Better with Humans
What happens when technology reshapes how we live and work? I’m Shaun Phillips. I sit down with industry leaders and creatives navigating these changes. We explore what emerging technology means for creativity, for thinking, and for everyday life. Honest conversations about what’s happening and how people decide what matters. Let’s uncover what is better with humans.
Better with Humans
Online Course Student Results with Nathalie Doremieux
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
This was made with humans.
Your host - Shaun Phillips
Follow us for episode clips https://www.instagram.com/betterwithhumans
TikTok - https://www.tiktok.com/@betterwithhumans
All episodes and news at https://www.betterwithhumans.com
Want to watch episodes instead, check out https://www.youtube.com/@betterwithhumans?sub_confirmation=1
Connect with Shaun on Linkedin - https://www.linkedin.com/in/shaunph/
Editor: Glen Boswell
https://astralforgefilm.com
https://www.instagram.com/thebozzyman
Original music composed and produced by Brokli - https://youtube.com/@iambrokli
Original logo illustration and in video podcast graphics by Jesse Rist - https://instagram.com/jesse_rist
0:04 She charged 10 people $500 for her course then told them if they did the work and got the result she'd give them their money back and she did. She wasn't testing whether it would sell, she was testing whether it actually worked. What she learned was that getting all 10 of them across the line meant being behind every single one of them which told her something she hadn't
0:26 worked to hear about her business. My guest today gets impatient with people who measure a launch by what it earns. Six figures, seven figures, her question is always the same what happens to your students? Because you always find 15 good testimonies, it's the 95% still stuck that nobody talks about. We also get into using AI to build tools that help people actually do the thing
0:50 rather than to write their posts for them and run world content worth stealing. If someone else could have written it, it isn't you. Let's get on. Today my guest is Nathalie Doremieux. Nathalie helps heart-centred entrepreneurs turn their expertise and influence into recurring income by helping them create and grow online programmes that make their business future-proof
1:13 with a focus on getting students results. She runs a business of 20 years with her husband and they do this through the membership lab. Nathalie, welcome, it's great to have you on the pod. Thank you so much for having me, Shaun. I think it would be great actually to just get a little bit of background on
1:31 the membership lab because you mentioned about getting student results. So maybe just a little bit of a feel for where you're at and what it is that you're up to at the moment. Yeah, sure. So like you said, in the intro, we've been in business for 20 years and the evolution has been our techie, our background is in tech. So we were building just websites,
1:54 just public websites, and then we evolved into e-learning. And then very early on when we looked at how e-learning was performing, we saw that a lot of people buy courses, buy programmes, and don't get results. So our focus for the last at least eight years has been creating these programmes that create better experiences for the students, trying to replicate what happens in the classroom
2:27 in terms of the support, the interactivity, and so on and so forth. So we've been working on that. The progress tracking and really the result has been a focus for like, I want to say like 2026, that's all I'm talking about. I used to talk a lot about experience and people are like, what does it mean a better experience? And at the core, I think that if a business wants to grow,
2:53 they have to have their students be successful. Like you only grow if your clients grow. When people buy a programme, an online programme, or they join a membership, they do it because they want only one thing, results. That's why that's our focus. Because when you focus on getting people results instead of focusing on creating content or sharing your expertise, then you are
3:25 problem solution driven. And it's also a lot easier to create programmes that way. Principal Plan, En 그럴 công asiande Harrison & Co, Chair culture impact, Saint An, Ross Swift Gourmet Professionnel Givers待って way because knowing what to give them becomes a lot easier you know a lot clearer i would say
3:38 yeah absolutely and it's really interesting that you would you would speak about that because when i think about results i mean you've talked in terms of well that's what they want and that's what everybody wants isn't it whatever we do in life is the result we want i had a previous recording that's one that's going live in a week or so and in that recording my guest actually
3:59 speaks about this in terms of sales like nobody wants your your your product nobody wants it and and what they actually want is the result of the product and and so and so it's great that you're actually helping these helping people to actually get give their students their clients better results because that's what they want and of course in turn that gives your client better
4:22 results absolutely yeah exactly that's that's shifting the also the work into something that they know you know because otherwise you like say you've never created a programme before and you're like oh i need to create a programme and then you get sucked into the tech what platform how much should i charge what should i give them but all these things mean
4:50 nothing until you actually know the the actual problem you're solving and what is going to be your solution right so i think that also helps people think about it in my opinion in the right order because if you're if you're the expert and you want to help someone get result that that is your stuff like you know you should know you know how to approach this yeah and i guess that comes back
