Episode 32

E32 Blending Human Creativity and AI in Podcasting with Colin Gray

Unlock the secrets of podcasting with industry pioneer Colin Gray and discover how the fusion of human creativity and cutting-edge AI is transforming the landscape. As a tech-savvy educator turned podcasting virtuoso, Colin Gray joins us to share his journey and the innovations that are making podcasting more accessible than ever. We delve into the essence of what makes this medium intimate and unfiltered and how platforms like Alitu are breaking down barriers for creators everywhere.


  • AI has made podcasting more accessible and efficient with advancements in transcription, editing, and content creation tools.
  • The future of podcasting will likely involve a balance between AI-generated content and human-created content, with the latter being valued for its authenticity and uniqueness.
  • Certification of human-created content may become a distinguishing factor in a world saturated with AI-generated content.
  • The use of AI in education is expanding, with applications in instructional content creation and personalised learning experiences.
  • Quality content and the human touch will continue to be valued by audiences, even as AI technology advances. Human content provides value and unpredictability that AI-generated content lacks.
  • The decline of authenticity on social media has diminished the enjoyment of platforms like Twitter.
  • Embracing imperfections and avoiding over-editing can lead to more genuine and engaging content.
  • AI-generated clips and automatic clipping tools still have limitations in selecting compelling moments.
  • The ethics and morality of AI and its potential consciousness raise important questions for the future.

Links relevant to this episode:

Thanks for listening, and stay curious!



Tools we use and recommend:

Riverside FM - Our remote recording platform

Music Radio Creative - Our voiceover and audio engineering partner

Podcastpage - Podcast website hosting where we got started


00:00 - David Brown (Host)

Hello everybody, welcome to the Creatives with AI Podcast. I'm your host, david, and on today's show we have Colin Gray. Colin's a podcaster, speaker, phd, and he's also the founder of thepodcasthostcom and Alatu. The podcasthostcom is a huge audio video and written resource on how to create a successful show. I've been spamming the videos in preparation for this today and I can tell you he's got some great advice and there's some great stuff in there for aspiring podcasters and videographers and YouTubers, if you want to do that as well. So well worth a look on that. And Alatu is a podcast maker tool, so it's a platform that helps people create and produce their podcast in a really easy way, and I'm sure Colin will be more than happy to tell you a little bit more about that and instead of me butchering it. So let's get cracking, colin.


Welcome to the show. Thanks for having me, david. Good to be here. Awesome, that's probably a good point. Maybe if we just quickly. It's the standard podcast format, so if you can just give a little bit of your background, just so people know where you're coming from, and then we can get into a conversation from there, yeah, sure, I mean, the short story is I was an educator years back.

01:16 - Colin Gray (Guest)

I well, I started as a freelancer doing web design, e-commerce, that kind of stuff, helping people set up sites, but I ended up teaching that at universities. So I was teaching a little bit of web design, a little bit of animation, all that kind of stuff, and I eventually ended up in learning technology, which means just really teaching teachers how to teach with technology.


that came out back in started:


that? When was that? That was:

02:13 - David Brown (Host)

Okay, yeah early, early days.

02:15 - Colin Gray (Guest)

Exactly, yeah, it was the time when you still had to go and buy a like a USB stick MP3 player plug it into your computer and like drag and drop the files over.


It was horrible. It's an awful experience. Nice, yeah, but I ended up. Despite all of that, I just I just loved the medium. I mean, it's all the things that people love about it these days, which is the kind of the intimacy of it, the personality that comes across, the real in-depth, you know, information and entertainment you get from it. So I just kind of fell into it and started writing about it and really that writing is what became the podcast host, which is the blog. That's really the kind of central hub of our business today a whole bunch of blog articles on there, a bunch of podcasts, videos and, like you said, led to the developing the software as well, because that was one of the things that our readers were really struggling with was the editing side of things. So that's kind of a potted history of how I got to where we are today.

03:11 - David Brown (Host)

Yeah, it's. It's been really interesting. I've talked to a lot of people who've done it for a long time and and the progression, and I think you're gonna have a really interesting perspective on it as well, because when you started certainly, I mean even like you said, you know, when you started in the beginning it was a very manual process and you know part of the reason that I suspect that we've seen this explosion of podcasts is well, one I think people like. I think I think people like having an uncurated feed and an uncurated conversation.


So broadcast media is very, very, you know it's very controlled. You've got different media channels have their own sort of agendas and you have, you know, some stations are very conservative, some stations are very liberal and they have audiences and they tailor their messages and I think people like having an unfiltered sort of you know, experience. I think that's one from the audience perspective. I know I certainly do. But the other side is the tools that are available now. Like you know your tool, alatu, and I think there's it's so easy to do these days and if you sort of sat back and just reflected a little bit on how the industry has changed even in the past, I would say three years, what, like? What do you think about that? And and how do you, you know, how do you think that's gonna go?

04:39 - Colin Gray (Guest)

oh, yeah, it's, it's. It's just really. It's almost weird to look back at the progression, like, like you say, not even that long ago, how different podcasting was. I think you're totally spot on there about why people fall into it. Is that kind of. It's the one place, I think, where you get the transparency, the honesty, the openness. It's the one place where people, no matter who they are, from celebrities to complete normal people like ourselves, like just talk about you know their daily lives, their experience, their, you know, their, their loves, their hates, all those things that we're just really interested in.


date sometimes. It was about:

05:46 - David Brown (Host)

yeah, I was somewhere around there, I think, something like that. I mean that was a watershed moment.

05:50 - Colin Gray (Guest)

Yeah, totally like there was just this whole region of people who suddenly had a podcast app by default on this device that they have in their pocket every single day, connected to the internet too. So it meant that you didn't have to have this horrible process that I used to do, which was literally you would plug in your MP3 player, delete all the episodes that you've listened to, upload the ones that you have, and it's just nasty. But suddenly it was just. It turned it into this thing where you can just pull out your phone at any point and there's a new episode there. The latest stuff is just there waiting for you. It was just. It was so life-changing that part that was back then, wasn't it?


yeah, yeah, exactly, you could use it on iPods and but it was yeah, it was kind of the the automatic syncing. That just changed the game. So, yeah, I don't know the last few years, though let you, yeah, what have you seen changing most in the last few years in podcasting?

