🎙️ Episode 37: The AI Devolution

Episode 37 – The AI Devolution: Broetry, Bots and the Death of Voice

Everyone’s calling it an AI revolution. We’re asking whether it’s becoming a devolution.

In this episode of Awesomely Off-Topic, we take a proper look at what’s happening when creativity gets outsourced a little too eagerly. Not from a place of panic. Not from a place of fear. But from lived experience.

We use AI. We value AI. But we’re also noticing something uncomfortable – when you lean on it too heavily, your own thinking, confidence and creative muscle can start to atrophy.

We talk about:

• Why so much content now sounds eerily the same
• The overuse of rhetorical “broetry” patterns
• Why AI is only ever as good as what you feed it
• The difference between complementing creative professionals and bypassing them
• Art scraping, genre styles and where the ethical line sits
• Why the least interesting thing ChatGPT can do is write your LinkedIn posts
• And what the real power of these tools actually is

This isn’t anti-AI.

It’s pro-standards. Pro-voice. Pro-thinking for yourself. Pro not using really obvious triads, like this one! (Well, it was before we added a fourth :D ).

If you’ve ever wondered whether you’re using AI – or whether it’s starting to use you – this one’s for you.

🎧 Listen now and join the conversation.

Transcript

Speaker 1 0:03

You're listening to Awesomely Off Topic, the podcast where we talk books, brand, business, and everything else we're not supposed to say. Out loud. We're Taz and after ex-gernos, now coaches, creators and chaos navigators. Let's go! Welcome to another episode of Awesomely Off Topic. Today we are talking about the AI devolution. Broetry, bots, and the death of voice. Sounds really dramatic, doesn't it?

Speaker 2 0:34

It does do do. Feels like you need a little soundtrack to go with it something.

Speaker 1 0:39

It does, it does. Um so Ash and I were talking before we came on air about this. We've been talking a lot about AI. We love AI, and I proper go into my geek zone with AI. But while everyone else is calling this this massive AI revolution, we want to talk about the AI devolution. So what's that about then, Taz? Well, it it's not that the tech's getting worse per se, it's what happens when we are letting our own thinking, our own creativity, yeah, and actually our own confidence slide. And I've noticed this in myself. We were talking about this, weren't we, before we recorded this today? Yeah, yeah, yeah. If I lean on AI too much, my own inspiration starts to f shrink. Yeah. My confidence, my creativity start to kind of atrophy, and I start to lose the belief in myself as a writer, which is bonkers when you think that I was a journalist from the age of 16 and then went on to you know, edit and be be editorial director over global titles for a global outfit. I've written seven books, contributed to goodness knows how many others.

Speaker 2 1:56

So, what does that say though? Isn't that I mean if we're if you're feeling like that, because I must confess I feel like that too. I was saying to you the other day that because we can get into the habit of using AI, it never replaces us, and we need to remember that, and certainly not in our creative um areas when we're doing our marketing, it should never replace us, it can work alongside us, and we'll be talking about some of the ways it can help. But that feeling of of losing confidence because you've got out of the habit of doing something, and we're both professionals that have been doing writing, creating content for you since you were 16, for me since well, similar age, really, you know, going in from school.

Speaker 1 2:33

Plus, we're not even thinking about the kind of writing and creative projects where we're getting on as on with as as kids and as young adults before we ventured into the city. All our lives basically. Yeah, the thing is, if I'm noticing that it's having an impact on the way I behave and approach everything from content creation to allowing my imagination to to open up, and I don't rely on AI half as heavily as a lot of the of other people, or judging by a lot of the content I'm seeing out there, yeah, then this has got to be worth talking about. Because again, if it's having this impact on me, and I'm not using it to that extent, and you're absolutely right, it's it's about and I've said this so often, I've said this until I'm blue in the face, and I swear I'll I'll continue saying it forevermore. AI is an assistant and not a replacement. I think that's where we're going wrong. What about you, Ash?

Speaker 2 3:36

Well, I've got to think about firstly when I think about content, one of the things that I see again and again is this kind of this bro broetry the B word, the B-word broetry that I see in every other post, and it's like, and people that I I've known for quite a lot of years that I've got to know their style of writing and suddenly disappears, and I'm like, what happened? Did you have a personality transplant overnight?

Speaker 1 4:01

And you know what? We had an i an influx of broetry style writing on LinkedIn quite a few years back. Because some probably American content coach was out there telling people to do it and it took off, and it was just I mean it was really hideous.

Speaker 2 4:14

Yeah, it was really good.

Speaker 1 4:16

I thought we'd got rid of it, and now it's snuck back in again. AI is bringing it back, and it's not just broetry, it's it's kind of rhetorical overuse as well. Yeah, so what I mean by that, and and that's not that rhetoric is bad, but we see these patterns in in content that's been developed by AI. So false contrasts everywhere, yeah. Um lots of emotional presuppositions, i.e., you're not broken. Hold on, nobody said it was broken. It presupposes that you thought you were broken, and lots of other examples. Triads in every paragraph, there's three series.

Speaker 2 4:56

The rule of three, which actually, when we talk about that, we will talk about that a bit later, won't we? Of having that where we we training as journalists, we were shown the power of that. Yeah, so that's been around for a long time.

Speaker 1 5:07

But when it's in every post and in virtually every paragraph, yeah, um it becomes really confidence, clarity, creativity with a little full stop between each one.

Speaker 2 5:19

Yeah.

Speaker 1 5:20

And those awful things where you see it's there's there's a really, really short paragraph, like half a line, and then it does a new paragraph, but it doesn't leave a line gap in between. Yeah. It's like, what are you doing? It's trying to create this kind of I don't know, rhythmic cadence, but it's false. Yeah. Um what are plastic? Plastic. Plastic, manufactured vulnerabilities. Plastic flows. Plastic flows. Didn't they have some of those on only fools and horses in the back of a car? Yeah, manufactured vulnerability. Oh dear, it makes your skin creep. I have been talking about the power of using vulnerability in a way that is is ethical and makes a real positive impact and creates real change. So now to see this awful manufactured in air quotes 'vulnerability' popping up in all these AI posts and yeah, the stacked dramatic lines. But again, we're not anti-rhetoric, you know, we've been journalists and editors. Rhetoric can be really powerful.

Speaker 2 6:29

The technique is powerful, but the whole pattern of it is lazy, isn't it? It is lazy AF. Come up with something a bit different this time, think about it, rather than just kind of shoving it all into AI and letting letting it write it write it for you when you can all write, you can all write, and it's also about knowing how to construct a proper prompt.

Speaker 1 6:50

It's about being able to be the architect of those prompts in such a way that they that the AI delivers what it is that you're looking for. But if you're just writing away and just typing in or speaking in a load of old guff, if you don't understand how to use rhetorical devices, she says sounding like a proper old journo, AI defaults to them. Yeah. And when it defaults to them, it always overuses them. Always, always, always, always. And there's there's it's not just in posts, is it? I just for a while now, for as long as AI has been around in kind of the the mass uh available to the masses, yeah, you've started to see lazy AI creeping more and more into some of the manuscripts that you've been editing.