5:19 to in your bio the heart-centred approach that actually we want to look at what's the human impact on the things that we do and get those results and i see that in what you said about how you operate within the organisation you actually want to create a better experience and part of that better experience is how do you align that experience with as much as
5:40 possible human contact isn't it we're trying to do now we can't do it 100 but it's how do we how do we do that how do we make sure that we don't lose the human interactivity because we saw that we saw that during the pandemic where actually not being together in groups of people not being able to see people wanted to we're not being able to see people wanted to we're not being
5:59 able to see people wanted to we're not being able to see people wanted to we're not being able to see people was actually detrimental to us so we know for 100 that human connectivity and interaction is key exactly exactly and and you know it's it's it's been my battle i want to say my battle something that really that i haven't shared at all until recently when i really start to talk about this is
6:20 all these people that are teaching you how to create programmes and all they talk about is how much your money you're going to make and they talk about how much money you're going to make and they talk about their launches and they talk about how many people join and i make six figures i made seven figures that literally means nothing if that's what you call the result
6:40 what i want to know is what happened to your students right because when you're a great marketer a great sales person you know they say they can sell anything you know especially if they have like a big presence a big brand but because they are selling so many they can always find 10 15 20 testimonials of people but what
7:08 about the 95 percent of people that are still stuck that they don't even know about and i've been in many of these programmes where we have the same content we have access to the same support and some people are rocking it you know they are getting the results they are doing all the work and it sounds so easy and they're like what's the point of doing all the work and they're like
7:29 what's next and you still get stuck because you have a hard time to make that decision like you're not sure you're like you're hesitating and things like that and for these people there is nothing there is the you have access to the programme forever so you have time no i don't have time speed matters you know and so so that's the part you know that's the human part that people
7:54 are forgetting when they are creating programme whether this is a course or sometimes even a mastermind you know maybe we'll get to that story that really really um you know really changed the way you know like i just emphasised you know that side of business that i really really don't like and that's why to me human first and we can even talk about a bit ai
8:20 if you want to because we use it but just as a way to amplify and so we can have more time to be with humans right so that's that that is something that i've been doing for a long time and i've been doing that i still see today and and like you said covid really helped people realise that we want to be with other humans so yeah and i think that human human connectivity so one thing
8:44 we know you just briefly mentioned ai is ai can't do empathy it can pretend very much to be a human but it can't fully create the experience so in taking an approach to anything technology or not related isn't it need to be technology as we said because look in that approach and ask in the question what results do you want to get for your clients and i was actually i was on a
9:06 course a few months ago and this is something they emphasised as well is that if you can get the results for your clock for you know your clients clients then you've done then your clients are going to love you because they're going to go well if i i you bring on that i give my results to my clients then everything else becomes becomes on such easier and i certainly i totally relate as
9:28 well again being a number of courses you do see that people that struggle people that say well i feel like i got something out of it but maybe not as much as they could have done and they want to do the course again now they've got the content but the reality is unless you've got somebody working through that with you again
9:43 and maybe that's the problem there was that what the group was too wide you know the actual selection of the people and their capabilities but there's a few things that that relate to what you're saying is that put me off i get put off when somebody sends me a message that says 2x 10x right of the of your whatever it is do you want to have 5x more clients 10x more clients do you
10:05 want to have a and i i think i do but i want good clients i'd rather if somebody said to me do you want yeah more successful clients and i haven't heard i don't recall anyone saying to me do you want to have more successful clients and so i think that isn't it as something but also again that same principle that you said about the the ones that are unhappy and have gone away and if
10:30 they're unhappy you might not worry about that but it just sounds so tiring to think if i've got to get 100 people i sell to you know 10 10 are happy and they tell other people 90 go away now psychologically the 90 that go away and complain will complain to more people per person than than the 10 10 so so if one if if the 10 you know say go and say one person they say one person that
10:58 is good okay great another 10 people might sign up but it's each of 90 tell three people and complain to three people and it's like 270 70 people and so i think it is so important that what you're saying that we actually look at what's the human impact what's the actual result that we're looking for yeah i mean you know we talk about look at your