06:42 - David Brown (Host)

I, I think it's that. I think it's the audio tools and the editing tools yeah for me personally, I think that's made the biggest difference.


And also things like transcription. You know now, or even a year and a half ago if you I mean, there were specialist tools in the market, yes, but they were very expensive and I think they, you know you had to be really serious about doing it. So maybe if you're an agency, you might have that software because you used it across a lot of clients and things like that. But you know, even to generate a transcript of an episode and do your show notes and all that sort of stuff in summaries, it was very, very time-consuming totally now.


It's a big idea yeah, you know we'll, we'll record. Obviously you've got your own tool that does this as well, I think, but you know we'll record this. And we use Riverside, and you know I've talked about it before. But you know we'll have this recording when we finish. It'll be five or ten minutes, and then the recordings will be ready to go and I click a button and it will generate show notes, transcripts, suggested titles, you know, keywords, everything that I need, and then it will. You know I even have the ability to, you know, to add intros and outros to the beginning and end. I can edit the whole thing. I can do all of that. I mean I could. We could literally have a produced show and I could have it ready to go for tomorrow, and that's just that. You couldn't do that until about a year ago, really, I think.

08:07 - Colin Gray (Guest)

And it was. It was much harder. Yeah absolutely yeah, we kind of got to that point with Alatu maybe a couple years ago, but it's yeah, it's been. It's very recent, very recent, and it's it's terrifying, isn't it? Like transcripts? We've spent thousands of dollars on transcripts over the years. I get an episode, transcribe video is all that kind of stuff and that's all human. It was all like using revcom, that kind of thing like a dollar an hour, sorry, a dollar a minute.


Yeah, yeah, yeah, but but now it's just, I mean it's funny if you think yeah, you know that's.

08:37 - David Brown (Host)

That's 60, 65 pounds for a normal episode and then you know, I don't even pay that for. You know my subscription to the, you know to my, to my platform and yeah, absolutely, yeah, yeah but it's like that that's just one of the things that's been disrupted.

08:50 - Colin Gray (Guest)

Isn't it like that's a whole industry that we we tried some of the automated tools in the early days and we thought now it's still worth the proper like human-based ones, but nowadays, actually, it's just, it's just about as accurate.

09:03 - David Brown (Host)

I had a really interesting conversation with a guy actually named Steve Dunlop who's gonna be on the show later. We're gonna record a chat in January and I don't want to steal all of his thunder, but one of the things I heard him say on another podcast was he made the point we were. He was talking about voiceovers and obviously you know there's a tremendous opportunity for people to get much better sounding voiceover work because you can go to something like 11 labs and you can drop some. You know, just put some text in. It's got an AI generated voice that's actually very, very high quality. Yeah, and you know there was this, the, I think the.


The person whose podcast he was on you know, said isn't there a worry, you know, in the industry about this? And he said actually no, because the companies that are spending big budgets on doing ads and that sort of thing they're like the cost of a real person to do voice over is like not even non-existent in the budget. Like they've got a four million pound budget and to pay 200, you know, 200 pounds out of four million to get a human to read something is not a bit like they don't even think about that. So there's no kind of downward pressure on rates and those sorts of things for for real voice over actors, but it's it's the small companies, and this is the same that I'm seeing with copywriting as well.


It's not you know virgin media or you know big companies like that or scar, whatever, like they're still going to get people real people to do the work for them. They're going to write their own copy, they're going to do all of that. You know they're going to pay copywriters to do work for them and and videographers and all that sort of stuff. That's not where the issue is. It's the issue with the small companies, which make up like 90% of all businesses, and it's it's those of us and I I've said many times on the podcast before. You know, ironically, I have a podcast where we talk about, you know, ai eating into people's jobs. Yet I use all the AI tools I possibly can so that I don't have to hire anyone yeah, but.


I don't know about you, though, I've found.

11:11 - Colin Gray (Guest)

I mean I am absolutely certain that it will cut some jobs. It'll cut, it'll cut some earnings from certain people, probably a lot over the years, but I it hasn't stopped us, it hasn't cut down what we pay real humans at all. In the last year, the last two years, it's just made me do more with with the tools, because we've kept the same budget for employing actual people to do real stuff, yeah, and I've just done more using AI myself with a lot of the stuff I've done.


So yeah, I'm looking at it just now, it's just, it's, it's kind of given us 10 times, 20 times, whatever yeah we can do and I.

11:47 - David Brown (Host)

The other thing that's been really interesting is to see who's doubled down on it and who hasn't, and it's for me it's going to be interesting, I think, to watch over time to see what happens.


So, for example, adobe has just they haven't even doubled down, they've like tripled down. They've put AI, generative AI, into all their visual tools, their audio tools, like everything, and they did it so quickly. It's unbelievable how quickly they did it. And you know their position was is we work with creators, we're going to give them tools. This is just another tool that they can use and you know we're gonna we're gonna be the lead on it where you have other platforms and other industries where they've said absolutely not, we're gonna resist it the whole way, we're not gonna use it. And I just I wonder if it's gonna end up sort of like the whole Napster and music thing, where it's almost like some of those people miss the trick and they're gonna miss the wave in the beginning and those companies that are the early adopters are really gonna end up being even more successful in the end there's a.

12:54 - Colin Gray (Guest)

There's a funny balance there, though isn't there because we so we did implement, we experimented with it early. So, as soon as you know, open AI, particularly their API, became quite open and we start playing with it, see what we could do, and we've got a few implementations built into into our tooling so you can, you can get show notes written, you can get titles described, all that kind of stuff. But there's a lot of the work that we did, we've we've also so it kind of goes into that. We created a show planner which is, in a way, it's just a wrapper for chat, gpt to help you plan your podcast okay, I'm totally free tool that we wanted to build just to help people create their plan. It asks the right questions, it puts some of our experience around, like how you can plan one well for success, all that kind of stuff. And we've we built it.