Speaker 2 7:41

I can s I can see why it can happen, because we all know if anybody listening has ever written a book or is writing a book right now, it's a hard slog.

Speaker 1 7:50

And it is I've done it with some of mine, yeah.

Speaker 2 7:52

It can be a long slog too, but so it can be tempting to run it through AI. There's nothing wrong with that per se, but it's when you've got your own unique way of writing, your own unique style, and then to put it in there and just take it word for word from the AI perspective, it immediately jars. If you've got half of your book or two-thirds of your book written in your unique style, and say your style is a flowing style or something like that really flows, it's longer sentences, and then suddenly in the last couple of chapters of the book, it's got all the things that we've talked about here, all the rhetoric that we were talking about, all those things, suddenly it's like there's a jarring, it's like, what happened? What happened?

Speaker 1 8:33

I'm guessing though, Ash, that your authors that you're working with, yeah. And actually it goes further back than that, I remember.

Speaker 2 8:43

Oh, I'm not talking about people that I'm working with recently. I mean, I'm going back a fair old while now.

Speaker 1 8:49

Before that, if I think about when you were still a curator for TED events, yeah, didn't it become obvious that some of the speakers there were turning up with AI scripts as well? Yeah, exactly that. How can you tell though? Because presumably if you're if you're working with someone as as their editor, how can you tell that they've used AI in that manuscript? What are the gifts for you? The Alexa, stop. See, we we just keep this really real. We do.

Speaker 2 9:19

We're just gonna get everybody will want to know what A L E X A was gonna ask me then or remind me of. Oh no. And you'll just what I've here's an idea. We're just going off topic, but this is so us. Have a think about it, guys. Let's have some really funny responses to that. What do you think AI was gonna prompt me on?

Speaker 1 9:36

You mean A-L-E-X-A? So I mean A-L-E-X-A. Yeah, it is A-A. It's still an artificial intelligence.

Speaker 2 9:41

If it was, what would it be saying to me?

Speaker 1 9:44

Hmm, interesting. Sorry. Anyway, how can you tell if if you're working with someone who's writing a book, presumably they're not gonna fess up and say, Oh, I ran this one through Chat GPT or Claude or Gemini or whatever. Wha how how can you tell that it's happened? For want of a better word, it's voice.

Speaker 2 10:02

It's the voice.

Speaker 1 10:04

Um Can you explain that too?

Speaker 2 10:06

So that's so the style of the sc of the words, it's the way that the expressions, things, you know, every writer will have certain expressions that will crop up again and again in their manuscript and they will use it, and it'll be, you know, that will be their language pattern, their way of explaining things.

Speaker 1 10:22

So is it that it suddenly changes tone?

Speaker 2 10:25

Yeah, that too. And the pace as well. Very often it's the pace. So for instance, if you've got somebody that's got a more a slower kind of style of writing, and suddenly you go into all of these really stacked sentences and things, it it will feel very different when you read it. And one of the things I always talk about, everybody's heard me say it a thousand times, I'll say it again. But my great way for me, anyway, of looking at um a manuscript is to read it out loud because you get the pacing of it, you get the feel for it, you get the feel for the writer's style, all of that. So it's really, really powerful to be able to do that. And when I do that, what I notice is that kind of that whole thing, there's just a flatness to it. Um, and it's almost like we're smoothing out all the imperfections, i.e., by that I mean that kind of we end up, we replace it with kind of almost like a monotonous tone.

Speaker 1 11:15

So the t all the textures disappearing. Yeah, you know, smoothing it all.

Speaker 2 11:19

I'm seeing it almost like a I don't know, like a I don't mean a graphic equaliser, but one of those things, you know, it's just it's peaks and troughs. The last thing you want there is a flat line. Exactly that. Yeah. I want to see that that up and down. I want to see that. Yeah. And I'm not seeing that with it. So and it's almost like the whole kind of texture, I don't know, disappears. It's just it's really difficult to explain, but it's just it's a sense, and when you get to read as many manuscripts as I do, um it becomes it really jars. It just feels really, really outpill.

Speaker 1 11:51

So if we can tell, then the readers can tell.

Speaker 2 11:53

Well, this is one of the things that we've talked about as well, isn't it? Because obviously we've worked with um words and worked in commu you know in communications for a long, long time now. And I often say to you, is that just us because we've worked in it for a long, long time, and so we see that. So again, playing devil's advocate, does it really m matter? I mean I think it hugely matters and that you can spot it. I think it does it. And it matters that you can spot it, but to other people, maybe it doesn't. What do you listeners think about that?

Speaker 1 12:23

Does it bother you? I know for me, if I'm scrolling through again LinkedIn, for instance, or Facebook or or any uh medium or or any kind of blogs or article platform, if I see something and it's very, very obviously been overly eight-eyed, I scroll past, I don't want to read it. Because it's it's everybody else's. You you came up with a brilliant analogy earlier, Ash. You were talking about when we go into a high street now. Or a shopping centre. And it's the same shops. Yeah. It's the same shops.

Speaker 2 12:55

You know that if there's a certain coffee house on that corner, there'll probably be that sort of store over there. You can always map it out before you go in, can't you?

Speaker 1 13:02

Yeah, there's going to be plenty of um bookies, bookmakers, gambling shops, really, where do you go?

Speaker 2 13:09

Every town centre, opticians. Yeah, but I was talking more I was thinking sorry, I was thinking more shopping centres, but I get what you're saying. But they're gonna be shopping centres. But you can map it out, can't you? You can map it out. You could literally say, right, I've gone round the corner here, same way that you can with certain service stations, with certain chains, you know what you're gonna get with them. It's like that with our high streets now, isn't it?

Speaker 1 13:27

And that's so it's gonna be much of a muchness, which is why we love going to the shambles in York, yeah, or the lanes in the bright. Or the lanes in Brighton. Because they're little individual shops, independence. Whereas if you just go into a main high street, and you say, we know you're gonna get a boots, yeah, you're probably gonna get a Greg's. You you you're gonna get again your optician your opticians and your bookies. If it's big enough, you'll get a Holland and Barrett. But it's the same wherever you go. Yeah. And that's the problem with using AI too much for content. And you know, later on I wanna I wanna rant about this because of all the things that we could be using AI to to do, creating content is the most basic, boring just oh, lightweight element of all of it. Anyway, anyway. Um let's move on a little bit. Okay. I've been ranting a lot about imagery. You ranting, surely, not. No, I've been ranting a lot about imagery lately. I wrote a piece on on LinkedIn a few weeks back about um I use AI for images. But I try to do it as ethically as I possibly can. So for instance, my AI photo images are only as good as they are because number one, I know how to create the prompts, and number two, I have a good and regularly updated stock of photographs from my amazing, awesome, brilliant photographer Vicki Head. So my AI generated lifelike imagery is only as good as it is because of Vicki Head. So I have I'm of the opinion that we should use AI imagery to fill the gaps that sit between our professional photography sessions. Once we have the budget to have professional photography sessions, not everybody's there yet. But I do not believe we should be using AI image generation at the expense of photographers everywhere. There's one there's one thing using it to plug a gap. Or, you know, I wanted a picture of me standing in front of a bright pink vault, for instance. Well, I'm not gonna have a photo from Vicki of me standing in front of a bright pink vault. Oh, I'll have to AI that one then. You know, and some of the ones where I was promoting uh the AI power lab weren't when we did the first the first uh opening of the doors for that, and I wanted all these magical kind of AI conjuring AI symbols and luminosity all over the place. Luminosity, luminous shapes.