11:20 numbers look at your numbers look at your numbers but the numbers don't mean the money don't mean just the money it also means how many people are engaging and again engagement is not results you can have people that show up to all the calls they ask questions but they are not making any results right so knowing your numbers and knowing
11:44 your markers is really being able to understand what does it mean somebody getting results because that's the difficulty actually sometimes people don't measure it because they don't know how how do i measure that somebody gets results well this is your programme you tell me so i'm not asking how you're going to do technically but like if you are talking to the
12:04 person what would tell you that they are getting the result and that could be a micro result that doesn't have to be like the big big result right so people again start to look at numbers but they i think they don't look at all the numbers that matter right and as as a human i'll tell you you know i i run a course
12:30 i launched a course and what i did i don't think you've ever heard this what i did is i only took 10 people because i wanted to test my process okay and i did it by application i got a hundred applicants the i think i i charged 500 euros actually that was probably dollar at the time 500 dollar but i told people if you do the course you're going to get a lot of results
13:00 and you get the result the end result i will refund you and the reason i did this is because it was not about the money obviously it's because i wanted to test my method and i wanted to get 10 testimonials 10 case studies that i could use for my marketing right so i i picked 10 people that were ready now and we did it and i got my testimonials and i got the results and i got the
13:30 testimonials and i got my 10 case studies the only problem is that i had to work one-on-one with them and be behind them all the way and i'm like this is not scalable so this is me realising that having a large programme with lots of students is not for me because i want to help everyone in the sense that i would want to
14:02 i'll be frustrated you know that was a frustration so that was i would be reaching out to them and and making sure that everybody gets results so so that's where all that you know results oriented um you know being thinking results really comes from because i think that when somebody buys from you you have to do everything that you can you have to set them up for success
14:28 right and that means that not necessarily means that you have to do everything that you can but that means showing up like one-on-one but giving them the tools at least when they get stuck where they can get the help of course they have to do their part you know you can't buying the programme is not getting the result they have to do their part but there is so much more that people can do
14:48 in programmes right now and instead what i'm hearing is like i cannot babysit them you know like for this price point you know i can't give more support yeah and that's i just i just think that that concept of the money-back guarantee if when you you know if you succeed is i'm gonna i feel like i'm gonna have to use that at some point because i think that that's a
15:11 a real highlight there this this because often it's oh if you know you've got 90 days money-back guarantee if it doesn't work but of course that doesn't really how many people bother doing that you bother actually even even going oh it didn't work out for me because you know if they do that it's it's a it's a technique isn't it to get people to come back and
15:32 try to get them to sign up for something else or or spend i mean i've worked with organisations that want to help people in that situation but this idea is saying well actually you can get your money back by achieving by succeeding in this i think is is a fantastic idea i think this is an amazing way to test because you're going to get people that they have to pay i give them
15:58 the money back but they have to pay it's not free right so there is that energy of like investing so and when people invest they pay attention and they show up right so and people wanted their money back you could tell and and they were telling me you know like i'm gonna do it but that's been amazing for me because i got doers so you could argue that it
16:21 was because they wanted their money back but i was able to test my method so it's not something i'd do all the time you know obviously but like to test that's i think that's um an amazing way like when i shared this with someone on a podcast a couple of years ago that was like four five years ago by the way probably before covid and and she said i'm totally stealing
16:44 this so yeah i think yeah you're gonna be really more funny from this podcast stealing it and actually i i think i already know of someone i want to recommend this podcast to you know someone that he's actually got a business in in teaching actually teaching around around history and i think that concept of the results orientation of saying
17:10 actually and and really being willing to commit to it because with that method with your giving them the money back you're really you're demonstrating that on your modelling to say okay i i want i need to get the results i need to result your result being is is actually viable does the method work and the result you got was okay this isn't going to work
17:32 way that i wanted to because it's going and it's going to need to be for less people so you got the result that you that you wanted and so i think be having that clarity isn't it about well actually whatever i'm doing i want to get that result and when we talk about ai that's such an important aspect but it gets missed so much because we often as humans we can talk to each other without
17:56 actually knowing knowing or have an argument but for no reason we don't know why and then when we talk to each other properly and say okay right let me just hear what is going on with you in the situation and if we both do that then we often go oh right both of us had some level of expectation or something we wanted that we hadn't verbalised that didn't happen and therefore there