What we started maybe six months ago, maybe even closer to a year ago, on that one, and we've just found ourselves refactoring that again and again and again as things have changed with the API and particularly things like like one example is the, the chunking. So initially, you know we wanted somebody to upload a transcript, for example, to get an idea of their voice and what they talk about, and we had to try and figure out how to chunk that into pieces that we could actually upload to the API, because it can only take a certain amount of right of words and then. So we put a lot of work into that and then a month later, suddenly the you know the size of the chunks that you can put in are much bigger and actually there was no need for it whatsoever so we


keep spending time solving these problems that open AI and all their competitors to solve a month later. So there's, there's definitely like a balance between being that early adopter enough to experiment with it, but also not putting so much resource into something that's just entirely changing entirely. I don't know it keeps the effort you put in can often be thrown out straight away as soon as it's, as soon as something does change in future well, that's just that.

14:53 - David Brown (Host)

That's the nature of a new technology, isn't it like you've got a, everybody's got to figure out how to use it yeah, and you know, and the landscape is constantly changing because, like you said, you know the the requirements are changing, the prices are changing.


nk that we'll start to see in:


When I started using Riverside, there was no AI, there was no anything built into it. It was literally just a recording platform and you download your video and audio files. Now it does all this stuff that it didn't do before and you know, shout out to podium. I use podium as well to do show notes and and and all that sort of stuff, and podium in the beginning was literally like drag and drop a file and then it literally just gave you a zip file of a few text bits back and that was. You know, that was. It was very rough and ready, but now it's got a much nicer UI and it's a lot easier to use and you know and and all this stuff is developing really quickly. So so what, how much do you actually use AI yourself in your sort of day-to-day and what sorts of things do you use it for?

16:28 - Colin Gray (Guest)

yeah, a lot, to be honest. I mean we I use it so much for planning and ideation and stuff like that, like as a as a creator. You know, one of the biggest things we do is come up with new ideas and topics and things like that. And just so, something I was working on today, for example, we run a newsletter, the podcast pointers, where we send out, you know, tips for podcasters once a week and I was putting together a new series for for that. And something I like to do, and I try and encourage a lot people to do in their podcast, is actually to think in sort of larger chunks and seasons.


So you link your episodes together. And I do this in my podcasts, I do it in my emails, I do it in my youtube videos, that kind of stuff, because there's a whole lot of advantage in having you know five podcast episodes that actually link from one to the next, to the next, rather than you know a different topic every single week. And it's the same with blog posts, the same with emails. So AI has made that a whole lot easier for me because it helps me to kind of pick apart the parts of a topic I'll I'll use it to research the questions people are asking just now. Give me some ideas of things I could write about and then I'll use it to tease out.


How do you break this down into components and obviously I can put my own thoughts into there as well but I'll end up with, you know, one big question that breaks down into 12 different sections and then I'll actually take every section and put like five to 10 bullet points on every one of those sections and, yeah, by the end of that you know 20. That used to take me maybe an hour, an hour and a half doing it by myself. So you can do it by yourself, but with the help of someone like chat gbt I can do that in 15, 20 minutes and I still put my own spin on it. I still put my own experience into it. But it just creates this plan for the next. What if it's 12? That's like three months worth of content. So I've got 12 episodes all planned out ahead of time. I can turn that into, like I said, an email series, but I can then convert it into a podcast episode series as well and then for.


YouTube and all the different places, and that's one of the biggest places. It's actually saved me. You know, headache, time, stress, just helping me to distill those ideas that are somewhere in my head. But I just need prompts exactly.

18:33 - David Brown (Host)

Well, yeah, finally enough. Yeah, it's um. It's chat gpt. Is that your favorite tool or do you have others that you like to use for different tasks?

18:41 - Colin Gray (Guest)

I actually, yeah, I subscribed to that in the early days, as soon as they had their paid plan came out. I subscribed to it just to play around with it and I've ended up sticking with that. I've played around with some of the other ones, like Claude in particular, but yeah, I've mostly stuck with chat, gpt actually, especially since they've made it more multimodal. You know, I don't have to go elsewhere for my images or to upload files and all that kind of stuff too. So, yeah, what's your favorite these days?

19:06 - David Brown (Host)

I have different.


I use different tools for different things so to do the ideation in the outlining and all that sort of stuff. Chat. Gpt is genius at doing that sort of thing. I feel like they've they. Well, they've definitely tinkered around with how it works in the background, because you used to be able to say you know, suggest a. You know, uh, I don't know, suggest a summary of or an introduction for, you know a business plan for a podcasting studio, and it would just write you a bunch of stuff. What it does now is it says if I were writing this, here are the topics that I would include. So it doesn't give you the answer. It steers you towards the answer instead of actually giving you the text which is opinion yeah, which is, but it's it's.


It's more of that bullet point kind of ideation thing that you're talking about, and I find it really good for that. I also find it very good for analyzing text that I already have. So I might write something and then I'll say what analyze this like, summarize this for me, and what I want to see is is the summary that it generates what I'm actually trying to say. So, and I do that a lot with things like cuz, I work a lot with public sector and I have to write a lot of information that goes into Bids and stuff like that. So I'm like I have a question and then I need to write an answer for that and I want to know did I actually answer the question? So it's good for doing that kind of stuff as well, and you don't have to worry about it hallucinating then, because it's it's only analyzing the words that you've given it. Do you know what I mean? And so I find it really good for that limits that yeah.

20:43 - Colin Gray (Guest)

I like a lot for critiques. Actually, I often Use it for if I'm writing up, let's say, a page to sell one of our courses as an example, and I'll put in some of my copy, I'll put in the headline and a few paragraphs and I'll just say, from the point of view of a sales page. You know, take the point of view of this copyright. You're that copywriter, their style, give me some critique on this and then ask it once it's given the critique. Often I'll feed back in and say, using this critique, can you rewrite that?


Yeah landing page and it's funny when you have you tried to have given it a few levels of this in the past. Yeah, in terms of yeah, so you've just written that one. Can you critique your own page from the point of view of this other copy? Writer? Oh, based on that critique, please rewrite it again.


I mean write it yeah, yeah, exactly, and often you can do that two or three times and sometimes it ends up going off Off the rails, but actually more often than not you get something really kind of refined out of it that way, I think yeah, I remember about a year, almost a year ago already, there was a.

21:46 - David Brown (Host)

There was a professor at the Wharton School of Business and he was one of the very first Sort of university level professors that came out publicly and said that you know that they had integrated it into the MBA program like immediately.

