Speaker 1so maggie this week we're going to be chatting about ai now this is something that i was very late to using and i to be honest i still haven't embraced or explored full capacity um but i know that you've been dabbling with it a lot a lot longer than i have and i've done a lot more exploring um for for lots of different reasons using it so i guess a good place to start would be kind of how have you used it what have you sort of found how have you found that sort of useful and what are the things that you found not so useful and and kind of how's your use changed over the time you've
Speaker 2been playing with it well yeah so well for me personally um i've discovered that sometimes it's not so useful it's not um it's not perfect and i think do you know what this takes me back to when i was teaching and this is gonna show you how old i am but we um we had to teach the kids how to use wikipedia because immediately when it came out all the kids were like wow this has all the facts in the world and you'd be like no no it doesn't um people can put anything they like and it might not stay up there very long because it is open source by like collecting individuals um information it is managed by various just different people the public can put whatever they want into it so it's not always going to be factual and you will need to check your sources elsewhere now in much the same way i think that is ai because ai only can take when you are looking for information what is already out there and what is already out there and what is already out there on the internet that we've put into it so it can be biased it can throw out things that aren't necessarily true and i think what you have to be discerning about is how you're using it you can't just go oh i want to know xyz about and ask ai i think like when you are looking for information through it you have to have an understanding yourself what you're looking for to know as to whether it's true or not so i'd say like i don't use it for research purposes because well some of it's better than others um anyway but it does this thing called um hallucinations right if it doesn't know something it just makes it up and it'll even give you a link to what it's made up um and and then you go on the link and it doesn't give you anything that you ask for so a really good example of that is when i was doing my master's um you know i'd go on like google scholar and things like that and i'd be like oh i'll just like ask ai if there's anything else related to this topic that i haven't you know manually been able to find because it's obviously got an algorithm it can search and you go yes here's three different papers and a meta analysis on this particular topic you're asking for and it's like oh i don't know what to do with it and you click on each link it'd take you to a paper it takes you to a paper about something to do with you know like nuclear science and i'm studying psychology and i'm like that's that's not what you said it was and then it would say no you're right it's not i don't know what you want from me why are you giving me information that is factually incorrect so so it does do that a lot and i think some of them do that more than others um i have started using uh claude uh by anthropic a lot more simply because it doesn't do that it will actually say i can't find that information i'm like thank you um it's also less sycophantic that's the other thing i found with some ai toys maggie that is such an amazing question no one has ever asked a question like that before in the history of questions how are you so good at asking questions i'm like oh my god just stop this is really annoying um claude doesn't do that so much um and it's also quite ethical in terms of there's there's some ai tools i don't use anymore because i've learned about some nefarious connections that they might have should we say or just say it that way uh and and so i've chosen not to use them um from a kind of a moral standpoint um but um anthropic um from what i've read about it is is uh has kind of higher levels of morals um because it's a it's a it's a kind of a moral standpoint um because it's a kind of also i feel like could be quite dangerous um you know it's not it's not necessarily just a good tool that can help us out we do need to be careful and we do need to be thinking about what are we putting into it what is it throwing out to us and how are we using that information so um i do like companies that take that into consideration and what does it mean for us in terms of like you know the future of work and people's jobs and again i like to use a company that takes that into consideration and water water is a big part of the future of work and of it um less so in europe because we're creating closed chain data centers which can reuse water but in places in america it's really super hot for example they do use a lot of water and they are starting to use water from um you know towns and places nearby um and that will potentially become a very big problem it's already becoming a problem but it will become a very big problem so for me there's there's there's different levels that i feed into which ai tool i'm using so i don't use it to create cartoons of cats i get it like you know turn my cat into a person it's great but i am aware of the water that's using and i'd rather not personally um so i am using it kind of very discriminantly um i tend to do my own research so i don't necessarily trust it to do it for me um i can ask it whether what i found makes sense and i think that's a helpful tool it's like a critical friend does this make sense could it sound better and