Speaker 2 16:17

I know what you're meant. I think hopefully the listeners will meet, it's fine.

Speaker 1 16:20

I haven't got a photo from Vicki where I'm essentially doing imaginative electronic juggling. Artificially intelligent witchcraft. So I used it for that. Yeah. And it's the same with when when we're when we're taking art. You know, there's been a lot in the in in the news over the past couple of years about AI scraping art and essentially ripping off other people's work. There's a difference between asking it to create something in the style of a genre, yeah, and asking it to go and create something in the style of a specific living artist, for instance. So yeah, it it and it's the same with content. You only get good content from AI when you have good content to feed into it that you've actually created yourself. You have to keep feeding into it, right? You have to keep feeding it.

Speaker 2 17:08

It's not a one it's not a one-trick point, you don't do it once and then think it's got it sussed.

Speaker 1 17:12

Well, interestingly, I'm running some experiments with creating custom chat GPTs at the moment to try and get it to more of my content style.

Speaker 2 17:19

Yeah.

Speaker 1 17:20

And to see if it then keeps resetting. And it's promising, but even using that model is fascinating. That when we're running the tests, the kind of the screen splits into two, and you've got um a preview of the chat GPT model you're creating, the custom one, yeah, and then you've got your existing chat GPT that sits on the other side who knows you're writing. And if I put the prompt into the custom, even though it's been fed loads of generic content of that I've written of lots of different styles, it still slips into Broetry. And if I then put it across into the other side of the screen with the with the chat GPT that knows me, it can immediately see where all the issues are and correct it. Yeah. But as a standalone, that model that knows me will still default back to the open AI preferences. Okay. But in that scenario, it can do it well.

Speaker 2 18:14

Anyway, it's fascinating. So it's yeah, yeah, yeah. You're geeking out now.

Speaker 1 18:17

There are ways to explore, there are ways to explore it. But AI essentially is is only as good as the content that that you feed it if you're gonna use this stuff ethically. You know, it's if you feed it Pinterest images, it's gonna come up with your versions of Pinterest images. And let's not even think about well, well, hold on, who owns the rights to to that photo? If I go and grab a picture that I like the look of from look of from Pinterest, plug that into whichever image generation AI I'm using, yeah, and say create that, but with me, have we not just ripped somebody else's image off? Oh, that's a bit icky, isn't it? Yeah. And sooner or later, if people are doing that, we're gonna see more and more images appearing on social media, and we'll be able to go, oh, that's come from that program. And how long is it gonna be in a kind of Mel Robbins and Let Them scenario where the original I knew that was gonna come up again at some point. The original creator pops up and says, Hold on, that's my work you're ripping off.

Speaker 2 19:28

Yeah. Yeah, yeah, yeah.

Speaker 1 19:30

It's very important. Or somebody claiming to be the original creator pops up.

Speaker 2 19:33

It's murky, isn't it?

Speaker 1 19:34

It's mur potentially very murky. Yeah, so it's if you feed it generic images, if you feed it generic prompts, if you feed it what's the word for things well not thin thinking. Thin thinking. Thin thinking, thinking that it's not fat enough. We've not put enough meat onto the bones of our coats. Okay. We're giving it a little skeleton with with no muscle on it and no lard on it. No original voice.

Speaker 2 20:00

Yeah. It's the same thing, it comes back to that again, doesn't it? Whether it's photographs or it's, you know, whether it's images or it's written content, originality. That's what we're in danger of losing, that, aren't we?

Speaker 1 20:12

It's a bit like, to tie this into some of some of our wonderful woo elements for a moment, if I think back to the first um shamanic teacher I ever worked with, I will always remember her saying, woolly intentions lead to woolly outcomes. Yeah. And it's the same with feeding stuff into Chat GPT or whatever, LLM. Somebody asked me what LLM stood for earlier, by the way, it's large language model. Um, if you feed it your real writing, your lived experiences, your your nuance, all the stuff that you're talking about, it's smoothing off with your author's ash, your patterns, then you're gonna get sharper assistance.

Speaker 2 20:48

Yeah.

Speaker 1 20:49

So I know I complain that I see lots of content from AI that sounds and looks generic. It's because people are giving it generic input. Yeah. You may hear one of our dogs trying to join in in the background at the moment. Bless her, slight overshare, slight overshare. You might well she might not because we've got our super duper new DJI mics. So we have today. This is their first outing, it might filter it out. But if you hear whimpering in the background, TMI coming, brace yourselves, Gladys is in season. We're trying to keep her and Bailey separate, and they really want to be together. We want to be together. And again, there is to be even more TMI. We're keeping them separate. Not because Bailey has the Bailey has no nuts. However, he got so excited last time she was in season. Oh dear, we're going off. We ended up at the emergency vets in the middle of the night with a lovely net vet nurse having to um uh put his pieces away again where they had become It was like the veterin equivalent of the singing detective.

Speaker 2 21:58

Do you remember Joanne Wally with that you Know with what's his what's his name? The actor he played it was like trying to get some inflated balloon through a toilet roll.

Speaker 1 22:06

Let's just leave it there. So we're trying to avoid that. Anyway, so back to AI.

Speaker 2 22:12

You wouldn't get that with AI. Well, you probably would actually. You might.

Speaker 1 22:15

You might. So let's go back to what we were saying about this this this devolution. Okay. If you allow AI to get you to the point where you stop creating your own, I don't know, generating your own thinking, your own inspiration, imagination, ideas, creativity, you have nothing good to feed it, so it's gonna poop out generic shite. And there's your atrophy.

Speaker 2 22:44

Yeah, I like it. Well, I don't like it, but I like the way you're saying it.

Speaker 1 22:46

Yeah, mate, that's a scary thing though, isn't it? We started to talk about imagery, didn't we, didn't we? So let's just double back to that for a minute. Gladys's whimpers took us off off topic for a second there. So again, it's it's it's one thing to use AI to fill the gap. It's another thing to start seeing AI as a way to bypass professional photography entirely. And you know, this is a bit like the rant we had in in the last series, and the last episode where we were talking about um when we are trying to get speakers to pay to work. We're literally we're taking food out of their kids' mouths. If we get to the point where we start to think we don't need photography and we can just do it all with AI, we're starving whole families. Can we just not? That's not evolution, that's evolution. We start when we when we get to the point where we start where we as human beings start to frame any human art, craft, area of expertise, Ash. Yeah, all of the above. As unnecessary, and we start to excuse that by saying things like it's uh it's overpriced or I don't like having my photo taken. We start to make these we start to make arts.

