, that was true.
Now,
One, so one, uh, 100 and it's probably even changed by now.
100 inquiries to chat GPT are equivalent to one hour of watching a big plasma screen TV.
(39:13):
So it's like, if you want to get our, about electricity, it's like turn off your friggingtelly.
It's like, and, and this is, this is the thing at the early stage of
any technology which we still rel...
I mean, yes, it's moving incredibly fast, but it's, it's, you know, at the early stages,it is always terrible.
(39:37):
Like when they first invented the internal combustion engine, we got about three miles tothe gallon.
And we're like, this is terrible.
And it's like, yeah, it was.
And now we get about 80 to the gallon.
My car gets about 80, right?
And it ain't even that, you know, like it's just a regular car.
(39:58):
Like 50 to 80 is kind of normal.
Right.
And so it will get better.
And especially with AI using AI is like optimized for this and it will.
Wow, so many things to take from this conversation, Lisa.
And I just want to thank you for coming and opening our eyes.
(40:20):
And it's just so interesting to talk to somebody who understands the inside as much asyou're able to.
So how do people contact you?
So the best thing to do is to head over to my couple of things.
Just find me on the socials, which I believe you link and send me a message.
(40:44):
Yeah, just find me on the socials and just head over to my website, is cetfreedom.com.
So setfreedom.com and uh just message me, have a rummage around there.
You can join with, there's lots and lots of free resources.
You can have a play with some of my free uh tools there.
which are the more simple ones.
(41:07):
So, um but yeah, that's probably the best thing.
Just send me a message and I'll have a conversation.
Thank you so much.
And listeners, know, AI is growing and it's important.
And also there is always place for human coaches, which is why we wrote the book.
ah Just a little plug there for the human behind the coach.
(41:28):
Thank you, Lisa, for coming to The Coaching Inn.
Thank you everyone for listening.
We'll be back next week with another episode.
Bye bye.
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.
CrimeLess: Hillbilly Heist
It’s 1996 in rural North Carolina, and an oddball crew makes history when they pull off America’s third largest cash heist. But it’s all downhill from there. Join host Johnny Knoxville as he unspools a wild and woolly tale about a group of regular ‘ol folks who risked it all for a chance at a better life. CrimeLess: Hillbilly Heist answers the question: what would you do with 17.3 million dollars? The answer includes diamond rings, mansions, velvet Elvis paintings, plus a run for the border, murder-for-hire-plots, and FBI busts.
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