21:59 - Colin Gray (Guest)

Yeah and.

22:01 - David Brown (Host)

But what he found is is his coaching to his students was that you need to go Basically five levels deep so you can ask it for something. It'll give you an answer and then you have to narrow that down, and then you have to narrow that down and you have to narrow that down and he's, he's like it. Once you get to about the fifth time of doing that, you're now at something that's actually really useful, but you can never take the first answer that you get and yeah you know which, which I've found as well.


It's. You know, if you just ask it something, you usually need to refine and refine and refine.

22:34 - Colin Gray (Guest)

Yeah, yeah, keep asking questions of that.

22:36 - David Brown (Host)

Yeah, absolutely.

22:37 - Colin Gray (Guest)

Yeah, I find we. I. The other thing I use it a lot for is Repurposing our media. So we talked, before we got on this around, what you'll do with the video, what we do. The recording will probably turn into clips, will put it in other places, that kind of stuff, and I find it really quite useful for that. But similarly, that needs a lot of guidance.


I find, like show notes are the bane of every podcasters life. You've recorded an episode, you speak for an hour and that's what we like to do. We like to speak, we like to chat, but we don't like to write. But a podcast is like ten times as effective in terms of a growth kind of vehicle if you have good written material Alongside it. It just expands so much how you can grow a podcast if there's Show notes, which you know good show notes are really just a blog post that says the same thing as the podcast episode. And even better, you'll have things in there, like you know, a list of the resources mentioned with links. You'll have the people mentioned with links. You'll have Takeaways like here are the five or six things you can do with it, here's the tasks you can follow up on, and maybe chapters as well, like with time codes and stuff like that, and that all used to take me like an hour hours.


Yeah, absolutely, but with a good bit of guidance. I find if I just give the transcript and say, give me a summary and give me the chapters, it's, it's always a wee bit hit and miss. But if I actually kind of give it a little bit of the context, like this episode was about this, this and this, you know, I think the first half of it we covered this, but first, second half we covered that and then pick out chapters that maybe match those topics and give me some suggestions and then you can start to kind of narrow down based on that. I find, yeah, it takes a couple of rounds, but that's actually one of my biggest time savers. As long as I can kind of narrow down, refine it a little bit, you can get some really good results out of it that way.

24:26 - David Brown (Host)

Yeah, I've seen some people doing some really amazing prompting with it as well. I was. We had an event here in Tumbridge Wells not too long ago that I was sort of emceeing for, but one of the speakers was works in marketing and he was actually putting in all the audience codes, right. So getting extremely specific to say this is targeted at and then giving all the different sort of internal industry codes, for you know how they refer to different demographics and all that sort of stuff and I was like I'd never even thought about that, you know.


But but for someone who's an expert in an industry right and, and they use those metrics, and he knows those codes off the top of his head because that's how they target their content, that's how they write their copy, and you know you can feed all that stuff in again to get more specific, more tailored answers.

25:17 - Colin Gray (Guest)

That's interesting, actually, one way I've used it that way. I've been learning Spanish recently, so I've used that for a bit for language learning, just asking questions, having a conversation, and I've discovered one day that there are like official classifications In Spain for the levels of Spanish, like there's. I had no idea about this until I discovered it just like a year into trying to learn Spanish, and suddenly, with chat GPT, then I could say give me a conversation, but only up to level what was it? A B2 or something like that. Like that's a kind of the classification of the level that I think I'm at or pushing towards. So yeah, you're right, it's like it's hard to find those, but maybe that's what you can use chat GPT for as well, or whatever you know model you're using isn't it like?


to actually try and discover some of those frameworks, some of those classifications within your industry that actually then give it really good guidance on on Refining, on getting more specific, on really making it much better quality.

26:15 - David Brown (Host)

Well, and I didn't really realize that you had an education background and I don't know if you've listened and to any of the the educators that I've had on, but I've had a couple of educators on and and it's this the stuff they're doing in education at the minute, and the stuff that's coming is amazing, not just from the fact of Sort of teachers and educators using it to create New and better and more engaging content for their students, but also for students to prepare for things like their A levels and their GCSEs, which are the standardized tests here in the UK.


If you're in the US, that won't mean anything to you, but it's sort of like SATs and ACT. So, but you know those written answers though those are written answers mainly, and the kids these days I mean my son just did his GCSEs. You know they. They spent nearly a year practicing how to answer the questions and all that sort of stuff. But now what the students are doing is they're saying they're taking the mock question, writing their answer and then saying great it and Suggest a way that I could have done it. I could have written a better answer, yeah, and they're using that Almost like a tutor to be able to get better and better at answering the questions and I'm like that's genius. I mean, that's such a genius way to use the tool and you know, as an ex you know sort of educator in someone training educators, that I figured that must be interesting for you as well.

27:40 - Colin Gray (Guest)

Yeah, yeah, absolutely yes. I mean we're I'm quite far out of the education world these days, but we still do our own education. Of course, we still create our own courses and I even seeing Some of the platforms like you know, teachable and podium stuff like that like which one was it? Well, I think it was thinkific. Actually, I was playing around with create some new courses and you could just type in a course title now and it basically Creates the framework for you to then fill in the gaps and obviously you'd want to put in your videos and things like that, but arguably, you could get them done by you know, ai and some motor form as well these days. It's, yeah, I find that one of the most fascinating bits of I find is the kind of explosion in faceless videos on YouTube and how they actually how they actually still really engage people. Have you come across a lot of them?

28:29 - David Brown (Host)

I'm sure you have they're the bane of my existence.

28:33 - Colin Gray (Guest)

I hate them with a passion that burns of a thousand sons. I get that Yep.

28:40 - David Brown (Host)

No, it, it. I Run across them all the time and you know I'm trying to build my skills as a as a podcaster and as an interviewer and as a host and and you know, and also trying to, you know, do the audio processing, because I want to do more of that myself, and then I want to get more into doing video next year and so trying to learn things like Premiere Pro and how to do basic Video editing and that sort of thing, and so I'm constantly looking for instructional videos and and you know how to use and all that sort of stuff. And the sheer number and proliferate, proliferation Lately of the faceless you know AI Scripted and read Videos is doing my head in because they're terrible.