Speaker 2 24:07

You said I was just gonna say obsolete, that's the word, isn't it? And that's the scary bit. We've seen it. We've got to get a supermarket checkout of people. What we have bring back the people at the supermarket, yeah, but where does it all end? Then we get into that whole thing of I don't know, perhaps it's just me, but I'm thinking like, you know, I've watched too many and too many versions of Terminator, it's like that, the machines take over. It's not just devolution, this is devaluation. Yeah, yeah. Honestly. And it's all terribly depressing, really, wasn't it, when you look at it like that.

Speaker 1 24:35

Let's think about because we were we were both journalists. If we think about then the the c the the collaborations we had with photographers as journalists, amazing. And it is a collaboration, isn't it?

Speaker 2 24:46

I miss those days. I've spoken to Vicky Head about this quite a bit. Yeah. And it's one of those things that I loved as a journalist on regional press was that you used to go out with the photographers and and go out and work together and come up with ideas for shots and to look at you know, building it into the story of the stuff.

Speaker 1 25:00

It's the thing again though, isn't it? When you're coming up with ideas for shots, sometimes you get a little bit of creative tension. Yeah, yeah.

Speaker 2 25:06

But that usually produces exactly that, and that's what we used to do, and we would get we'd come away with all these ideas, and then of course you'd have more content because you'd have more images from the photographer's perspective. That then, if there was a follow-up to the piece that you were writing, which there sometimes was, you'd got even more images to work with. So it was a win-win for both the photographer and the writer, the journalist, and it worked beautifully.

Speaker 1 25:30

Sounds that you you get that degree of interpretation, creative interpretation as well, so you can play with it a bit more. And I don't know about you, Ash, but I've never worked with a professional photographer and ended up with my hands on back to front.

Speaker 2 25:43

Or so I can't think about it. I've I've been asked to do some weird stuff, like anyway, let's not let's not go there. All above bold, I hasten to add, you know, given what's going on in the world right now, but nothing dodgy, but just you know, just some yeah, some embarrassing pictures of me somewhere, I would imagine, in the archives, in certain local press.

Speaker 1 26:03

Again, to tie back to the title of last week's episode, if we talk about an ecosystem, AI imagery for me, yeah, is the strongest when it complements the ecosystem. Yeah. But it becomes massively problematic when we try to use AI to completely replace it.

Speaker 2 26:21

Yeah, never to replace.

Speaker 1 26:23

No, that's that eco versus ego coming back in again, isn't it? Isn't it? So we started talking about art scraping as well. Um It's a great phrase, that isn't it? Art scraping. Yeah. For anyone who doesn't know what we mean by that.

Speaker 2 26:36

Yeah, Gungs, you actually explain it to me. I didn't quite know what you meant by that.

Speaker 1 26:39

So remember there are different types of AI, and I'll touch on this a little bit more later on when I'll get well I'll get a bit geeky and Asher will look at me, go, well I'm probably doing that a bit now.

Speaker 2 26:48

I'm going, I mean it's a cool, you know, phrase, but it's like I didn't know anything about it really at all.

Speaker 1 26:53

So you've got your large language models, which are the ones that are largely text-based, and yes, they can handle things like image generation as well, and they are improving all the time. Remember, if you first tried to get Chat GPT to create an image with words in it and they'd come back with all kinds of kinds of random hieroglyphics, now you can do it quite well. Um so you've got your large language models, but you've also got AI image models. And just as a large language model, an LLM is has that name because it's literally been able to uh absorb every bit of written work there's ever been in the world.

Speaker 2 27:30

She's gesticulating like madness, listener. She is a hands, and you took me nose off then. There's still plenty of it left over. I knew you were gonna say that, and you know I don't like my nose. I hate my nose. I think it's lovely. No, I don't like my profile, it's not good. I think that's why I didn't get that Kellogg's gig.

Speaker 1 27:52

Anyway, image models. So, in the same way that LLMs are trained on these massive, massive, massive data sets of text, I am gesticulating, I get carried away with this. My geekiness is on fire right now. Not my nose, like not your nose, not your nose. AI image models are also trained on vast data sets, often including copyrighted artwork, and that's where the ethical tension sits. So there's a difference between playing with it at home. So, you know, a few months back I remember sitting with some members of of my family, yes, and we were deliberately giving Chat GPT over-the-top descriptions of one another and asking it to produce a an image based on what we just described in the style of Beryl Cook, for instance. Yeah, and it was amusing, and we had fun with it, or go and dispaint that person we've just described in the style of Picasso. Oh, the ones you did with me and all the um movies, didn't you?

Speaker 2 28:55

Yeah, yeah.

Speaker 1 28:57

Yeah, not quite a few people are those on those, they're fun. But you're playing with genre styles privately and you're learning from some of those aesthetic uh genres, movements? Yeah. But if you are then commercially generating images, i.e. generating images to use in your commercial work in the style of ex-living artist, are you then effectively nicking their work? You're borrowing that person's idea. Idea and skill base and creativity and style. So again, it's muddying the waters a bit, isn't it? And again, it's a difference between asking it to create something in a genre. So for instance, I sometimes I'm a bit of a comic book nerd. Not to the not to the extent of Morgan Gleave if he's listening.

Speaker 2 29:48

I was gonna say, super cool Morgan Gleave.

Speaker 1 29:50

Super cool cartoonist. Yes. So I might say, uh I'm doing this presentation for a talk I'm I'm doing, I need some slides for the backdrop. Can you describe this scenario and do it in the style of the golden age of comics? That would be different to me picking on a specific artist who specifically drew for DC or uh Dark Horse or Marvel. Yeah, say give me one in the style of that specific artist, which would be ripping off that artist. Give me one in the sorry, I just went some of the. You know what I mean. That's the line, I I think for me that's one of the lines. Well, I've just crossed it, I've just jumped over it. Just because a tool allows us to do something, yeah, it doesn't mean that our responsibility suddenly becomes optional. Your responsibility is still there. Yeah. Now, is it the same if it's in the style of an artist who's no longer here and lots of works are in the public domain? So if I said go and create an image of this photograph of our village in the style of Van Gogh. Yeah, would that be okay? Maybe. Because is that now a genre because he's no longer here and again the the the works are in in the public domain now?

Speaker 2 31:19

I don't know how I feel about that. Yeah. You know as long as but uh uh could it be in just being really upfront about actually saying I was inspired by this and this is what working with whatever gave yeah I'd feel like I'd need it in I need it the context around it and I'd need the story around it.

Speaker 1 31:39

What these guys a hand painted piece of local art that we brought back from Athens on our last trip. I know Gladys, which is a picture of um the Acropolis at night in the style of coming a green fart.

Speaker 2 31:58

I don't know who you mean. I don't did I ever know? Yeah, because you remember when we chose it. I know, I can I can I'm actually picturing us where we exactly where we were.

Speaker 1 32:06

Anyway, forget what the genre is.

Speaker 2 32:08

But I can't remember what the genre is.

Speaker 1 32:10

In the style of a very, very famous, well-known artist who is no longer shuffling this mortal coil. Yes.

Speaker 2 32:15

It's not Lowry.

