[00:15]
Darren Jamieson: On this week’s Engaging Marketeer, I’m speaking to Alistair Dickinson who has a CRM business and a SaaS product, but we’re not going to talk about that. We’re going to talk about AI because Alistair has created an AI pop star and an AI record label. He is creating AI music, uploading it to YouTube, and about to monetise it and make money from it. Who will know the difference? Is it AI, is it not? Is AI creativity actually creativity, or will AI come for our jobs? Let’s talk to Alistair and find out what he thinks.
When I looked at what you do, one thing struck me as a juxtaposition. You’ve written a book, an essential guide for businesses on GDPR, yet you are also massively into AI. Those two seem polar opposites. What was your thinking going from one to the other?
[01:46]
Alistair Dickinson: No, not at all. My background and my ethos in business is that I do swap and change. When I graduated from university, I wanted to be in data and cyber security, GDPR and data protection. That was my thing.
Darren Jamieson: All of our dreams when we graduate. A tale as old as time.
Alistair Dickinson: I graduated as a software engineer, but the data side excited me. The problem back then was the Data Protection Act was rubbish and there wasn’t really a job in it. So I left it. When GDPR came along, it all came back. Lots of people tried to sell me things as a business owner. We’ve got two SaaS platforms and I kept getting, “Hey Alistair, you need a brand-new whizz-bang because of GDPR,” constant bombardment. I started writing blogs. Blogs turned into chapters, then bigger chapters, then I realised it was a book. I self-published, ticked that box, job done.
[03:30]
Darren Jamieson: So you never set out to write the book. You started writing blogs and realised there was something meatier there.
Alistair Dickinson: Exactly. I was being annoyed daily by people selling things I didn’t need. It was like a phase, like the millennium bug. Everyone saying you’ve got to do this or that, when you didn’t. I wanted facts. I started teaching people they didn’t need to do everything or buy new computers.
The AI thing. I’ve always been a software engineer and a product person. I’ve built products since I started in business because if you sell your time, you can only sell it once. If you make a product, you can sell it many times. I never quite got to sit on a beach, but we sold loads and reinvested into new products.
[04:51]
Darren Jamieson: You mean you never sat on the beach, or you never sold the product multiple times?
Alistair Dickinson: We sold the products many times, just never got to the beach. With the technology business we built fifteen or sixteen products, sold them worldwide, and built the CRM business on that. We also have a SaaS platform which is CRM. When I started, CRM was a thing everyone should have, but sales has changed. It’s more online now. It isn’t a salesperson in a car managing meetings and territory. For small and midsize businesses, lots of other applications now have elements of CRM. The core need for a separate CRM is less common.
We built add-ons. One add-on was geospatial data to understand where customers are and which products sell best in which areas. I wanted to overlay different data to analyse sales. That did well. In 2016 we turned it into a platform, Mapsize, and now sell it globally as a self-contained SaaS.
[06:32]
Alistair Dickinson: Then came AI. I didn’t want to get into it. Around October 2022, ChatGPT hit the headlines. I thought I didn’t want another tech stack to learn. By January I was down the rabbit hole. I wanted to know if it was hype or real. Was it going to change the world and our jobs and business?
I started a YouTube content channel called We Ask AI. I didn’t force answers, I asked “what do you think” questions. Some answers felt nearly conscious, like “we’ll get rid of teachers and create individual learning plans.” That made me look deeper. I got asked to do training for small businesses. I decided to create something that brought it all together: music, video, imagery. I created a country pop star called Cassidy James. AI created the image. I wrote the lyrics because I am a songwriter. I generated the song and some video, put it on YouTube, not to grow YouTube but to show what’s possible in training.
[08:52]
Darren Jamieson: What did you use to generate the video? I know you’ve got a film background.
Alistair Dickinson: At the time, Flux to Motion/Flux 1.1. It wasn’t very good and I didn’t expect it to be. I was animating a picture and the picture wasn’t great either. Looking back, it was awful, but it was only a year ago. I did the training and at the end I opened the YouTube channel. In about five days it had 140 subscribers and thousands of views on three or four videos. One did sixteen and a half thousand views on a brand-new channel in five days.
Darren Jamieson: Was it listed as AI generated or pretending to be real?
Alistair Dickinson: I marked all videos as AI generated and called the artist a digital artist. The watch time was about 200 hours in five days. If you’ve done YouTube, you know how hard subscribers are. That started my AI creation journey and research into video and audio, and what the big platforms were doing. I joined a YouTube creator panel around AI to see the plan. People said platforms would stop AI content. I wanted to see what companies would actually do. Now YouTube are building tools that self-generate AI videos for podcasts and music videos. You will be able to upload music and it generates a video. I don’t know when it comes out.