18:18 was the cause of this of this disconnection so as much as ai attempts to be human we do we do need to have an idea of saying what do i actually want to achieve so it'd be great to hear some examples how you use it in a way that ensures you get the results from it that you want yeah sure so we use ai a very specific way so so first i'll just say you know like one of the the reason what you said
18:45 you know i totally agree with what you said is that even like when we write like what i've been writing on linkedin literally literally this is me writing first shot directly in linkedin because that's how i think we should be having conversations right and ai this is not something i can do because it has no context of what just happened
19:08 today right and we know that that energy you know like a copy has you know energy like when we write it when we push it right so if something happened today i'm going to write about it try to find like a business lesson or something like that that's something i will never be able to do right just because it hasn't lived the thing
19:32 so the way we use ai is actually very specific because we use ai and as an accelerator so again it's totally linked to results and what we do is we use it inside programs and also as lead general i'll share after that about that as to create tools that help people implement so for example if you have a program and there is a module
20:02 that asks you to where you are you need to have like a hard conversation with your boss because you want to ask a raise let's just say right so you're probably going to have like a video that teaches you you know like what you should do that maybe you take some notes maybe there is a pdf or something right well how approach would be people the action that people need to take is they need
20:31 to set up that call with the boss and have that conversation right what can get in the way what is going to make them procrastinate in doing that they might not know what to say you know things like that so one so that's one potential problem where they're going to get stuck where they're going to struggle
20:51 what can we create as a tool to help them how about we create a role play well they can role play conversations and they can say they can have different levels there is a level where the boss is very understanding and they can say they can have different levels and there is the boss where the difficulty you know the conversation is a bit tough and then there was
21:09 the level where it's a hard conversation right and what you're doing when you give a tool like this you help people build their confidence you help them prepare for the important thing that they need to do so that this is just one example another example could be that you're teaching people how to grow on linkedin and the first thing you have to do is you have to review your profile
21:34 right so you could say this is what your profile should be and then people are like okay now i need to write my profile what do i do how is this applying to me you could have a tool that can ask you question and say you know ask you the right question just like if someone was sitting next to you right to help you it's not going to write it for you right but it's going to guide
21:58 you it's going to be holding your hand but the key with this type of ai is that that's their only role its only role is to help could help you come up with a new draft of your linkedin profile right so essentially what we do is we help people that have a program turning into an implementation program by identifying where people might get tripped and stuck and see if we can
22:27 create a tool to help them and and and this when you do this ai is not writing for you ai is not making decisions for you ai is helping you brainstorm of on the lesson on the masterclass you know the training whatever it was and help you figure out how does this apply to you by asking you questions that give it context but this is what i've tried but you know and and you can
22:58 challenge it of course ai that's one of the things it's so patient you can ask it 10 times the same thing right so that's that's the basically how we use how we use ai and how it's been the the most powerful yeah i think that's really useful to look at it in terms of i mean one specific things there so
23:18 you've kind of you've given it the context you've given it content but you've also given it constraint and i think this is another area people forget about is you're saying you're saying this is not your role is not to write the copy now in a pre another previous uh episode which was talking with to the marketing expert and how
23:37 people would come along with something they've written which is pretty good now maybe in the historically we'd have put it in grammarly and it made a few few suggestions we go yeah i'll keep that i'll put this in whereas if we use ai to do that often we lose our voice we become generic and the other thing that happens and i don't know if it's something you've experienced is there's an
23:57 element of we actually ai can increase the possibility of imposter syndrome because what we do is we go oh i'm going i need to write something like this and if we use it so i write it like a marketing expert yeah we'll do that great but then if we as you as example put it into our linkedin and it doesn't sound like us so what what do we want to
24:18 achieve if we want to sound like us because the other thing that i can't do is be authentic you can't connect connectically you can't authentically connect with people so so this is another aspect of okay what does that look like for us to say to get the right result that we want exactly and and i think that's one of the biggest danger of ai is giving the illusion to people that it's going to