Speaker 1 32:17

Not Lowry. I can't even remember what it looks like now. I can. Remember that, so so there you've got Oh it's No, it's alright, it'll come back in a minute. Just carry on. There you've got a local artist who is paying homage to a style of a long deceased painter. Is that okay, rather than putting it into AI and asking it to generate it? Ooh. There it's there's a there are massive grey arrows.

Speaker 2 32:51

What do you think, everybody? It's hurting my brain a bit actually. I need to think about it.

Speaker 1 32:54

There are massive, massive grey areas. I think for me, it's one thing doing something in the style of a genre, it's another thing finding uh taking the work of a living artist or a living photographer and just picking up some of their work. Living photographer isn't is an artist, isn't it? So picking up some of their work and saying, do that. Yeah. Does Yeah, that doesn't feel right somehow. Do this, create this image of me presenting on stage in the style of in in an urban graffiti style. Yeah. As opposed to create this image of me speaking on stage in the style of Banksy.

Speaker 2 33:33

Um one is a kind of straight rip-off and one is a one's a genre, one's a style. Sorry, or rip-off.

Speaker 1 33:44

And one is a rip-off.

Speaker 2 33:45

Yeah. That's that's where I'm noticing the difference for me.

Speaker 1 33:48

One of them, you're looking for you taking everything that Banksy has ever created work. Yeah, and then AI is looking at all Banksy's pieces and going, yeah, we can we can make that look like a Banksy. It's a subtle line and it's a grey area, but I think I think it's a really important one.

Speaker 2 34:08

Yeah, interesting discussion though. Yeah, do let us know what you think as well as.

Speaker 1 34:12

Let's come full circle again. The the real power of of AI and let's go back to LLMs, because most people are famili more familiar with with the likes of Chat GPT. Yeah, okay. Or Claude or Gemini.

Speaker 2 34:23

With these new, sorry, just a thought, with these new sorry to go off topic slightly, but with these new mics, am I able to rove just to let the dog out? Yes. Okay, so I'll do that then and I'll carry on talking to you. And it shouldn't affect our um sound. No, you should still be able to don't go dancing around the garden. I'm not going to, I'm just going to let her out.

Speaker 1 34:42

Should be able to at least walk to the back door and I'm trying not to make a noise as I do that. And if she is still wearing her doggy chastity belt. What? I thought you meant me. Please make sure you remove it before she goes out. Oh god, no, she's not got it on. That's okay then. She's not got it on. I'll just let you, dear listeners. Imagine your own scenario, guys.

Speaker 2 35:04

I get through the gap.

Speaker 1 35:05

Where I say doggy wearing a chastity belt, of course I'm in a uh a doggy season nappy. That's the dog gate going, right? Okay. That's the importance of remembering to take said chastity belting air quotes off before they go outside to use the loo, which I should have knocked earlier on.

Speaker 2 35:21

We are chastity belt free. Go on, go.

Speaker 1 35:25

So as I was saying earlier, the least interesting thing you can do with ChatGPT or any LLM is say something like, Can you write me a LinkedIn post? The real power is pressure testing your thinking. Um getting it to argue the opposite of your stance with you so that you get to see different perspectives, spotting inconsistency in your messages, um structuring messy ideas where you've just had a massive thought drop, brain fart on a piece of paper and you give it your scribbles and say, make me something out of this, help me to find the pattern in this. Um turning chaos into systems, synthesising research, and you know, if you've noticed there are different models of Chat GPT you can use now according to what you want it to do. The research model might not be as instant, but my goodness, it's really brilliant at going and doing all the up-to-date research and gathering it all in one place and um giving you links back to it so that you can be sure that it's real and it's not hallucinating, synthesising that research and modelling scenarios. It's about using AI to extend your oh cognition, there's a good word.

Speaker 2 36:43

Yeah, I was gonna say thinking, another good C word.

Speaker 1 36:46

Yeah, not my favourite C word. No. Use AI to extend and expand your your your cognition, not to replace it. It's it's it's using it in a grown-up way, isn't it?

Speaker 2 37:00

Taz, just to let you know and let the listeners know now, I'm actually sitting on the other side of the dog gate because Madam's feeling a bit left out because she's usually in with us in the evening. Bless me. So I'm currently sitting in the dog bed with Gladys, just so everybody knows. So if you hear any heavy breathing, it's not me, it's my dog.

unknown 37:17

Yeah.

Speaker 1 37:19

I suppose with any conversation about AI if we did not mention the environmental impact, it would be um What's the phrase when something is Remiss? Yeah, but remiss, but when something is significant by its not being there. Anyway, there's a saying, I'm having a I'm having a manopause.

Speaker 2 37:48

Well sorry, I was too busy trying to get the dog to sit down, so we didn't get the tail back to the door.

Speaker 1 37:51

It's gonna come to me at three o'clock this morning. Everybody just message us with it, because you'll you all know all know what we meant we mean.

Speaker 2 37:57

I think you've had a moment to interesting Yeah. Da da da da da da da. I'm not I don't I'm sorry, I wasn't. Why don't you know? I don't know, because I was half listening but half sorting out the.

Speaker 1 38:08

How many ADHDs does it take to create a podcast? Probably more than we've got. More than us. Significant by its absence or something.

Speaker 2 38:18

It would be What's the phrase? I don't know. Doesn't matter anyway, back onto the topic. Now I've got performance anxiety again.

Speaker 1 38:27

You're gonna do a backwards roll? No. That's another podcast for another day.

Speaker 2 38:31

I think we've talked about that before. I'm sure we won't bore them with it again.

Speaker 1 38:34

And anyway, I'm being very serious because I don't conspicuous by its absence! Thank you. Right. So, yes, AI consumes energy.

Speaker 2 38:43

I'm talking to the dog, sorry.

Speaker 1 38:45

Just remember that you're wearing a microphone, so you might be lowering your voice over there, but that's gonna go straight into your mic. Oh, I'm sorry. And it's not furry. I'm in trouble, glad.

Speaker 2 38:54

I'm in trouble. She's not happy because she's out here.

Speaker 1 38:57

We're talking about AI and the environment. I know we are, sorry. Okay, over to you and I'm listening. So, yes, AI consumes energy, but so does the entire digital accommod a con a comedy. So does the entire digital economy.

Speaker 2 39:12

This is a digital comedy.

Speaker 1 39:14

Economy then. Economy was almost conspicuous by its absence. The thing for me is if we're gonna have that conversation, let's have it honestly and not selectively. So if you are condemning AI on environmental grounds whilst participating in disposable consumption patterns elsewhere, are you? You have been impressed.

Speaker 2 39:38

Have you written this down?

Speaker 1 39:39

No, I was doing my research. If you're having a great big old rant about how you're not using AI because you know the environment, but you're sitting there doing that in an outfit that you bought from Primary. Tilly, I mean Gladys. Have a word with yourself.

Speaker 2 39:54

Sorry, Taz. Gladys just got my mic, so apologies if that did you swallow it? No, she just went like that and ate it. So good.