90, 99.8% of them are terrible and it's just, it's clogging up the people like you and and Mike Russell who I've had on before, and you know, there's tons and tons and tons of people who do amazing instructional content out there and it's just getting absolutely blown away and and drown in all of this Frankly terrible video, you know content and and people are just churning it out by the thousands, you know, because it's a numbers game and it's, oh it's, terrible.

30:04 - Colin Gray (Guest)

I don't know, and I'll tell you how I really think in a minute. Yeah, and this is the sad thing. The sad thing is that you that like point zero two percent that succeeds is getting a you know so many views and real profit, which makes it worth. It's like it's like spam email, isn't it? You wonder, like why are you sending me this? Like whoever actually clicks the link to buy this crap that you've just spammed me with? But there must be like point zero, zero, zero, one percent to do and therefore it's worth it. It's gonna be the same with content from now on.


But I don't know, do you, do you see a world in the future where there's Certification of some sort like humanity, of humaneness?


It's kind of what I'm thinking towards with our, because we, we do all types of content, like our blog is one of our biggest traffic drivers. It's just written content. Obviously, written content is so under threat just now, but equally, like we talked about, like even video and podcasting and audio, is still a little bit too. But I think that keeping our humanity in there, keeping our stories, our personality, all that kind of stuff, it can still be copied, but there's something around we've committed to no AI writing whatsoever in our blog and therefore we'll be able to have some kind of you know Guarantee if people look in future that this is written by a human. I don't know, I'm not sure if that's idealistic, optimistic, but I can see a world where that is a. There's a small percentage of people 10, 20, 30 percent who actually go out with their way to look for places that are certified human written human created that kind of stuff and maybe even willing to pay a little bit for it.

31:37 - David Brown (Host)

Yeah, it's. I had this conversation with someone the other day and it's a question I've thought about a lot and and and something I want to dedicate a whole episode to at some point. But I Think where we'll end up 20, 30 years from now is I think we will end up where I, you, will have AI content. That's amazing and if you need, you know, if you want imagery for your show or your website or whatever, you can just go and get AI to make it. But there will. There will be, exactly like you said, there'll be this human, you know, sort of human created element, but that's going to be like the super. You know, that's going to be the super high level, the super expensive stuff. Do you know what I mean? Because it's it's just going to get to the point.


You know where it's a race to the bottom on a lot of that stuff, you know it's. It's sort of like music. Now, if I need music for for the podcast or I'm doing a video and I want to get some music or whatever, I can just go get some AI music that's been created that has no copyright on it, because you can't copyright it, because it was created by AI and it's totally original and so that can just be used. But then if I and if and if that's just my short form social media videos and stuff like that that I do, that's fine.


If I do a film, documentary though I'm not going to use that kind of music for that I'm going to get a, you know, I'm going to get a human to do it and a professional to do it, because it's a different form and it's a different kind of thing, and so I'm trying to get across a different message. So I think we'll just end up where we have different tools for different things, but I think there'll always be the human element in there somewhere. But that's going to become, you know, that's going to become the super deluxe, the really expensive, you know, luxury, the luxury tool?

33:33 - Colin Gray (Guest)

Yeah, potentially. Potentially, I wonder. I mean I suppose there was already. Was there already a bit of a flight to quality over the last 10 years? Let's say, like um, sub-stack is a really good example. I think like for years there's been free newsletters, free content on just about any topic you can think of, and yet sub-stack newsletters have grown and grown. There's a there's a growing cadre of people who make a decent bit of money out of just charging you know, five or 10 a month to um for people to subscribe to newsletters, and you don't even need that big a subscriber base to make a decent living out of that as an individual creator. So I suppose we've shown that there's a not insignificant number of people who are willing to pay a smallish amount, even a big amount, I suppose, for quality content, and I suppose the hope is that that continues even if we're all drowned in free information from all angles?

34:29 - David Brown (Host)

Yeah, I mean I, I think it will. Um, and I know there's some tools for podcasts that are coming out with it that allow people to make micro payments as well. So you know, you can offer to, you know to pay in in in some sort of satoshis or whatever digital content or uh currency, to say, you know, if I listen to the creatives they are podcast I'll give them, you know, I don't know a pound for every episode that I listen to or whatever, and then it just sort of you know, it will, it will do the payment in the background and all that stuff is coming and I think people are.


I think I think people are willing to pay for it, at least at the minute, because there is still a difference in it. I I worry that at some point, you know, five years down the road, if the AI content is really going to be as good as most humans can do, then it gets a bit wooly. But I think we're.


I think the thing about the human content and this goes back to social media, but in a second I'll I'll link it in but the thing is is that people are crazy and that's what's fun about us, right? Like we do crazy, random things that an AI would just never do. Because it does it, it's the supreme of working at averages, right. It's taking all of human behavior, putting it into a bell curve, and then it's basically taking all that stuff in the middle. That's the most predictable, and then it throws a little bit of variety in on purpose, but other than that it's it's always in that.

35:59 - Colin Gray (Guest)

you know it's always in the middle.

36:02 - David Brown (Host)

And so you need the humans to to provide the crazy on either end, and that's what used to be fun about social media is before we had all the bots, before we had all the marketing, before we had all the ads, it was, you know, Twitter. When it first came out, it was like it was a bunch of people and they were all crazy. And that's what was fun about Twitter is because you could go see just people being nuts all over the place, but you knew it was a person, Because now you don't know what it is and you don't know where they come from and you don't know if it's real or not, and so it's kind of lost its luster because of that. And it's again I keep using this phrase today but it's, it's like a race to the bottom, almost so like I don't even enjoy it anymore. It's, it's not fun. I don't.


I like Instagram on a personal level, Because it's pictures and so you know, and short videos and stuff. I don't do tiktok, so I do Instagram reels instead, which is basically the same thing, but it's a visual thing and I quite like that, but I don't know, I just yeah.

37:09 - Colin Gray (Guest)

That's really interesting. Actually there's there's a, there's a concept I teach in our podcasting, all of the podcasting courses we do. It's one of the one of the things I think is it's almost the most simple but also the most. But it seems to hit really a nerve with just about everyone that comes through all of the stuff that we teach and it's it's just the Live recording mindset I call it, whereby when you're in your first ten episodes of podcasting, just pretend you're going live, don't edit, commit to not editing at all, so you just hit record and you just go as if you're doing a webinar or you know.