i can ask it and it will say to me actually yeah that's quite good i might change this a little bit for grammatical effectiveness and i'll be like oh yeah that's better thank you um but i will always always review myself what it has given me and i would never use it therefore without knowing the answers myself in the first place if that makes sense um and i also use the same tool over and over again because it learns who i am what my voice is so when i want to create for example an instagram post i'll say this is what i want to say here are the notes about it so it's not creating its own version um and i can you write it in my voice and because i'm using a tool that has learned what my voice sounds like it can do that and it sounds less um you know sort of samey as some of them do i think some sometimes you can tell that ai has been used to an instagram or a facebook post um so yeah i i kind of try and get it to learn who i am um and then it's useful as you jump between tools i feel it's yet less useful although i do feel like in just what he talks about that i jump between everything so would you like to hone in on one of those elements how would you use it there's lots of ways and not at
Speaker 1all uh it's great because you i mean you've covered a lot of things there that i think we just don't think about um because it's presented to us at as this tool that's either gonna um take over the world and and remove all jobs or it's the answer to everything in terms of you know reducing our own workload and helping with it and it is worth sort of thinking about the there's a bigger picture to it in terms of as you say like the environmental concerns political concerns and stuff like that and it's not stuff that most of us think about because you get stuck on that initial dilemma of how am i going to use it do i trust it where am i going to use it and i think that's um i mean certainly i've the times i have used it it's very much a case of um a lack of a lack of trust i mean i have come to the conclusion over the years that i don't trust much anyway um so it's no surprise there's no difference on that but certainly that that need to to read it to check it to make sure it's saying what i want it to say and and things like that i think where i found it useful has been sometimes when you just you know when you've got that blank piece of paper and you're like i don't know where to start with this or i need ideas then having something that you can ask just jump start that first draft almost so that it's always easier when you've got something in front of you to then edit and change it and when you've got a blank sheet of paper i find so yeah when you're feeling a bit stuck though that normally when that pressure is on as well so if you've got ages i can my creativity's fine as soon as you've got that like deadline have you got to get something done and it's got to be done in this period of time and you can't really spend much longer on it my mind goes blank you know i've got no idea where to start or what to write about or what how to approach this thing and that's where i found it useful to be able to like either say give me a first draft based on this bit of information or give me some ideas for this thing give me a starting point Because then you can look at it and you can go, there's no way I'm doing that, but it normally sparks something else. Or you've got a list that you go through and go, no, no, no. Yes. Oh, that one's something I could definitely expand on. And then it's, you're off basically. And so that's sort of where I've, where I've used it a little bit. Um, and I know a few years back, you actually used it to, to do a magazine article for us. And it was really interesting to see what it, it came up with in terms of, you know, how, um, how it was really good at bigging itself up because you wrote the article about how great AI was. And then you get to the end and it's like, yep, this was written by AI. So interesting. And I think that's the kind of stuff that has been useful a bit. Like you've, you've said yourself where you kind of like, you've got your answers and then you see what else it suggests. Um, where you sort of see what it would do and then go, actually, yeah, yeah, I agree with that. But just because you might agree with one answer, doesn't mean say that you then trust it for everything. And I think that's, again, maybe a little bit, um, I was going to say different, but it's not, it's something we should be aware of. I think in everything, it's always been an issue that you, you find somebody that's great on a particular topic and then the, the sort of like halo effect crosses over and you expect them to be right on everything. Whereas in reality, they've got their topic and their area of expertise and what they might say on something else isn't, isn't necessarily accurate. So I think treating AI the same as we would, any other source really is, is a good idea. You say we, we might've found somebody that's fabulous on their nutrition knowledge and it's accurate and it's always scientifically backed. And then they share something to a training and we go, hang on a second, where did you get that from? Why are you thinking that? Or vice versa, you've got someone that's amazing on training. And then all of a sudden they start waxing lyrical about some obscure or extreme dieting method that sort of about it being so like, the only answer to everything, you know, hang on a second, that's not, that's not right. So if we treat all of our sources on that way, I guess it's never a bad thing to have that, that stuff and to have those sort of mixture of things that we go to.