Speaker 1 40:01

Because that would have been really interesting, would have we'd been able to get the sounds of Gladys's entire digestive system until it went clunk as it fell onto the patio at some time tomorrow at some point tomorrow. Yeah. Yeah, if if you are if you're complaining about the ethic the environmental ethics of AI whilst participating in fast fashion consumerism, for instance, then you need to have a word with yourself. You just do. Because there are so many things about the environment and the amount of energy that AI uses, but there are so many other things that that equally use a load of energy. You know, the biggest amount of energy in AI usage is right at the beginning when the models are being trained. When we're using it day-to-day, it's much less than that. And I think if we're gonna have that conversation, we need to not be selective. It's a bit like these people who say that, you know, who who quote the Bible for an excuse to be homophobic when they're covered in tattoos and eat shellfish. It's like, come on, we can't be selective. We've we've gotta we've gotta choose deliberately and be far more uh conscious in our thinking, you know. Streaming also uses energy, data centres use energy, endless device upgrades use energy. So selective outrage isn't helpful. There's a time and a place for that debate, but it's got to be in balance, I reckon. Ashwell, what do you think?

Speaker 2 41:46

Um I'm just I've got a slight problem. Keep talking, keep talking. What are you getting what's going on? I don't I'm trying to get the dogs mobbing me. Is it because you're sitting in her dog bud? I know, but I'm trying not to make a noise because she's now wagging her tail and panting down the I thought it was a good idea to sit in the dog bed.

Speaker 1 42:09

Gender readers, so I'll just carry this for a little while. Listeners, listeners. They might be reading the transcript on Bossfeed. Oh dear. If you could see what I could see through the the the bars of the dog gate, she'll stand up because her tail is whacking the microphone madly.

Speaker 2 42:33

Right, okay.

Speaker 1 42:34

Let's go back to the core of this episode.

Speaker 2 42:36

I do apologise, I'm sorry.

Speaker 1 42:38

When you outsource you're not going in there, no, the messy first draft every time instead of I think this is a messy first draft.

unknown 42:47

It is.

Speaker 1 42:48

I think I've got it. But at least we're not running it through AI. This is at least live. If you start outsourcing the first messy draft of everything that you do, you stop being able to develop that idea on your own. You you use the you lose the muscle memory of your thinking and your creativity. That's a nice way of saying it. Yeah. If it strengthens your skills, it's an assistant. Definitely. If you'll if it weakens your skills, that's where you're replacing yourself. I'm back now.

Speaker 2 43:15

Well done. Thanks. Well done. Sorry. Is that inattentive ADHD anybody?

Speaker 1 43:22

Yeah. So back to what this episode was about. The AI revolution only really becomes a devolution if A, we let our own voice disappear. Yeah. And B, we start using AI to the detriment of our fellow human beings, our fellow professionals.

Speaker 2 43:46

Yeah, that sums it up, doesn't it? And that that's my absolute worst nightmare is that we we lose that creative spark that's in all of us because we just lose confidence in actually testing that out from time to time, actually regularly. Not just about testing out, but having fun with it, enjoying it, enjoying being creative and not losing that ability to be able to do that, yeah, and how much that brings, and how much that brings for your customers as well, you know, when you when you do it yourself rather than relying on a third party and AI.

Speaker 1 44:16

Yeah. And there's another part of that ethics and energy debate somewhere about um if we're genuinely using AI properly to be able to work smarter and faster, but without replacing ourselves, yeah, then you could argue that we're saving some of the energy there. But if we're gonna be constantly playing with it to generate images for it, because it's because it takes a lot more energy to generate image or video than it does to create text.

Speaker 44:46

Yeah.

Speaker 1 44:47

So yeah, every now and then there's one of these fads that that does the rounds like create a caricature of me doing my business or you know, who will be my celebrity parents? Oh, that was quite a recent one, wasn't it? Yeah. And if we generate just, you know, one or two of those, okay. But if we're spending, if we're just constantly reinventing it and churning out 40, 60, 100 images, that's where we perhaps. What are people really doing that then?

Speaker 2 45:12

Is it really got to that stage already? Because I always used to think, well, occasionally, we all do that occasionally, as much as to have a bit of a bit of a laugh and a bit of fun with it, but to actually do it that much and rely on it that much, is that really are you seeing more of that all the time now?

Speaker 1 45:27

Yeah, yeah, absolutely. And it's and again, it's I think there's much better things that we can be using it for. Yeah. Back to that thing I said earlier, that the least interesting thing you can do is get it to create your content.

Speaker 2 45:43

Well, it's all about, isn't it? So, for instance, if I don't know, you've got an event or something coming up, you know, and structuring how that's going to look, you putting your marketing plans together, things like that, it's brilliant for that.

Speaker 1 45:53

It'll help me to create a marketing plan.

Speaker 2 45:55

Yeah, exactly. It's loads of things, and it will give you those stages so that you can keep on track with it, you can keep on you know But not to the expense of working with a marketeer.

Speaker 1 46:04

Most of the self-employed people who'd be asking you to do that would frankly be fumbling through and doing their own marketing plans anyway. It's not like we're talking about big corporations who are suddenly going to start getting rid of their marketing teams.

Speaker 2 46:14

No.

Speaker 1 46:15

Well, I hope not. Um It's about um if you create your own content first and you know there's not some there's something that doesn't quite feel right with it, there's nothing wrong with them putting that that into your chat GPT, into your L LLM and saying things like, um, why does this paragraph feel flat? Or how could I strengthen the context of this argument or the contrast of this argument? Yeah, that's sharp that's using it to sharpen your technique rather than replacing. Does that make sense? Yeah. So given that I mentioned it, what else are some of the things that we could use it for instead of, or as well as creating content? I said earlier, pressure testing your own thinking. So again, that could be things like argue the opposite of my position, find a weakness in this strategy. Um using it to find um to find unconscious bias. What assumptions am I making? Oh, that's a good one. Yeah. Um how would a sceptical audience challenge this thing I've just written, for instance? Um you can look for use it for pattern recognition. So we've done that with with with all of our episodes. We put all of our episodes of this series in so far and asked it to break it down and analyse the content and the segments. That was a really useful exercise and look at what we should be bringing in for the final episode arc. And we even went, how many episodes should this be in this series? Before do we need to do series two or do we just carry on with series one forever? That was really useful. Um, you could uh give it a whole list of your content that you've created that you have created, not Chat GPT, say for the last year, and say analyse this content, tell me the themes that I keep defaulting to. Spot the repetition in my messaging. Um and here's a brilliant one for spotting gaps. Where am I inconsistent in my positioning for my business?

Speaker 2 48:18

Yeah.

Speaker 1 48:18

And there again you build more strategy. Um scenario modelling. That's another brilliant element of using an LN LNA. So things like um if I raise my prices by 10%, what might happen? Right, okay. Or um if you're launching a product, can you map me uh three likely audience reactions to this launch? Oh, okay, yeah. Or um what are the what could be the consequences of this decision, not just at the first level, but at the second and third and fourth levels. Right. So you start to build foresight, which helps you in in more of your your planning and your evolution of of your business and your strategies. Um it could be using it for ideas expansion. So um thing I said to you the other day after you you'd you'd written a brilliant idea down in your notebook, and then you were too busy to to see it through, and I was like, well, take that, give it to Chat GPT, yeah, and ask it to expand it. So one of the things I sometimes do is take this kind of half-formed thought and expand it into three different angles. Okay. Sometimes when I give I can give Chat GPT the the transcript of our episodes.