37:44 - David Brown (Host)

Yeah, okay, instagram live or something like that.

37:46 - Colin Gray (Guest)

Yeah, and you just go with the flaws and you're just totally open and honest and you trip over your words and you mess things up and you say um and I and all that stuff, but you put it out there anyway. And there's two reasons why we do this. One is one is that it makes it really easy to meet your first ten episodes. It makes you Get them out there and you improve so fast as a presenter During those ten episodes because you know you don't have editing as a crutch to rely on. You know that you have to listen back and improve those arms and those eyes and those crutch words and those slips and all that kind of stuff. You get good and fast. So, yeah, it's a great way to get it's. It's harsh.


But it's a great way to get good and fast. It makes it easy, so you get, you don't have to do editing, you get out, you get past that kind of initial barrier of just kind of learning what the heck you're doing. But the other part is the kind of secondary benefit is kind of what you're getting at there, which is it's just human. It's like you connect with people because you hear them making mistakes and you. It is utterly possible and I've heard it done so many times to Sanitize a recording, to sanitize a bit of media by taking out all of the mistakes and too many ums and as and you just sound too perfect. Yeah, and maybe that's it. You're gonna sound too much like an AI or a robot or something like that.

39:00 - David Brown (Host)

Well, you then start to sound like a broadcaster.

39:03 - Colin Gray (Guest)

Yeah, I think that's the difference is you.

39:06 - David Brown (Host)

You start to get that when it's, I would say, over edited, which I was, I've been guilty of, and I did exactly the opposite when I started. But in my first five episodes I had a co-host Name didi and didi was excellent because she used to do radio show and stuff when she was in Spain, so she did that for several years, so she was an experienced broadcaster and she was able to give me feedback and and some support and things on the, you know, on the back end, while I was learning a little bit. But yeah, I did exactly the opposite of what you recommend, you know, and I spent hours and hours and hours.


Like like three. It took me three times as long as the recording. So if I had an hour long recording, it took me three hours to edit the episode.

39:52 - Colin Gray (Guest)

That's so common, and that's exactly why I talk about this so much because that is the biggest killer of podcasts. It's the biggest killer of so many of YouTube channels, of blogs, of anything, anything creative. That is what kills it is you spend in too much time, not you the broader.


Yes yeah, spend in too much time editing, cutting, like, worrying about making it perfect, and you never get past that. First, you know five, ten, fifteen, twenty releases, which gets you to finding your voice, gets you to figure out what you're actually going to talk about. It gets you to a few listeners that you can then ask, like, what do you actually like about this, get some feedback. All of those things require you to do 20 episodes, get some reps and and actually do it. But the over editing, the perfectionism, it kills that. It stops you getting there.

40:37 - David Brown (Host)

So, yeah, that's totally it's exactly why we talk about it. And on that I have a question, and I've wondered this in your experience because you've you've seen a lot of people go through this process. My working theory, my personal working theory, is is that the older someone is when they start, the more of the that perfectionist that they feel they need to be, whereas the the younger people who kind of grew up with mobiles and very easy access to record stuff on a mobile phone, say, are much less worried about that. Is that something that you see or am I? Yeah, absolutely no.

41:14 - Colin Gray (Guest)

There's definitely a trend that way for sure, maybe, maybe just the fact that you know is, even when I grew up like you, know that you didn't have rough media. Every bit of media you consumed was a TV program, it was a film, like you said, it was a broadcast run a radio station. There was no rough ready media, whereas, yeah, so much of it now is. It's just like that. But I don't think it's entirely. I think even the younger people we work with like even stuff 18 to 20 year olds they still have. They still often do have a big bit of perfectionism in them. They're willing to be more rough but they still they still have a lot of perfectionism in them. They're willing to be more rough but they still then want to put too much into it sometimes in terms of the time to edit and all that stuff. So it's not been eradicated entirely, but you're absolutely right I think in my experience it definitely we tend more that way.


Yeah because that's that. I think that's when I started to dig into it just within myself.

42:02 - David Brown (Host)

I, you know, thought about it a lot. Sorry for the background noise, but I I thought about it a lot and I think it's because I was so used to the only time we ever saw anybody was because it was on TV or it was on the radio and so I felt like the standard was so high that I have to meet that standard and that's what was almost holding me back and I wanted to do some live shows and I wanted I really want to do video and I worked out that I'm not very good at video editing. I don't know how to video edit although it's something I'm learning but if I want to do a video podcast, I just need to take the video that I've got and I can't edit it because I hate jump cuts.

42:50 - Colin Gray (Guest)

So you just have to go wrong.

42:52 - David Brown (Host)

I just, I just have to put it out there. So I've really really tried the Ums and Oz I've tried to get out of my speech. I have a few other. I call them ticks. I have a few other ticks that I say a lot that I've noticed. So I've replaced the Ums and Oz with something else that now annoys me to death. So I have to work on that. But it's, it's something for me to constantly work on.


But I've got to the point now the last few episodes where I don't know if it's. I've just become comfortable with the fact that when I say something boneheaded or I make a mistake or like I'm talking quite, I feel like I'm talking quite a lot in this episode, but, and that kind of bugs me.

43:34 - Colin Gray (Guest)

But I just gotta go with it, right.

43:39 - David Brown (Host)

But anyway, I know what you mean and and I can feel it and I've I've relaxed into it and, like you said, I've tried to find my own voice and I'm a little bit more comfortable with my voice, not just hearing my voice, but my voice, and that makes it easier as well. So I don't feel like I have to edit as much out. I do very minimal editing these days. I usually, you know, you'll know we had a conversation that was being recorded at the beginning that I'll cut off and I might, you know, chop a bit off at the end. But basically it will go pretty much, as is the, the audio, when I might tidy up a little bit. But you know, the video one, if I, if I put up the video, I'll just put it up. I mean, you know drinking, looking around, you know being stupid in my side, I had like, whatever, it doesn't matter.

44:25 - Colin Gray (Guest)

I think there's a funny, there's a, there's a what do you call it? A barbell shape of payoff in that, in the editing, in that there's real strength at either end and nothing in the middle. So people should either, I think, in my opinion, not edit at all, like just put out raw just like you're talking about like just a trim start and end.