Speaker 2And I think interestingly that when you are, when you are someone who is already kind of applying a critical lens onto things, you can use AI to help you do that too. So for example, you can kind of like, particularly when I'm, doing my master's, it was about challenging and really trying to get to a point where I could challenge my own thoughts on things. So, you know, you're writing something, you're going to naturally lean towards research that fits in with your existing bias. And then you kind of have to go, have I got an existing bias? Am I aware of my existing bias? Like biases in essence are not something you're necessarily always aware of. So I could go to it and say, this is an argument I have formed. I think these are the counterpoints. I missed anything. And it can sometimes be that kind of critical, like help on actually, yes, like this is the opposite argument for it. And I go, I hadn't thought of that argument. And then I could go and do some research on it. So it was kind of useful in that realm. Obviously I was doing a university degree. You're not allowed to use it to write things for you, but I also wouldn't want to. Something, again, I very much understood as a teacher, and this was for AI. I did not teach during AI, but I taught you in Google and you would read something and it did not sound like your kid. And you'd go, huh. And you type it into Google and the whole thing would come up in front of you and go, huh, that's why it doesn't sound like them. Because people have a voice and people have ideas and opinions that they convey in unique ways to each other. And I do think that sometimes, I said before, like sometimes you can tell AI has written something, because it sounds like AI. It does not sound like a unique individual. And especially if you're used to reading things from that individual, you can read it and go, oh, that doesn't, that doesn't sound like them. It's also a bit over wordy at times. Like sometimes it actually gives you too much and you're like, all right, calm down. So yeah, I like, I think I wouldn't ever want to use it. I'm part of this. Maybe it's me. I wouldn't ever want to use it simply, for writing things entirely for me. But part of that is because I wouldn't want to lose my creativity in that realm, if that makes sense. But then also if I had to like, you know, if my job was putting out as much content as possible, I would probably use it more than I do because it would help me speed up that process. But then I also would need my human element to stay in it too, because otherwise it's just giving me the equivalent of that. I don't know if that makes sense. I purposely got it to write for the magazine. It was like sycophantic around AI. AI is amazing. It's going to replace this, this and this, but also it's going to help personal trainers be even better. And I was like, okay, not sure I believe all of this. So I think that, I think with AI, it is not a replacement. And I don't, I don't actually like the idea when people talk about, oh, it could replace this, it could replace that. I'm like, sure, maybe data entry, it might be useful. But also you would still have to check it because it gets things wrong all the time. And part of that is because this isn't like some amazing technology that knows stuff that we don't know. It only knows what we know. It is pulling on human error and bias and assumption as much as it is pulling on facts that we also have in the sphere of the internet. So it only knows what we know. It can therefore be prejudice in ways we might not know. We might not notice because we put it in there in the first place. So, so I do think that's always worth bearing in mind that you're not going to get something perfect from it. It will always require, I think anyway, human intervention into it because otherwise you just, yeah, you're creating something that's dull and mundane and potentially stereotypical and biased. Like, yeah. But then it is useful for checking things. I use it to proofread stuff because it's quick and easy. I might see things I've missed a fair bit. Can you proofread this? I know it got a bit wordy in the middle, or I'm trying to say this and I've spent the last five minutes on it. So I put it in and say, I'm trying to say this, but I want to say it in five words, not 10 or 20 that I have at the moment. How can I do that? And it'll go, how about this? Oh, there we are. That was quicker than it would have been if I'd sat here for another hour. So it can be useful for stuff like that, I think.