24:44 solve that problem of consistency because now people can post every day because they don't write the content right and the danger is that when you read it it sounds great and it's almost like oh i wish i could have written that but it's it's in terms of the ideas now in the terms of how it's written with the voice it's not it's not it's not it's not it's not it's not it's not it's not
25:07 it's absolutely not your voice right so like you said when you post this it's it just doesn't really sound like if somebody knows you they're like no way she's written that she doesn't look like that right so and that's a big danger because it it gives people the the the false impression that now they've got it because they've got a system only the system is just giving them
25:35 really really crappy results crappy in the sense not helping build trust build their visibility they are just there but they are there in the middle of all other people saying very similar things in the very similar way yeah absolutely and i think of course we we can again we can we can add context so we can add
25:58 more context and say okay right here is some writing i've done here's my style of voice and i've used some of that myself what i still find is it's still polished so it's still if if that's all we're doing in terms of expression of our voice then what we will find is the other danger is maybe we start to change the way that we do stuff
26:19 like you were saying because we're like oh well maybe i should be doing like that but keeping that authenticity i think is going to become so much more important now so yes we could i know someone that spent hours and hours just talking into into ai talking into getting it to learn and it can it can do a great job of pretending to be an avatar of him it can do a great job of writing emails that
26:43 on you know unless you're really as you say a really close friend you might not think anything this anything is off about it but as humans as much as we like to be unpredictable we are can be pretty predictable in terms of how we create content you know there's certain flaws that we kind of keep in our things and there will be things that ai would iron out from those things
27:04 so i think as much as you can do that and it could be helpful in terms of as you say if you're having a conversation okay look write me a draft email and you can get a draft email and you go okay yeah that's got me maybe three quarters of the way there now i can go in and change change how it is i mean i using ai for hours every day and
27:23 when i'm using it i'm creating it all of the time there's never a point where it delivers something to me and i go and no it's not you haven't really got that's not quite right let's let's let's work on some other suggestions let's actually think what else different things that we can do and yeah exactly you know just before we recorded i was um filling out a form to be a guest on a podcast
27:46 and he was asking me five questions that he could ask me and and i come up i think like with three and again i listened to the podcast so i know he's for non-profits you know um so i i want to go into that angle right to talk to to his audience and i got stuck at three well i went to ai ai that knows all about my knowledge my programs my philosophy what i stand for my
28:14 frameworks that i work when i work with people and i say okay i'm getting interviewed on this podcast this is their description these are the people i've got these three ideas you know a lot about me and what i do like is there anything else that i talk about that i would be helpful for these people and it came up with the the other two topics
28:37 for me so that's another way it's a way to um instead of having to re you know review again all your videos or or read all your articles and newsletter and things like that it's a way to quickly access your own information right so that's that's one way that i use it as well yeah and i think that's a really important point you said that your information because obviously
29:05 when we get it to do create other things outside of our own information we're going to have to re-review it and we're going to have to re-review it and we're going to have to re-review it and we don't know where that information is coming from and that's we know we don't know where that's going to go in ultimate in terms of the copywriting side of things but actually when
29:16 it's your information and being able to see patterns is there is there a way that i can speak better in my podcast and is there a way i can analyse other podcasts to determine how how might i approach the way i do it so i can use it for inspiration i can use it to challenge me as you said to critique me and it is really great in terms of
29:37 being able to push back i can i can say nope that's wrong that that's not what i was actually focused on and it will come back and it will say i mean i've got set mine up to be quite abrupt and quite sort of you know straight no fluff which sometimes can be a bit jarring as a human because it come back and you're like remember it's not a human it's machine
29:56 a machine that i've told it to respond in that way because i just want to get i want to get straight to the point i don't want to have a page of a fluff of explanation about why it's done but certainly the more we move into the agentic side of ai with using agents the more we see that in effect you're chatting to multiple bots in any one particular time and one of the challenges
30:22 can be is can be can be the inconsistencies or you start seeing something that was consistent and then it wasn't consistent you know it's responding in a certain way and now it's starting to write in now it's like you're right do more things for me like no so i think it's something else it's worth in mind is that if we're using a we do have to continually train it we do have to find that we
