It took me a while to realise that I am a Vibe Coder, here’s the story.
I have been working with AI to code for 2 years now, before it became a meme. It just seemed an obvious way to use the tool.
Career Leading to Prompting
I should probably rewind a bit for background. I graduated in 1996 and spend a career working within corporate communications slash marketing departments. I was the buffer between the tech guys and the marketing folks. I hand-coded my first website in HTML for Comcast back in 1997 for some pocket money whilst at college. It kinda snowballed from there.
By my 20s I was managing complex websites working with teams of developers, designers, copywriters and translators for other multinationals. I loved the conceptualisation and building process but hated the corporate bullshit that came with working in these organisations. As a reaction I shifted into writing, I thought I hated making websites, but I just hated the environment I was doing it in. I worked as a journalist and an analyst in the emerging cannabis industry for over a decade and finally settled as a freelance corporate communications advisor, writer and photographer. It took a while before I started tinkering with websites again in my spare time as a hobby. Naturally AI found its way into my workflow.
AI Ate My Hamster Career
Try being a freelance writer in the age of AI. We were the first to fall, alongside the accountants.
The problem was not that I lost my work, the problem was I was charging by the hour and AI made me so efficient there was not enough work available to make enough hours to make in an income.
Initially I thought I could ride it out, if I embraced AI I could surf the wave, as I had done with the Internet 30 years ago or the cannabis industry 15 years ago. I used my writing, conceptual and managerial skills to write great prompts. As a result, the team I was in shrunk to just me. I was the guy they would ask to get AI to do stuff. In the end, I thought fuck it and just explained how to use AI – thus losing my final client, but freeing myself in the process.
My 2025 new year resolution was to never work for anyone else again and to only spend my time working on creative projects that would generate a passive income. I found one quickly that brought in enough to keeps things ticking over. It was using my photography skills shooting VR content. This gave me a boss-less income with spare time to spend on other projects.
I made that creative work as efficient as I could and got it down to a few hours a day. This freed me up to explore other avenues.
Building an iOS App with AI
Each day as I walked my dog in the local forest I would listen to podcasts. One of which is Rick Rubin’s Tetragrammaton. One episode in particular grabbed my attention, an interview with Jack Clark, CEO of Anthropic. In it they discussed using AI to code. To describe an idea and get AI to build it. It struck me that I was already doing this! I had built several simple websites using AI to build the HTML, CSS and javascript. Having a deep understanding of these language helped to keep things on track. Using AI sped things up immensely. I’d always wanted to make games. From my experience in building websites I could see how I could work with AI to guide me through the process of setting up the necessary environment and even writing the code.
Retro Solitaire Legacy
I’ve played computer or console games since I was a kid and have an affinity for retro games reminding me of the innocence of gaming back in the 80s and 90s. As I got older I became a big fan of solitaire. My grandmother had taught me how to play Klondike Solitaire with real cards when I was a kid.
Recently I would download and play solitaire, game after game on my phone, but could never find any that satisfied me. I found them over-engineered, full of spy marketing code or with exploitative business practices such as high monthly subscriptions. I thought they were ugly with too much going on. I wanted a game that stuck to the core principles of solitaire. The equivalent of having a pack of cards in your pocket. So I set about building it with Claude.
Nothing Else Mattered
I felt compelled to drop everything I was doing to work on this. After a few days the core gameplay was almost done. I thought I’d be finished in a couple of weeks. I did finish the game quickly but I wanted to set it apart from other games out there. This involved a complex backend to process the data. This turned into a monster, I spent months reworking the code and learning along the way how to make it as efficient as possible.
I realised that the big solitaire companies, Microsoft etc, have huge databases of solvable deals. I built an app to help me discover solvable deals but soon realised that I didn’t have the computational power to run through the billions of combinations of play to find enough solvable days to make it work. I learnt that there are more combinations of deals than there are atoms in the known universe! It is estimated that around 78-82%, roughly four out of five deals, are solvable.
Eureka! Harvest Player Brain Power
The whole point of a good game is that you don’t win it all the time! That’s what keeps bringing you back.
I realised that if I were to record the deals that people solved from a completely random deal then I could build up a database of human-proven solvable deals. This could then provide competition for people wanting to beat previous high scores (low move counts). Therefore crowdsourcing my computational power by harvesting player brainpower.
As an incentive, players needed to be rewarded for the games they solved. What better reward, than to name that deal after them? Like discovering a new species. Hence “Solitaire Legacy”, which then morphed into Solacy: Solitaire Legacy. Players are rewarded for finding solvable deals from the random deals that they play. The high score system for the solvable deals gives completed games replay value.
It’s now September 9th, I listened to Jack Clark’s interview by Rick Rubin less than an hour after it was released on the 4th June. The seed was planted. I stopped all other work around the end of June to focus purely on Solacy.