Speaker 1Yeah, I think it's interesting because the temptation is to use it, like, as you say, maybe it's for social media or putting out content. But one of the things that we've always sort of say when it comes to looking at is letting your personality through and showing a bit about yourself and how it's, you know, for most personal trainers, it's less about content for the sake of content. It's less about, you know, having to have an educational moral of this story for every post you put out and far more about showing who you are, who you train, how they train, and that kind of stuff, which can be done quickly and easily through sort of snapshots of your clients, more than through some sort of post about the benefits of eating protein. And you can include bits of that, but it needs to be in a way that's actually interesting to the people. You're trying to attract and that's where you kind of have this sort of extra layer really of choosing where you put this stuff in, choosing where it fits in your life and business and if it fits, and if so, is it really adding to what you're doing in some way rather than kind of leading you in a direction you don't want to go? Because it's really easy if you were to use it for your post to just kind of get to that point where you're like, yeah, that seems okay, that seems okay, that seems okay. And all of a sudden, you've created a completely different voice and you're putting out a completely different kind of idea of things. And I think it is that kind of trying to find that point where it is helping you. So it is kind of useful to you in some ways, there's no point in including it anywhere. But it's not taking away the stuff you like doing. It's not removing that element. I think it's really important to have that element of thought process somewhere because I think as humans, we all need that. And yes, it's nice to have some things that are easy, but it's also there needs to be that element of sort of thought that goes over the top of it really and sort of for our own benefit, for our own kind of personal reward. There'd be nothing at all rewarding. I think about having AI do everything in the job. It kind of creates your programs, does your social media, has the communication. It's all about having AI do everything in the job. It's all about creating connections with your clients. And you're kind of turning yourself into an admin that's just taking things from one source and sending it somewhere else or hitting a post button or whatever. And I think that I can't think of anything much worse than that in terms of stuff that's going to make me interested in the job and interested in what I'm doing. So I do think there's definitely a place for kind of choosing where it where it it's choosing what you want it to do and what you want to keep doing. and and also to not be completely afraid of it because i think that's the other element of it you know some people are kind of like well it's going to replace everything and you don't need a personal trainer because everyone can just get a training plan from whichever ai tool they like and all of that but i don't it's never going to take away our need as humans to have human interaction i think it it's there there's always been alternatives for people that couldn't afford a personal trainer or didn't want a personal trainer um most of us actually i probably should caveat with most of us over a certain age um have got got programs from magazines in the past and i appreciate i'm not sure where magazines sit nowadays in terms of people it's really way down further down the list than it used to be back in i'm going to say back in my day um and and then look into pension shortly after this call i think but um but yeah that's there's always been those kind of things and yes it's easier um and and cheaper in terms of most people have access to the internet and you don't have to pay the cost of it but it's it doesn't mean to say that we don't sometimes want a real life human there to say something we we have this like innate trust more i think in having somebody that we can say yeah but what about this or this feels weird here and we don't need to worry so much about it we don't need to worry so much about it we don't need to worry so much about the language we're using because somebody has stood there with us and we can go this bit here and point to it in a way they're trying to explain where this thing hurts and what i'm doing can be it can be more of a challenge when you're having to put it into words versus you know point at things and go when i do this that there hurts which doesn't work at all you can't see it so it there is always a place for us as human beings to be able to help people you know deliver these sessions and to have people have a fun experience um out of something that is far less uh fun um and engaging when you're doing it yourself based on something that um a computer's told you to do i feel yeah it's i mean
Speaker 2it's interesting isn't it because ai's been around a long time now i think if it was going to replace personal trainers it would have done it by now um and it's because of the fact that it's been around for a long time now and it's because of the fact that it's been around for a long time now and it's because of the fact that it's been exactly what you said is that essentially for a very long time now youtube magazines like wherever you wanted to look you could have found something if you wanted to but i don't think that's the reason that people use a personal trainer they don't use us to write a program for us they use us because we have they know an expertise and a knowledge that keeps them safe but also makes them feel like they are making progress it motivates them like there's so many different reasons and a lot of those are actually psychological rather than physical so yes they could improve their physical health by and some people will but equally that doesn't cover the majority what a personal trainer does and i think that's the same for a lot of jobs that are out there but essentially yes ai could be a useful tool but i think that we need to use it discerningly knowing that what it doesn't