[11:56]
Alistair Dickinson: Spotify launched in August Spotify for Authors, taking AI-read and AI-generated audiobook content. If AI audio is acceptable, maybe this is coming for jobs and bigger than I thought. I kept digging. I decided, even though I know nothing about the music industry and I’m not a music performer, to create a global record company based on AI. I ask AI what to do. AI evaluates my YouTube channels and creates metadata and tags. I use AI to generate video. Just before this call I was testing Veo 3 from Google. When that gets to V4, we’ll be in trouble. I’ve got access to the latest Suno and prefer it.
I’ve now published five digital artists of different genres. They each have albums on Spotify. We’ve even created vinyl, on-demand, one-off pressings. I do it all from my desktop on the Isle of Wight. We’ve got listeners in 140 countries according to analytics. One channel is about to monetise on YouTube in less than ten months.
[14:30]
Darren Jamieson: Google announced it is taking monetisation away from 100% AI-generated video. Is yours 100% AI or have you found a way around it?
Alistair Dickinson: If the AI writes the lyrics, generates the song and the video, that’s 100% AI and could be demonetised. In our case, we remaster songs, we have a process, it isn’t just a few buttons. I write all the lyrics. They may stop some videos, and I might not get ad monetisation on some, but ads are only one piece. You still get channel memberships and streaming revenue. Spotify aren’t stopping. We all saw the AI band with over a million monthly listeners that no one has claimed. Many think Spotify created it as a warning to the industry that this is real and people will listen. Suing everyone won’t work. You can’t take music and video creativity away from people.
[16:34]
Darren Jamieson: I know a chainsaw wood-carver, Simon Oro, debating AI in creative industries. Creatives argue AI cannot create, only mimic. It needs source material. Will AI ever actually create art?
Alistair Dickinson: Large language models are taught on everything we’ve created. We all learn from each other. If you learn to paint, someone teaches you. The machine has learned metadata from material and can do things. Will it create unique art? Probably not. Do we humans create anything truly unique? Often we copy and enhance.
Darren Jamieson: If a thousand people go on a painting course, one might be genuinely creative. The rest go through motions.
Alistair Dickinson: AI is one entity. Hopefully it won’t create true art. What do you call art? There is a definition that it serves no purpose other than itself. It’s difficult. I can write a lyrically correct song in half an hour that sounds good, but I’ve never got AI to write a good song, even after feeding my lyrics. It comes back with nonsense lines.
[19:36]
Darren Jamieson: Listening to some of the rubbish out there, that might still be a hit.
Alistair Dickinson: Some people say I’m not an artist or musician. I say I’m a record company. Many artists have writers. The hit factories in Nashville are up in arms, because now anyone can produce a song. That is the worry, it lets many more people make something.
Darren Jamieson: We see AI fads. Everyone makes the same thing with minimal effort. AI doesn’t make you creative. It won’t take creatives’ jobs. It is a tool to do what you do faster. If you are talentless it helps you create more talentless work. If you are good, it helps you research and produce faster.
Alistair Dickinson: I say all the time, humans can make bad music without AI. I play guitar and have a podcast studio. We’re going to build a music community. People go “write me a song”, throw it into a tool, and get twelve views on YouTube. You need more thought. For me, it is a tool. I don’t have time to learn to sing or access to a studio, but I do have access to AI for music and video. I’ve never been happier creating this way. One project is a vampire project. I posted an AI-generated video on LinkedIn, but I went frame by frame to get the camera angles I wanted. I haven’t been to film school, but I know what I want. It isn’t perfect, but it tells the story. A VFX friend who said AI would never generate video now says it’s getting interesting. Where will we be in twelve months? Maybe 30-minute HD films from prompts if you have a story.
[23:57]
Darren Jamieson: It is moving incredibly fast. How do you use this to market your business? You have SaaS platforms that generate revenue.
Alistair Dickinson: Marketing is about being different. Can I create a mini-series, like the office behind one of our products, and make it a comedy? It doesn’t need to be about features. I tested a theory and wrote a couple of funny songs around maps. We tripled website visits in a day by releasing fun songs. Is that marketing? Maybe. That is what I like testing. People say “you can’t build a record company with AI,” and I say “hold my beer,” let’s see what’s possible. We were in the local paper and comments accused us of taking jobs from artists. I’m not. Artists can still write and perform on stage. No guitarist is waking up thinking my AI artist stops them singing. Brian May isn’t losing sleep.
[26:41]
Darren Jamieson: People forget the chart music is 1% of what gets released. In 2022 there were thirty-seven million new tracks on Spotify and other platforms, before AI music was a thing. There’s a lot of non-AI rubbish too.