AI Hallucinations and Discovering Cursor
I was making progress using Claude to do my code, but there were limits. I would need to feed it my entire code base in order for it to know what the app does and then after a while it would get confused with too much info and I would have to start a new chat re-teaching it everything again. I would then cut and paste the resulting code into Xcode. Apple’s software for building apps. It was crazy inefficient. But it worked. Kinda.
Things changed in July, I went on my bi-annual pilgrimage to Boom music festival in Portugal. A place were the magic of synchronicity is commonplace. On the first day I was sitting in a shaded spot tripping lightly on psilocybin and chatting to those who came and went. I met a French music producer. We chatted for hours about our shared passion for creating stuff. Towards the end of the conversation I told him about the app I was building. He said it was great what I had achieved but pointed out how crazy inefficient the process was. He told me to use Cursor. He described how it could plug directly into my code. I could work with it to modify my code and it would never forget anything. I was so excited I almost wanted to leave the festival right away.
The AI Dojo
Installing Cursor when I got home I discovered it is indeed a useful tool but you need to be conscious of its limitations and learn the best way to tackle problems. I see working in Cursor like sparring in a dojo. I’m currently a blue belt in BJJ but regularly get my ass kicked in sparring. Despite losing time after time I am getting better. That’s like working in Cursor. As an example, the app is almost finished but just yesterday something went wrong and we spend almost 10 hours trying to get back to the last working version. Before coders shout out “Git!” I did and do – but the issue was in the data, so no matter how far we went back in the app code the same visual issue was still there. It took a long time to realise when the bug was introduced and the impact it had.
I find it amazing how smart AI can be but also amazing how dumb it can be sometimes. For the moment it needs a human with a helicopter view to get things back on track or else it gets stuck in the weeds. That’s like my sparring, I had my ass kicked for 10 hours yesterday in Cursor to get to where I was at the start. I learnt so much in the process. It is the final thing to implement. Today I can tackle the issue with a fresh perspective and I hope am able to finish up and start the process of launching it on the App Store.
The Unconscious Vibe Coder
So that’s how I became a full-time Vibe Coder.
The funny thing is I never even knew what a Vibe Coder was, I rarely dip into social media so completely missed the memes. I just had this overwhelming urge to focus all my attention on this project. The worst thing that can happen is that I end up with the perfect solitaire app for me. Which is actually pretty cool. The best thing that could happen is, well put it this way, there are two solitaire apps in Apple’s top ten list of games. Imagine the income an app that’s going toe-to-toe with a Microsoft product could generate for a sole developer! Hopefully I land somewhere in the middle. Either way, I’ve now learnt a new creative skill which I enjoy.
The Mother of All Synchronicities
Three days ago Tetragrammaton podcast released a bonus episode of Rick being interviewed by Ian Rogers. I read the show notes and was confused. It said Rick was recently named one of Time Magazine’s 100 most Influential People in AI and he’d written a book about The Way of Code: The Timeless Art of Vibe Coding. I have listened to all of Rick’s podcasts and strangely feel I know him well. He’s the first to say he is not a coder and does not have much of an idea on the purpose of code, beyond that fact that it helps make things and “some people are better at it than others”. I thought this episode was some kind of creative joke. A bit like his interview with Marlon Brando. I figured it’d at least help me understand what Vibe Coding is and I also love listening to Rick being interviewed.
It turned out that there had been a meme of Rick Rubin and Vibe Coding. Something I was completely unaware of. As a result Rick tweeted, “Tools will come and tools will go, only the Vibe Coder remains”, as a joke. Something I was also blissfully ignorant of. The tweet took off with millions of interactions. Rick saw the zeitgeist moment and decided to drop everything to write the book The Way of Code – which is a reinterpretation of the Tao Te Ching a book he has read annually for over four decades.
Synchronicity Surfing
As I felt the urge to drop everything and Vibe Code my app, unaware of Vibe Coding as a concept, Rick Rubin dropped everything and wrote a book about Vibe Coding, not even knowing how to code or what coding is. What is even more nuts, is that I have been using multiple texts of the I Ching or Yi Jing for guidance for around 20 years.
The Dao runs deep. Remember, it was a music producer that pointed out I should use Cursor!
I pulled out K’UEI or Contradiction whilst asking the I Ching “How should I approach the completion of my solitaire app?”. There was a changing line on the second line. Which changed to SHIH HO or Reform (Biting Through), it read…
“An unexpected or accidental encounter with an important idea or person will benefit you. There is a natural attraction at work here, although a direct approach would have been inconceivable or impossible.”
The I Ching Workbook, R L Wing
After I read that I then started on my walk and listened to the podcast that was lined up for me by Apple. Sometimes you just need to accept the serendipities. Trying to unpick them can distract you so much you miss more of them.
Ultimately you have to let go and surf the synchronicities.
I’m hoping to finish SOLACY this afternoon so should be hitting the Apple’s App Store soon.