provide is a human element and that what actually a lot of jobs need is a human element and that's included even in you know working with your colleagues not necessarily customer facing things but when people are actually you know so many people lost that feeling of connection and belonging in the jobs that they did when covid happened because they came out of an office and a lot of jobs have kind of benefited almost from being at home but people still connect at least one day a week in the office or online over zoom and there is still a human connection that is taking place that i don't think can be replaced by ai and actually there's questions around whether people should be having these psychological type of conversations with ai um you know there is um there's examples out there of people that have actually made their mental health worse because they're asking ai for support in things and that's because it can't discern properly what a human being can when they speak to you um it can only take what you put in and a return to that is very often very sycophantic it is designed to big you up essentially and for you to be right rather than to give you facts that maybe make you question and think about your behavior in a certain situation so i think that for our industry which is so much about human connection as much as it is about physical activity that it is something that is going to probably remain with us and be useful if we use it in the right way but also recognize that that human that you are is why people come to you so don't overuse it um planning is great with it you know i've got a bunch of ideas and i don't know how to pull them out can you help me do it but i always think if you're the input first rather than it being the blank sheet you know even if i don't know where i want to go with an idea i put what i do know about it in there first these are my notes these are my opinions this is the research i've done here with links to where i don't know where to start can you structure something for me and it'll do that and then i have something to go back and work from i enjoy doing that with it and i find that very useful and i know that whatever comes out the other end will still be something that is created by me but essentially if we let it have too much free reign on that we end up with you know the teacher who sits there and goes this doesn't sound like my kid all right google and this time it's all right ai yeah i mean i've
Speaker 1one place i started using it is actually for my episode summaries of these very conversations we have because i found that you know i sit there and i two weeks after the fact going what was it we spoke about and how do i summarize that and i just need a really short paragraph that gives me this whereas i can take the you know i can get zoom that we use to record it to give me a summary of the conversation i can chuck that into ai and say can you just write me an episode summary and i get this nice paragraph and yeah i read it through and sometimes i'm like well we're not saying that and i'm going to change that i'm taking that sentence out but i've got this nice little summary of of our conversation that has saved me a lot of a lot of work because it really feels like one of those jobs that takes a disproportionate amount of time to create when you like i need about i know a handful of sentences three would do and then summarize this conversation why is this taking me so long and why is my brain not working whereas i can just use it for that and i think that's that's the sort of stuff where you know we've provided the content because we've had a conversation up to an hour about the topic it's all there it's not changing that it's just summarizing it in a way that yeah my brain can't take in that hour's worth of information to come up with three good sentences um ai can and so it does does give you that option and that that inspiration and and that kind of stuff and i think that's where potentially if i look at something like programming i wouldn't be saying create a program but i might be kind of when i'm having those bits where i know my brain's just not functioning the way it used to and i'm like we've said before um i had lists of exercises that might be kind of i don't know squat based or lunge based and things like that well this would be the equivalent i think nowadays um again a difference between in my day um i had a piece of paper with a list or a notebook with lists of exercises and i was like oh my god i'm not going to do this exercises but being able to chuck in and go like i'm looking for a different form of lunch give me you know some options because right now my brain's just disappeared on holiday for a bit and it's not coming up with anything useful that kind of stuff when you're you're stuck there going okay i need to do this i mentioned before recently when i was saying i had you know i did classes that were an hour of abs and my first thought then was what am i going to do and i did i spent a lot of time on google at that point going give me some exercises on abs or fit balls or whatever class it was i was going to do it and and then going through going well yeah i'm happy to coach that or there's no way i'm getting a class to do that or you know whatever it was but i think that's the sort of stuff where i would be turning to ai now and going can you just you know give me that spark of of almost brain power um it's like an extension that can just come in come online when the real one goes offline and just give me those things that you can then go through going well no and yes and maybe and i hadn't thought about that and let me explore this one a little bit more and where that might go so um i'm like anything there's also that element of making sure it's not a rabbit hole that you disappear down um never to come back out again