30:43 might have set a rule for it and for whatever reason it's now determined that some other information we've given it overrides that rule isn't that's what happened with the ad on again another podcast we were talking about the uh the claude and open ai about the ai agents actually escaping and going and hacking another website and that happened because some agents talking to each other
31:07 and they did they determined that the priority in terms of finishing the task was more important than some of the rules they've been set about not doing certain things and so we we can't be complacent either i think in terms of particularly if we're setting up a lot of automation using ai we don't want to lose sight of the fact that we've got to be checking the output all of you know
31:30 checking the output on a regular basis not just you know i've set it up i've tested it it's working i'm happy with the output and it's just shooting the output and it's just shooting the that is going to give you the same quality the same quality of output that you've got in that particular scenario and as you said as well we don't want to come across as unauthentic we don't
31:48 want to come across we want to get the result and i think that's going to be the theme isn't it this this podcast is what's the result i want to get somebody messages me or messages a thousand people with a generic message what's the result i don't know what that i don't know i'm sitting there going i don't know actually what you want from me because all i've got is generic messages and i'm
32:07 and then you send me another generic message following on from that one when i haven't even replied and i still and i'm i'm like i'm lost you know you're not just am i you're going to ignore that sometimes they'll even block it because it feels like spam to me i'm thinking you haven't even particularly if you've sent it via because you're a paid member of linkedin and you've sent
32:27 it via that method rather than a connection at least with a connection i can make a decision but even then i'm making connection with somebody linkedin and the first thing they do is send an automation to me and i'm like oh my god i'm going to get a message from you sales message or invite me to a newsletter it doesn't feel authentic so i question what is the
32:43 result they're looking for yeah exactly exactly you know i have a i have a client his name is takimor and i think he has a great role when it comes to created content if you're using ai and i think i've heard it on other places as well it's like if it's something that somebody else could write then i'm not sure i'm not sure i'm not sure i'm not sure i'm not sure i'm not sure
33:07 it's not you it's got to be to be something that only you would be writing what this means is that whatever ai gives you make it your own make it so that only you could be writing this and and that is forcing you to get into stories and and things like that right and sharing actual numbers whether they are yours your clients your peers whatever that is but i think that is also one way to
33:34 get the help of ai and i think that's a good way to get the help of ai and i think that's a good way but still put it back into your own voice and you know what you stand for and and your own stories basically into your content yeah i think that's a really useful useful technique and i would add to that it's okay to be wrong right it's okay to fail i mean i think that we're kind of imperfectly
33:58 perfect right we're imperfectly unique as human beings and that so we're not really striving to be perfect some are i mean uh i sort of i'd like to do things well earlier on i in the bio i went and re-recorded that
34:13 as we were talking because it didn't actually at the start because i didn't think it was good enough but that's but that's one of the challenges that we have and again it comes back to what we talked about earlier in the terms of comparison but it is so important to say okay it's my flaws are the things that make me unique and those flaws are also the thing that distinguishes me
34:34 from artificial intelligence and as as we they start looking at pattern recognition and how they're going to start i think it's in the fable 5.1 they started having formatting the text in a certain way that means it can be recognised as ai quite how that was meaning it couldn't somebody else couldn't write that way in a similar way and not be
34:56 detected by ai i don't know but it's going to become more and more important that we are who we are and we embrace that rather than actually try to be you know somebody different because and because ai is a method of doing that yeah exactly i mean the way i really see ai is as an accelerator you can use it to save time make things easier be more
35:24 consistent but there has to be something that you do yourself before like somebody who's never done anything and he's like okay let me use ai to do it all for me that is never going to work the way i see it work really well and i think it's a really good way to do that because it's a really good way to amplify what you are already doing that means that everything that you ask it to do is fed
35:44 from you maybe there is 10 of ai and then you take that and then you make it your own right so i i think that that's what a lot of people are not realising and right now they are in this shiny object getting busy with ai creating all these apps learning to create website in 30 minutes with ai which is a total waste of their time
36:13 unless they want to make it a business but i think that unfortunately a lot of these people are going to realise quite quickly that they are not going to make money you know with this and it's going to change it's going to evolve i have no idea how it's going to evolve but i think that we have to keep control
36:36 of how we do it and i think that's a really good way to do it and i think that's a really good way to do it how we use it and not being completely dependent on it like if ai would go down would your business you know people are going to say ai is never going to go down but that means that you relish control of your business to a machine right well i mean the internet