Speaker 2 49:41

This is brilliant, isn't it?

Speaker 1 49:41

Because you've got it all ready there. Yeah, drop it in there and say, give me four different angles from this, give me direct or give me a series of direct quotes attributed to Asher or I from this ep episode that I can then turn into um into quote graphics. Yeah, that's great.

Speaker 2 50:00

And the sort of thing you could do that, but think how much time that would take you. Yeah. That would take you hours, wouldn't it, if not days, to create.

Speaker 1 50:07

But again, it can only do that because it's literally got the transcript of our spoken word. We're giving it our spoken word already there for it. Words.

Speaker 2 50:16

So it's based on again, we're feeding it.

Speaker 1 50:18

We're giving it, it's only because we're giving the information into it for it to then work, feeding it in for it to work with.

Speaker 2 50:24

Yeah.

Speaker 1 50:25

Um, if you've got there are some really kind of more strategic businessy questions you can ask it. Things like um I'm I'm doing a course, I'm doing a a uh a Mastery of AI course at the minute.

Speaker 2 50:38

Oh, which every guest has with a behind the scenes.

Speaker 1 50:40

So that's where some of these new ideas are coming from. So one of the things I've been working with there is this is my instinct around this launch, this new product, this audience expansion, this onboarding, whatever it is. What are the frameworks sitting behind this instinct? Okay, and that can be fascinating, or which thinkers would agree or disagree with this? You can go right into uh philosophy and strategy and and business mastery or loads of different areas because it's read more than we could ever read. Research synthesis, I touched on this earlier, that can be so so powerful. Summarise academic current academic research on X. Pull together five different perspectives on Y. Wow. Compare opposing schools of thought, yeah, and again, we're just saying, Can you write me a post?

Speaker 2 51:34

It's almost like insulting to the AI, isn't it? Yeah, you're almost like now I'm humanising AI, which is what I do with everything, but humanising it and going, God, I can imagine now it'd be like this, you know, this AI, obviously, which is now a physical thing, and it's like that, you know, going, for goodness sake.

Speaker 1 51:50

These things, when we use it in that way, rather than us devolving and losing our mechanism of thought, it encourages us to think deeper and yeah, and strategically and to think I hate this phrase, but outside of the box.

Speaker 2 52:05

Yeah, and also, yeah, those angles breaks our own patterns. Those different angles it's going to bring, perspectives it's gonna bring, that you might not necessarily think of or you might get there eventually, but you've now got all this beautiful we used that word, didn't we?

Speaker 1 52:16

Smorgers board of you know, I've sometimes used my journaling or even a dream I've had that I've then written down and taken a picture of it and said things like turn this really messy, messy brain jump into a logical arc, or analyse this dream for me, tell me what's going on what's on my mind. That's a good one to do, too. She's very different to our shamanic interpretation, yeah. But it's brilliant because it can say, Well, I can look at this through this lens or this lens or this lens, and wow.

Speaker 2 52:43

So literally, we're we're approaching, it sounds to me, like what we're doing at the moment. I know it's relatively relatively early days, although it's been around for quite some time now, but it's almost like we're approaching it like like the it's like the iceberg, isn't it? Yeah, we're only seeing the top bit, we're only seeing the top bit above the water, but there's so much going on underneath the water that we're not seeing, yeah? Yeah, we're not using it for, and we've just we're just literally, it is literally the tip of the iceberg that we're using at the moment, there's so much more it can offer. Yeah. So, and again, that's so important, isn't it? All the way through this. It's we're not saying we're anti-AI at all. If anything, we're for AI, but it's as an assistant to complement what we do, not to replace it. That's the important thing. Yes, otherwise, we can go down horrible rabbit holes, rabbit warrants and holes about you know the whole thing around get rid of all the we don't need humans, we'll just get get it together.

Speaker 1 53:34

I also think we've been through the initial stage of excitement, so you know, not too long back, it was fairly easy to get Chat GPT in particular, all of the LMNs actually, to to learn your voice.

Speaker 2 53:48

Yeah.

Speaker 1 53:48

You know, there was a time where I was actively teaching people how to how to get it to learn your voice, but the more AI updates, the more open AI updates come in, the harder and harder it can becomes because it keeps trying to default to the parent model.

Speaker 2 54:04

But again, so it keeps going back and I'm like, I wouldn't say that. Who's this you've just replaced instead of me?

Speaker 1 54:10

Watch what let's watch this this space in terms of the the the custom chat GPTs I'm I'm experimenting with at the minute. Yeah. Um but what if for now, instead of defaulting to putting something through chat GPT, because I don't know, you you don't think you're very good at writing hooks.

Speaker 2 54:27

Yeah.

Speaker 1 54:28

Or writing opening paragraphs, what if instead of getting it to do do it for you, you gave you some examples of your writing and then said, teach me how to craft stronger openings or stronger hooks.

Speaker 2 54:40

Yeah.

Speaker 1 54:41

Show me the common rhetorical devices I could be using more effectively. And again, explain why this this doesn't feel quite right for me. Can you explain why? So you so we're using it to build our competence instead of replacing it.

Speaker 2 54:57

But it's a bit like this is gonna weird connection I've made here, but it's a bit like when you take on a new pet, a new dog, a puppy or something, and you think you've trained them really well, and then they'll go backwards for a while, and you've got to keep reinforcing that, you've got to go back sometimes to basics to get to go if things are starting, if the behaviour's not as you want it to be, and you have to go back over old stuff. It's the same with this, right? You it's not gonna be one one shot at getting it right, you feed stuff into it, and then that's it.

Speaker 1 55:24

You've got to keep working with it because as it changes, as it develops, evolve in the way that we're using it. So, how many people in the workplace do you know who've gone from a solopreneur to maybe having a team around them? Yeah, and all of a sudden they need to have something like standard operating procedures to give to people, and it's all in their heads, yeah. And then they create this it's so easy to create this kind of way of working that that is self-defeating all the time because you think you don't have the time to get this all out of your head, and you don't know how to get it all out of your head, so that means the staff have to keep coming and asking you how to do it, which takes you off the type of the type the the the elements of the business that you're supposed to be doing, which was why you brought stuff in in the first place. Yeah, now with AI, we could say that this is the workflow that I usually use and talk it through with it, or even scan some documents and say, Can you draft me a clean SOP standard operating procedure from this workflow?

Speaker 2 56:25

Okay, yeah.

Speaker 1 56:27

So that is an intelligent way to use it, yeah. Or um I've got this. This taps back into the the epit episode from last week as well, where we were talking about how to get far more muscly sponsorship for events. Can you create me a sponsor sponsor package outline for this event? This is how much I need to raise, this is what it's gonna be. Tell me the sponsors I need to be approaching, what are the niche areas, what are the promises I can give, and help me to create a sponsor package outline. Can I ask a question as well?