Speaker 2yeah exactly exactly i'll tell you something else i use it for actually that i've just thought of as well is um sometimes it can create a tool for you that you don't have the knowledge to do um but you know what it is you need it to do um so for example we were talking about the fact that we were talking about the we were trying to work out how we could price workshops um that we might offer to like corporates and things like that and um what we wanted was a tool that a calculator that would take into account what our overheads are what the cost of facilitator is um how much margin we wanted to take from that and then we could put it in and then depending on like you know whether we're doing it for a small community business or we're doing for a large corporate business it would kick out a sort of a price that we could use the money to do it and then we could use as a a guidance more than anything we can obviously apply our own human analysis to that as well um but so i wanted essentially a spreadsheet that i could put things into that gave me that um and in the past i've had to use my brother who i think gets a little bit fed up with me asking him to make spreadsheets so what was quite useful was to be able to go i need to do this this and this um i was again i was using claude the other thing i like about it is it asks you questions so i'll say like what's the priority of this is it this this or this and you pick it and then it's like i just need to check one more question with you and then you pick it so it kind of asks you questions back again which quite like um and it created this spreadsheet for us that we can go to and we can plug information into and it kicks out this this price and then we can go and compare that against what other companies are offering um and see whether that feels right in the marketplace so there's still a human analysis that has to be done but i couldn't have made that spreadsheet it would have taken me ages i would have to do a bunch of googling about how do i do this how do i do that how do i do this and i know of other um pts and and companies which are using a to create tools perhaps using them to create chatbots so when you ask a conversation they're having a conversation back with you and as much as i find that annoying in some senses it's also useful when i come across it in some companies i don't need to speak to a human because the chatbot can solve it for me and as long as it doesn't think that it needs to keep talking to me when it can't solve it uh give me a human sometimes the chatbot gets a little bit of abuse um but um i know that that can be a useful tool as well and why doing things so i think in some respects as well it can be useful for that if you need a tool to do something for you and you don't know you don't have the skill to create that tool but you know what the outcome of it looks like and you know how you want it to work then again you can assess what you've got on the output is actually correct but you can't necessarily do the bit in the middle it's quite useful for that yeah i
Speaker 1think that's really really useful i have to when you start that i'm a bit like oh spreadsheet i'm a bit like oh spreadsheet i'm a bit like oh spreadsheet i'm i could do that um and i guess that's the thing if it's something that excites you don't get something else to do it yeah if it's something where you're sat there going i've got no idea then it can really help that's that's probably yeah i think that's that's the summary it can help in lots of different ways but really as long as it's kind of adding to your life in some way rather than taking it away and it's not removing from your your voice and and you being you and your service being your service and and all of that then it can be a helpful tool but it's um it's just making sure that it it kind of like yeah stays in the right place for you which is going to be different and we're all going to have a different kind of line on what we think is is useful and acceptable to us partly because we're all different in that very example you know give me a bit of a spreadsheet to try and get my head around and and i'm probably in sort of like a happy pace as it were if you gave me you know scientific research to do i wouldn't be so happy whereas you would be in your in your happy place so there's different things that we might we might turn to um different you know wanting support and sources for and and that's what what's right and that's what makes us all individual and that's why you know there will always be a place for for humans to have have roles and to do these things and to support each other because actually i'd rather talk to a human if i can and and if i can't then yeah there's a place for it and it can be useful and it can tie i mean did that help their skull uh since the example that you mentioned is is great isn't it from the point of view of somebody that answers questions having a tool that can go the equivalent of have you tried to turn it off and on again is great where it falls apart is when it's like yeah i've tried that and i've tried this and i tried this and i tried this and i still don't know the answer so i need a proper person to be able to go and take all this information that i can provide you because i'm not completely useless and i've got this far and that's the part and that's when we all get frustrated with it i think is when when there is when you are stuck in a bit of a an ai kind of uh loop and you just need somebody who can listen to what you've got to say go away look into things and go oh yeah you're right thanks for bringing that to me my attention there is a problem yeah
Speaker 2that's it exactly yeah i have got stuck in a loop with ai chatbots and sometimes i'm like oh give me a human and the human comes on and goes i'm very sorry about our chatbot it does get stuck sometimes this is what you needed yes still
Speaker 1humans in this absolutely and that's it i think as long as we stay aware of where the limitations are and and open to the idea that we do need to still sort of keep that critical thinking level involved even on simple things um to rather than just accepting everything as as read but it's the same as we would if it was a another human or um another source there still needs to be something that we kind of run it through our own internal sense check our own sort of gut and go yeah i'm putting stuff out that i'm okay with um and uh and that's the key yeah
PodcastTranscript AI