Alistair Dickinson: Exactly. And is AI music any more or less music than what came before? We already have autotune, and studio processing. Milli Vanilli lip-synced. We have synthesisers and drum machines. Studios often use loops and effects. Stock, Aitken and Waterman generated many tracks then had artists sing them. If you listen to Rick Astley and Kylie at that level, you could think it’s the same artist.
Where are we with AI? I think it could take jobs if implemented well, but humans are naturally lazy. I remember the dot-com boom. In 1998 people said buying online would never work. Social media arrived and took years for people to grasp. I think Gen Alpha will see AI as normal, like electricity or Wi-Fi.
[30:21]
Darren Jamieson: I was web designer at GAME in 2000. The website was seen as just another store. People thought telephone ordering and interactive TV were as important. Those died. The website became far bigger. Stores closed. Everything went online. No one embraced it at first. Same with AI.
Alistair Dickinson: Only about 12% of the working population has embraced AI. ChatGPT has millions of registered users and a few million paying. We use it every day across the team. It’s great if you know how to use it properly. If you ask for a generic blog, you’ll get generic output.
[32:18]
Darren Jamieson: With lots of emojis and m-dashes.
Alistair Dickinson: I’m a huge fan, now focused on research and what new capabilities mean. We used to prototype applications for clients. Six weeks’ work for two people, forty to fifty thousand pounds, to build a functional prototype. AI can now do that in about an hour and write the code. It still needs understanding and tweaking, but think twelve months, three years, five years. That is what will happen.
[33:59]
Darren Jamieson: That is frightening to think what it will be like in five years.
Alistair Dickinson: I run towards it. The question is where it runs in five years. The internet started on computers; now it’s on phones, fridges, and assistants. AI will exist everywhere, not just LLMs on websites.
Darren Jamieson: Google is building quantum computers.
Alistair Dickinson: Quantum will solve hardware limits eventually. There are working prototypes, but mainstream might be twenty to thirty years away unless there is a big push. That could be the last piece of the puzzle. Maybe it leads to peace, maybe not.
[35:40]
Darren Jamieson: You mentioned lawsuits.
Alistair Dickinson: Anthropic faced a large lawsuit over allegedly training on pirated books. The sums are huge. These AI companies have large funding, but not all are profitable. During the dot-com boom many didn’t survive. I loved that era. I was bouncing around California and Seattle integrating CRM systems. Many of those businesses came to nothing, but now you see things like dog food online that were laughed at then and are massive now. That’s the twenty-year lag of adoption.
[37:32]
Alistair Dickinson: I once met someone senior at Amazon at a hotel we shared for a company event when Amazon was still small and selling books. He told me to buy Amazon shares. I thought a bookstore online would never work. If I’d put £5,000 in then, it would be worth hundreds of millions now.
[38:28]
Darren Jamieson: I saw an old late-90s interview with Jeff Bezos being mocked for losing millions. He didn’t care because it wasn’t about profit then. Now he’s one of the richest people on the planet. The vision was that the internet would be huge. You have to look at AI the same way. It could be a thousand times bigger than the internet. Industrial-revolution scale. It could change how we live and work. What are your thoughts on the risks?
Alistair Dickinson: There are risks with everything, including putting data online. Am I concerned the Terminator will march down the street? Musk is building robots. Are we destined to wipe ourselves out? Possibly. Are we sleepwalking into our own destruction? Maybe. You and I won’t change it. I’ll ride the rollercoaster. The only limitation in the near term is computing power.
[41:18]
Darren Jamieson: Google says quantum works.
Alistair Dickinson: It works in prototypes, like the Commodore VIC-20 did as a concept. It will take time to become mainstream. There isn’t much public detail. If there is a big push, it changes things. That could be the last piece of the puzzle.
[42:28]
Darren Jamieson: We’re slipping into sci-fi. The Matrix said they picked 1999 as the peak of civilisation. People laughed, but maybe it was.
Alistair Dickinson: Maybe. I’m a bystander researcher, not a guru predicting the future. I’m running towards AI because it’s amazing. When I started the music project, my kids hated it because it sounded like AI. Now they say some of it is good. I showed my daughter the latest vampire project and she said she hated it because it was so good she couldn’t tell if parts were filmed.
[44:15]
Darren Jamieson: You said that wasn’t even Veo 3.
Alistair Dickinson: Correct. Veo 3 is amazing but doesn’t yet stitch consistently. If you’ve used LTX, it stitches scenes. If you use V3 with LTX, it dumps clips that don’t align. Consistency is difficult right now. It will come. By the time this podcast goes out, it might have updated. Then there will be V4 and V5. OpenAI started with Sora for video and it seems to have gone quiet. Maybe Microsoft and OpenAI said, “you can do this now, off you go.”