Make sure the star is like just starts with the voice. There's none of that rustling a paper at the beginning and same at the end and that is it. Put it out or do quite a lot of editing. But to the content itself. So take an hour long conversation, like we're having just now, and cut out half of it, pick the best 30 minutes from that hour and input that.

45:06 - David Brown (Host)

I can't do that.

45:07 - Colin Gray (Guest)

Someone else will have to do that. It takes so much time. It's like it's really hard. I've done that with many a show like we used to create highly produced stuff like that and it's you take an hour recording and it takes a couple of days. It's not just like three times that you're talking about.


It's a couple of days work. You get the transcript you, you spend like a couple of hours reading it. You start to cut things out, you start messing around with the order, all that kind of stuff, and you know it leads to something significantly better it does, because you cut out the fluff and you can get the real highlights in there. But for most of us I don't think it's worth the time now because actually the completely raw one is still good. But like, the point I suppose I'm trying to make is there's a bit in the middle where people do spend a few hours or half a day just cutting out arms and eyes and cutting out a few extra sentences and it barely improves it at all. There's a little bit of a quality improvement, but not by much. But it still adds so much time to the process. Is it's a funny one? Let's add net.


This is one area of the audio editing point where I do wonder if I will get to in a few years. I've played around with this with a few different LLM. That's like chucking in a transcript and saying, cut this down to 30 minutes, and usually it's just not very good. Occasionally I've seen it hit actually a pretty decent edit. It's similar with highlights, so you know, like automatic clipping tools like we're talking about, we're recording in the river side, it's got an AI clipping tool.


It does. Yeah, find the hit rate on that Not riverside in particular, but all of them is just really poor as well. It feels like Go on, go on.

46:40 - David Brown (Host)

No, it's good. Sorry On the riverside, when it was, I think when it very first came out, it was 50, 50 you think. I think it's much better now.

46:51 - Colin Gray (Guest)

Okay, which.

46:52 - David Brown (Host)

I find really interesting. But I think what they're doing and I'm and I'm not sure, I don't think they're using the content as such, right. So it's just a first of all, before anybody jumps on either one of us, I have no idea how it works, I'm just guessing. But if I were to guess, I I suspect what they're doing is they say okay, here was a recommended clip. Did they use that clip or not? Yes or no? And that's how they're reinforcing it. So so they're just. You know, the model is just learning. Did it use this response or not? So, was this a good selection or not? And and I I suspect that the model is training itself in the back it's not actually using that content in any way to do. You know what I mean. So it's it's not taking somebody's copyrighted content and using that content. It's just saying was that a good selection or not?

47:42 - Colin Gray (Guest)

Yeah, it's starting to train itself and it's training clips.

47:44 - David Brown (Host)

Yeah, and it may even do it on a user by user basis. So it could Say you know the creators the iPod cast or you know my account. He likes these types of things, and so we'll start doing more. So I don't know, but it seems to be getting better than it was in the beginning it was. It was quite woolly in the beginning, but yeah, it seems to be pretty good.

48:05 - Colin Gray (Guest)

Now I have to say I found it. Yeah, I suppose it depends on the type of content. Maybe it, maybe it's an indication of maybe in crap at short sharp sound bites. But I found a hit rate was more like One in a hundred. I really struggled. Yeah, like I've. I've tried them all because they're it's such a kind of In-demand tool for podcasters to be able to repurpose your audio, to be able to get out there.


Podcasts loads of strengths, but one of its weaknesses is it's hard to share, you know. So clips are actually a really powerful thing for podcasters. So, yeah, tried them all and they just I've never found any that can pick out more than you know. One clip out of every. Yeah, it's really bad I've found, but I'm sure that is. You're absolutely right, that's someone. It's gonna improve a lot.


I think so years but I also have a kind of an optimistic hope stroke, feeling that it might be one of those things that is really hard to solve. It might be one of those kind of last 1% that takes 99% of the time, just because it's such a Subjective thing finding a little clip that's really gonna capture people's imagination, it's really gonna hook people in. That might be the really human thing that actually is really hard to replicate, and I and and that's where the personality of a show comes out, isn't it?

49:20 - David Brown (Host)

It's Each show and each host is gonna be different. We're gonna pull out different, like, if you and I look at this, we might each pull out different things because of who we are in our background and whatever so, and and that's the human factor that that comes into it, kendall shout out to Kendall at Riverside. She had a really good suggestion the other day when creating clips, which is sometimes she'll just take the questions and and and clip all the questions together into a short, like 30 second clip. So you know, when you ask somebody, you know five or six questions or whatever, just all those little questions that come up.


Yeah as a teaser to say look, here's all the questions that I asked and if you want to get the answers, you're gonna have to listen to the episode and I thought that was quite clever. I haven't tried it myself yet, but I thought that that was an interesting way to go through and to sort of you know, Try and try and find content, but yeah, for sure, now earlier.


Cool, I'm conscious of time. We're sort of 53 minutes already, so we're we're running short, but I Remembered the question. So, for everybody who wasn't part of the conversation, I have a new question that I want to add into the the list of questions I ask at the end, and it's this I remembered what. It was great. Do we have to be nice to AI, or should we be?

50:44 - Colin Gray (Guest)

nice to AI. I mean, I do. I have to admit I've noticed myself doing this and I've not tried to talk myself out of it. I say, please and thank you and stuff like that when I'm talking to to any LLM I'm using and I think, yeah, I mean, I think it's just, it's just part of being a human that we should when you're having a conversation. If we want these things to stay useful, we need to treat them actually the way you'd want to be treated. All that kind of side. There's some kind of weird ethical, moral things in there. I think that if we're gonna treat them as a real assistant, then yeah, why not? Yeah, there's no harm, certainly is there?

51:17 - David Brown (Host)

well, no, and it was interesting to see. So I sort of saw a threat on this and it seemed to be very, very much 50-50. So a lot of people are like yeah, like why do I need to say please or thank you? Because it's a computer, it doesn't care, and it was all it was. It was all based off a comment when a lady said that when she uses it to ask to help with code, she said she feels gratitude like she would with somebody that helped her, but she doesn't know what to do with the gratitude Because she knows that the computer doesn't care.