Speaker 2 56:59

It's going off slightly, but I think it's still relevant. Yeah, just something that's dropped in. How do people work with AI in terms of you know, sometimes a lot of people text, they use, you know, they actually physically type into AI, but you get you can get a very different response when you use voice, and we're coming back to that again. And is that something to tap into to talk about?

Speaker 1 57:19

Because that type of thing is particularly if if you're still hardwired into wanting it to produce your content and you're worried about it losing your voice. One of the things that I've I've noticed over the past few months or so is that if you put your LLM into conversation mode, and I mean the one with with the two-way conversation, not just um typing or speaking into it and then letting it translate that into text and then get a text response response back. I mean, literally in the in the two-way conversation mode, because you are then interacting in conversation with the AI, you are going to get a response back that is far more human-sounding and would be far more like your level of writing.

Speaker 2 58:06

Yeah.

Speaker 1 58:06

But it's longer-winded, so a lot of people aren't doing it.

Speaker 2 58:09

See, I think that makes such a difference.

Speaker 1 58:11

It makes a massive difference.

Speaker 2 58:12

In the same way that when I talk to people and I say to them, have you thought about if you don't want to sit in front of a laptop and writing your book, use voice notes on your phone. It's in the same way, isn't it? Because it gives it to us another layer to work with.

Speaker 1 58:24

There's so many things. And and again, remember that while everybody's excited about AI, I'm guessing have nearly an hour. This was going to be another 30-minute episode. I don't think we have the ability to do a 30-minute episode.

Speaker 2 58:37

We're going to blame me going off and sitting in the blame glowing for at least 10 minutes. So don't blame glow. Actually, it was me that kind of made things worse, really. Yeah, you know. My tail's not wagging now. Yeah.

Speaker 1 58:47

I think again, we've got all really excited about AI. We've gone, wow, it can do this and this and this, and hey, I was there, and I've I've done it all, I've made all the AI faux pas, and you know, I think that's the other thing when you first start using it for content because you're not paying that much attention to other people's content, you don't realise just how A-I it sounds until you start to recognise that everybody's content looks the same as yours, and then you start to recognise how much you've lost your voice, but also remember that it's taken lots of retrograde steps. But it's also about finding that voice, it's taken lots of retrograde steps in terms of creating content, and we can get annoyed with it for that, or we can say, actually, this is the kick up the back side we need to recognise that we still need to be creating our own content, we can use it to help us expand our ideas and to come up with different um arguments and strategies and different perspectives, even to come up with an outline of that article or of that post. But just as an aside there as well, you'll also see a difference in the type of content it creates if you ask it to write you a post, rather than saying, I want to do I want to create a piece of writing on this topic. If you do not specify it's for a post, it will not then default into OpenAI's idea of what a post should be like, and you'll get a better style of writing.

Speaker 2 1:00:11

That's a nice tip.

Speaker 1 1:00:12

Don't tell it you're gonna write you want it to write for social media, because then it will try and optimise it for what it thinks that social media should be like, and it's technical term. Yeah, so you might be thinking, yeah, I've really levelled up now, I've paid for the the the paid for version of Chat GPT or insert your favourite AI here, but it's a bit like for those of you who go and hire a VA or more than one VA - Dee Atkins, we love you if you're listening.

Speaker 2 1:00:43

Yeah, we do.

Speaker 1 1:00:44

And Miss Make It Happen. When I decided to work with Dee, it was because I wanted to be able to access and tap into her specific skill set.

Speaker 2 1:00:54

Yeah, you talked about that on the last podcast as well, didn't you? Which is really important.

Speaker 1 1:00:58

So when that comes to when it comes to AI tools, in the same way that you wouldn't hire a VA and expect that V that same VA to just be able to do your accounts, your shopping, your dry cleaning, your graphic design, your client onboarding, your video editing, your podcasts, your car mechanics, you know, all of that, it's the same with an AI. They're not all built to be generalists. So you need to be really careful with that as well. And I and it's a bit like when we were when Ash and I still used to train people a lot on using social media. We'd rather see you learn to use one channel really, really well before you start jumping all over the place. And for those of you who are for I'm I'm leaning heavily into chat GPT because I think that's the one most people start with. If you have been using Chat GPT, but you've not yet, for instance, mastered the art of creating a custom Chat GPT, recognised um the difference between using projects and just using a new chat, uh, recognised all the different models and their different capabilities. What if you really geek out and really get to know everything that ChatGPT, for instance, is capable of before you think you then mastered AI and leapt onto the next the next one that happens to have a special offer on it. And you know, I'm just a beginner here.

Speaker 2 1:02:26

I consider that I know a fair bit more than most of the people that You certainly know a lot more than me that I'm coaching and supporting because I've geeked out a bit you do geek out, and I'm glad I'm grateful that you do. Thank you, ADHD. But but it comes back to that it's the iceberg analogy, isn't it? That there's so much under the surface that you're looking for. Learn to use it properly.

Speaker 1 1:02:47

Learn to use it properly, you use it to expand your thinking, expand your creativity to help you to really really get on top of thought leadership and get underneath it too, all the way through it, instead of just saying, write me a post for LinkedIn or write me a caption for this Instagram thing. You know, if all we are using AI for to use a really old saying that it that works really well, I think, with in in this scenario, if all you're doing is using link to using chat GPT to create content, it's like using a Formula One car to pop to the corner shop.

Speaker 2 1:03:24

That's nice.

Speaker 1 1:03:26

Yeah, so yeah, not a good idea. I keep saying it, the least interesting thing. The least interesting thing you can use Chat GPT for is to write your content. The most interesting thing it can do is to sharpen your own thinking. So we should probably wrap it there.

Speaker 2 1:03:43

I think you should. Um thank you. I feel like I've learnt so much today. I was really nervous about this over. Could you tell? I got slightly distracted with the dog because I was.

Speaker 1 1:03:50

I wish you could see her, um uh everyone listening.

Speaker 2 1:03:52

You have actually got a picture somewhere so you could actually put it out because it make them laugh.

Speaker 1 1:03:56

Asha is currently wearing one of the on-trends at Wake hats that's got the the kind of little satyr hats, you know, the ones with the ears and the little horns. I've got a pinky coloured one. Asha's got the green colo uh cream coloured one. And she looks really sweet.

Speaker 2 1:04:12

Thank you. She's a forest pattern, jungle pattern leggings. Yeah, because I thought I was going to go on my bike tonight, but I'm not going to.

Speaker 1 1:04:20

A t-shirt with the Statue of Liberty snogging a statue of Justice that says liberty and justice for all. Well, I hope that hasn't affected the mic, me dangly bits.

Speaker 2 1:04:31

Sorry, it's not fairy mic. Furry Mic ran away on my dangly bits. And the hat.

Speaker 1 1:04:36

And the hat, yeah. On that note with Asha's dangly bits, shall we call it thus? So until next time, we will see you next Tuesday.

Speaker 2 1:04:48

You've been listening to Awesome New Off Topic. Follow or subscribe to make sure you don't miss the next one. We're Asha Clearwater and Taz Thornton, and we'll be back soon.