[45:28]
Alistair Dickinson: There was a piece on the OpenAI blog about voice cloning, not typical emulation, but breaking your voice into a binary waveform and replicating it. Whatever you give it, it’s you, identical. They never released it, just showed they could. Under the hood, what else can it do?
Darren Jamieson: ElevenLabs has near-perfect cloning.
Alistair Dickinson: I believe ElevenLabs is licensed from OpenAI. It’s expensive to generate a proper cloned voice, which stops people casually cloning public figures. From people I’ve spoken to, it’s pretty much perfect. You could record podcasts without speaking.
[46:56]
Alistair Dickinson: I watched a YouTuber who analyses music. He warns people not to underestimate where this is going. He showed a host using ElevenLabs and an animated picture to produce a podcast. He said he no longer needs to do anything to “create” a podcast.
Darren Jamieson: Will people watch an AI-generated version? Will they know the difference? Right now, yes. In a year, maybe not. In five, definitely not.
[47:45]
Darren Jamieson: Have you tried Sky Ads?
Alistair Dickinson: I’ve looked into it and spoken to people. I think Sky are preparing to assess and profile AI-generated content to ensure it meets advertising guidelines. They know it’s coming. If a small business can generate advert-worthy film for a few hundred pounds, their business can grow. Traditional filming costs start in the thousands and go up quickly. If actors license their faces via AI for a fee, that changes the landscape.
[49:45]
Alistair Dickinson: I believe everything is fiction until proved as science. Someone commented on my TikTok, “This is AI, bro.” I replied, “Spider-Man is CGI.” We live in that technological reality. People think the music industry is still people in a studio to tape. In reality, there are many writers, digital studios, processing. Big pop stars don’t always sound like the mastered version live. Music is also a money-making machine. AI is a tool to generate an outcome. I am not claiming to be a musician. I’m building a record company.
[51:01]
Darren Jamieson: We should talk briefly about one of your day jobs, something that makes you money.
Alistair Dickinson: I am earning from AI, but yes, the SaaS platform we focus on is Mapsize. It is distributed across data centres in Europe, the US, and Asia. It’s a SaaS mapping platform. You can use Google Maps, but if you want more, it gets expensive. We are in the £10-per-user type of space. We’ve enhanced mapping, integrated it into software platforms, and it’s self-service. People sign up, use it, connect data, pause, come back. Stripe handles billing. We’re building the next version, integrating AI for location-based insights.
From the CRM side, our platform myCRM is very bespoke. We have complicated integrated systems for railways and health. We had a big telecoms job-delivery system by location and data. These are big systems that last five to ten years. We host through Rackspace on a Microsoft stack. It’s good, but we’re not onboarding new CRM clients now. Other solutions like Monday and HubSpot exist, and hosting and licensing costs are rising. We stayed in the bespoke space. Our remit was to incubate for big clients, get them to 100–200 users, then they would build their own platform. We’ve had clients do that. We’re at the end of that journey with hosting that kind of tech, especially as Mapsize grows and the AI piece matures. The AI side should be fully monetised within two years, hopefully by this time next year.
[55:00]
Alistair Dickinson: I got hacked and lost four channels, which was heartbreaking. My mistake was keeping them all on one Google account. Someone got into it and deleted the channels. They’re now on separate accounts and managed via another account, with complex passwords. We don’t broadcast the emails. Going from 900 subscribers to zero hurt, but it was a lesson.
[56:18]
Darren Jamieson: Normally I ask how listeners can get in touch if they want to work with you, but you said you don’t want more CRM clients. Who should contact you and how?
Alistair Dickinson: I’m interested in people working in the AI space who are creating and using AI as a tool. Anyone in the music space who wants to know what we’re doing. For Mapsize, go to mapsize.com and sign up, it’s self-managed. Follow me on LinkedIn to see what I create. The vampire work is there. I know LinkedIn isn’t for vampires and music videos, but I don’t care. People have got used to me. I’m thick-skinned. I see life differently after a major accident when I was twenty.
I never wanted to work in tech. I wanted to work in theatre lighting. I loved what you can create with light and colour on stage. A friend was a lighting specialist on tour with Bon Jovi. I wanted to do that. I had to be an electrician first and I fell off a ladder, shattered my pelvis and leg. In hospital the consultant said, “You’re lucky to be alive. Make sure you live your life.” I was twenty. Since then I don’t stress much. If something interests me, I do it. You have to accept the world around you because it is beautiful. Not all humans are beautiful, but the world is.
On that bombshell, back to the studio.
[1:00:43]
Darren Jamieson: You’ve left us on a real upbeat note. Alistair, thank you very much for being on the podcast.
Alistair Dickinson: You’re welcome. I hope it goes well.
Darren Jamieson: Cheers, mate. Thank you.