And I was like, wow, that's a really interesting Way to think about it, because I never thought about that. But that is what it is. When, when you ask, you know, when you use AI as a tool and you ask it to help you with something and it gives you what you need and it gives you that answer, you do you feel like this, this weird sort of sense of gratitude, like you would to somebody if they helped. You need be like, wow, that's really amazing, they helped me with that, but you don't know what to do with it. Like Can. Then you get this funny feeling. So yeah, it was. I thought that was expressed in a really interesting way, and it's interesting.

52:26 - Colin Gray (Guest)

Yeah, I mean, I'm with you.

52:27 - David Brown (Host)

I say please and thank you, because when it takes over someday I expected it'll go. Yeah, that guy was always nice, there is it.

52:36 - Colin Gray (Guest)

There is a tiny bit of me that thinks that.

52:42 - David Brown (Host)

So another thing I'm always curious about is do you, is AI male or female?

52:48 - Colin Gray (Guest)

for you. Oh Well, I have set my Siri to be Irish and female, and I don't know what that says about me, but that's, that's who my Siri Assistant is, certainly.

53:02 - David Brown (Host)

It's the Irish one is about as close to sort of a Scottish accent as you'll get, because there's probably not a Scottish one, is there you do?

53:09 - Colin Gray (Guest)

actually I think you do get a Scottish one, but I was like no, it's too grating. So the Irish little softer, okay, yeah, yeah and but no, I think I probably actually think of, like, when I'm chatting to, when I'm getting help from an LLM or something like that, I'm probably thinking about the male, just because, you know, we always tend to identify with ourselves, don't we? So?

53:28 - David Brown (Host)

interesting. Okay, and if, when you have a personal assistant that takes care of your life for you, what are you gonna name it?

53:37 - Colin Gray (Guest)

Good question, I'm not even sure. I mean, I think, I think, yeah, I'll just call it pal. That's what I call everyone I like, right, okay. That's a good answer actually give me some hand, give me some help with this pal.

53:54 - David Brown (Host)

Yeah, mate fella something.

53:56 - Colin Gray (Guest)

Yeah, always.

53:58 - David Brown (Host)

I've asked these two questions a lot and I'm yeah. I'm always really interested in the answers and and more from the psychology, of how people think about it, rather than what their answer is, and I think Female wins out. I think it's about 60, 40, hmm, and it doesn't matter what gender the person is that I ask.

54:19 - Colin Gray (Guest)

Is that right really?

54:20 - David Brown (Host)

so it doesn't seem to skew along gender lines, but I I suspect that it's mainly because so many of the sci-fi films and everything are all female. They often, and it's it's sort of ingrained that in our head all the way back from Star Trek and kind of the 50s or the 60s. Whenever you know Star Trek very first came out, it was a female voice and I think that's just put that in our brain and and, of course, all the male ones are evil, so what I was about to say.

54:51 - Colin Gray (Guest)

Maybe it goes back to that, you know when, that when it takes over the world as well, a well-erkin optimistic view is that hopefully it's more female inclined, as in it's kind of less prone to idiocy, like us males, and we'll treat us decently.

55:07 - David Brown (Host)

I was, I was, I was about to tread on very, very sketchy territory there, but Do you know I?

55:15 - Colin Gray (Guest)

had another thought about that, about you thinking about being kind to it as well. There's something around like I actually kind of believe the whole concept around your, your true nature, is who you are when nobody else is watching. Yeah, you know, would you, would you steal something if there was no chance you'd be caught? Or would you, you know, do this unkind thing or evil thing if you knew there was? Nobody would ever see it. And there's something around that as well, like I kind of I Really value courtesy and respect and all that kind of stuff. So I think, even with something like an AI, I don't know, maybe that kind of I feel like, yeah, even if it doesn't matter, even if it doesn't care, even if no one's ever gonna see, it's still worth doing just for that.

55:55 - David Brown (Host)

That's a great point. I love it. I like that question. I'm gonna definitely use that question maybe forward and see what people think on it. Awesome. Thank you very much. Before you go, I did a little intro at the beginning about your, your two different things that you're working on, but maybe if you give everybody just a quick, a Quick description and shout out to the stuff that you're working on at the moment and where they can find it oh yeah, I appreciate that, yeah, we.

56:22 - Colin Gray (Guest)

So we do all over writing and our content creation over at the podcast host calm. That's the podcast host calm and if you are thinking about starting a podcast and we've got a big, massive how to start guide there, which is under the podcast host calm for slash start. And yeah, I'll a to as our tool, like I said, it's got a good bit of AI built in in terms of the cleanup, does all of the the noise reduction, leveling, all that stuff automatically helps add music in the theme music. It helps you with your editing. We've got transcription based editing, like editing by a text, and we've got, you know, show notes, writing, description writing titles, episode titles, all that kind of stuff. So yeah, it's basically a designed to be an all-in-one podcast creator. So if you fancy creating a podcast over there, it's all a to comm a Litu and this is no joke.

57:10 - David Brown (Host)

I totally wish I'd seen that when I very first started, because it probably would have made my life so much easier. I'm sort of developed all my own workflows and stuff now. Yeah and, and and do more specialist stuff. But, colin, thank you very much.

57:28 - Colin Gray (Guest)

No, most welcome.

57:28 - David Brown (Host)

It's been great to have a chat about all really interesting yeah, so thanks for having me on and let me know if there's ever anything I can do to help you, and Until then, have a great Christmas and a happy holiday, and we will see you soon. Bye, bye.

About the Podcast

Show artwork for Creatives With AI
Creatives With AI
The spiritual home of creatives curious about AI and its role in their future

About your host

Profile picture for David Brown

David Brown

A technology entrepreneur with over 25 years' experience in corporate enterprise, working with public sector organisations and startups in the technology, digital media, data analytics, and adtech industries. I am deeply passionate about transforming innovative technology into commercial opportunities, ensuring my customers succeed using innovative, data-driven decision-making tools.

I'm a keen believer that the best way to become successful is to help others be successful. Success is not a zero-sum game; I believe what goes around comes around.

I enjoy seeing success β€” whether it’s yours or mine β€” so send me a message if there's anything I can do to help you.