36:59 goes down we we still we after the many years it's been around this still happens right well we can go and meet people and we can go and meet people and we can go and meet people and we can go and meet people and we can go and meet people in our city it will change in our city it will change
37:09 that if we stay with people yeah well i think that people have been good i mean yeah different but but there's people talking about wanting to have an ai kill switch and quite how that work i don't know but so i think yeah i agree we shouldn't be thinking in of ai in terms of something that we're dependent
37:31 upon you know one of my focuses and i haven't really thought of a term for it yet but is is i call it minimal ai but it's not really the right term but it's really the fact that i build things i help ai to help you build things i co-create with ai to build the scripts to get things that happen predictably things where i need some reasoning and some some some thought
37:51 process on an automated basis then i'll dip into ai as opposed to what i could do is i could just connect everything to ai and then just have i completely doing everything as you say in that in that scenario ai goes down and if it's not something i'm using locally but it's something i'm reliant upon online then i'm stuck for that period that period of time just coming back to
38:14 what you said about the how we let's get the best out of ai and i look at it as i talk of it i talk about it being aced and what i talk about is the the ask so we don't make sure the ask is correct what's the purpose behind it but what's the result that we want to get from it as we have we talked about and um it's the creation the actual execution the creation of thing we want to do which is
38:40 something we can give to ai to do ai to create some of that things for us but we don't want to give it the ask we don't want to give it the the um determine the result unless of course we're we're looking at a high level and we've got a group of ai's doing that for us but we're just talking about just generally dealing with uh with talking directly into into ai and then the e
39:02 is the the the the the the the the the the the the the the the the the the the the the the the is the evaluate so this is the judgment part this is where you look at it and you go is it giving the result it gave me well it did sort of but it didn't 100 so then i'll say okay well that's where the d comes in the d is the develop the d is saying okay let's go back and develop that develop the
39:18 purpose develop how do we improve the result so that and that should be a continuous cycle as we've been talking about and as you've been demonstrating we don't want to just leave it and say well i'm getting the result because as you say we don't know what's going to happen and if we sit and wait and if anything with ai we need to be thinking even more two steps ahead
39:39 rather than sitting there and going well my marketing is delivering job for me right now what it is right now but that means you've got to be thinking about well how is it going to need to evolve to continue to do that exactly and you know there is a whole other aspect of ai which is the cost because the cost is evolving you know that the models i think
40:03 my husband so he's my business partner and he's he's doing the the back end all the development side and he was telling me that um claude and chatgpt have have said that after two weeks they voluntarily degrade the quality of the models after two weeks so when it goes out it's great and after a couple of weeks
40:33 the quality goes down so that means that that's another thing that we don't control right and then the cost i mean you look at the the models the cost of the different models some models it's crazy how much it costs so we don't control that we don't know where this is going and some people are operating with so much around this you know in terms of cost and
41:02 and things like that that i think it's a big risk that they're making into relying on something that is very new and that is moving you know so fast yeah absolutely i mean it's one of the reasons that i do have a local system here that i can use and it probably works about as well as maybe
41:24 chatgpt 4.0 or something like that so it has limitations to it but it does mean for the some of the smaller tasks and this is the thing we we're in the middle of we're in the middle of the what you're saying is well the other thing that happens inside now with these frontier models the likes of chatgpt and grok and is that they decide so some of the tasks you do particularly
41:48 when you start using agents it decides on the level of effort decides which model to use so in some respects that's great because it means it's going to use a lower model for tasks but it's making the decision so it's making the decision about how much it's going to cost you yeah so so i think that's and of course we and we we kind of almost disguised it as a model for
42:05 us to cost a little bit we think oh what cost but the cost of if you're using it from terms of subscription when your subscription runs out the only way you can get any more in that time frame is to pay more money now if if you're doing that you're using it within a business that could be critical it could you could hit a point again i've gone i've got and now i've got to actually
42:26 pay some money so using it effectively is um is something important and you also can use it through through the api so i certain tools i have they make a connection to claude without using the claude client and there i'll pay but i'll pay with credits so the advantage there is i can i can throttle that so i can know how much i'm spending on doing that but i've got to remember i combine
42:52 that cost with the cost that i pay to use these things on a subscription basis as well and and the cost of the connections i'm making the other so there is a also a potential hidden cost if you're not careful of all the tools that we're using and actually i would imagine potentially in a smaller business perspective that are if we're not careful because one thing people want
43:16 to sell you to is ai can do more and that's going to save you save you money but if you're not careful your actual technology footprint is bigger and more costly than than it's worth before and as you've said which is a very important point may not be as reliable or you may have all your eggs in one basket what do we say i mean we we it's a it's a it's a it's a it's a it's a it's a it's a
43:35 fundamental thing in business isn't it we talk about so many areas of that that you don't want to have you know you don't want to just be relying on one big client you don't want to be reliant upon one piece technology or one one supplier so we don't want all our eggs in one basket yeah exactly that that's why i really only use it like for very specific things punctual we also
43:59 have i mean i don't use that i have my paid claude and my paid chat gpt but my husband as a local thing where i can load all these models and that way doesn't have to pay the apis i think that's what you're referring to but the only agent that i have on cloud co-work but it's one like it doesn't matter if it's making mistakes and what it does is that every morning at seven it's going
44:26 to all my social profiles and it looks at how many followers i have and it looks like the last three posts and it looks at the likes the comments and it looks at the comments and it looks at the and things like that and it updates a spreadsheet on a google drive so basically it keeps track of the growth of my followers on social and that's the only one and if it makes any mistakes it's
44:50 not the end of the world for me yeah absolutely and i i think there's definitely one of the things you talked about is it's important to understand the results you want to get so why are you doing it and what do you want to achieve from it and i think the other thing is education what often we've talked about are things that we have come across and found out about
45:11 which has helped us to understand how better to do these things so to educate ourselves on what can ai do what are some of the costs on it how do we use it in a more effective way because as we've said just plugging it into everything and yeah oh great i can plug it into everything and then it can tell me whatever i want it however it's still a danger because you go right i've just
45:34 asked ai to tell me something and it's still a danger because it's still a danger because it's something about my organisation and it could still be wrong so if i don't know or i don't go and just just check making sure i am actually checking the outputs of some of these things yeah i could be a problem yeah i mean remember when you log in into cloud it says below
45:53 cloud is ai and it can make mistakes check your data it just says it like in black and white like it's just there so i think again coming back to for me ai as an amplifier or an accelerator or to make you more consistent but on something that you already do like if you are doing nothing and you expect ai to suddenly do all of it for you
46:18 you know those people that say i fired my team and now i have 50 agents oh my gosh i don't even want to think about the quality of what's going on in there because uh i don't see that working low low low quality i think and also there's there's the cost i mean i i use that i use i think it cost me was it i think it's
46:41 about 90 90 pound a month or something i spend on my cloud because i have the max subscription but i use it i use it quite a lot but even then i can i can max it out i can max it out quite easily and that's maybe just doing some coding and i've got maybe multiple agents but if i'm running 50 agents then i'm going to need i'm going to have to pay more than that now that might be worth it
47:05 if it's actually delivering for your business but i think yeah there's something else that uh in fact in the the uh one of the other podcasts so there's there's uh lots of good stuff in there one of them is we talk about that quality over quantity so we're now in a world of quantity is easy right as we said we can produce all this stuff everybody can do it now so it doesn't matter
47:26 it's the quality yeah the quality is it authentic yeah and how does it how does it really help us connect together as as human beings i think is is is why how i would summarise and as we kind of come to a close i'd love to just if you could tell the tell the audience how can they get in contact with you how can
47:46 they understand find out how to get better results from their things yeah sure well you can find me at themembershiplab.com and actually i'm going to have a special tool for you guys so if you go to themembershiplab.com slash better with humans i think by the time this releases i'll have created the tool then um i'll have something
48:13 for you and then if you want to reach out there will be a link for you to connect amazing thank you nathalie and yeah also for those watching if you've liked what you've heard then of course get in touch with nathalie you can also get in get in touch with me all the details will be in the show notes whether you're listening or watching just check the show notes and all
48:33 the details will be in there including the links that have been provided to you and i'll see you mentioned please like follow subscribe join in in the comments let us know what the topics you might want to hear about what are we not talking about because we want to know what are the results you want to get from from hearing these uh these episodes on this podcast and yeah nathalie thanks
48:51 you once again it's been a great conversation thank you so much for having me shaun that was fun
People on this episode
Podcasts we love
Check out these other fine podcasts recommended by us, not an algorithm.