Gen AI was meant to side-line the artists, the tinkerers, the gut-led weirdos. Funny how it’s also handing us the tools…and nudging the tech geeks back in their (black) boxes…
If you follow (or just can’t seem to escape) any of the AI thinkfluencer conversations, you’ll know how vibe coding has emerged as the new thing in some of those circles. Or their parents' basements. In simple terms, instead of meticulously planning out architecture (is that what they call it?) or writing detailed specs, a vibe coder jumps straight into coding a prototype, guided mainly by intuition in the beginning. Using an AI coding assistant, you can just describe the app idea in a few sentences of a prompt, and let the AI generate quick-and-dirty code. Essentially, these more creatively minded developers are now building software based on a loose idea or feeling first.
In a few minutes, you can have vibed something up and running to play with - messy, rough at the edges, but alive. This approach is improvisational and exploratory. The key is momentum. Traditional software projects usually start with a bunch of complicated setups and designs, but vibe coding kinda flips the script. You implement first, then fiddle about afterwards. Riding a burst of creative energy from the get-go. You move fast not by recklessness per se, but by deliberately ignoring the usual upfront perfectionism to get to a sketchy first draft.
In essence, vibe coding is about capturing the spark of an idea immediately, optimising for flow rather than early correctness. The code might be held together with digital gaffer tape, but it works well enough to play with. It’s coding by instinct, with the faith that you can sort it out once the core idea proves itself.
In early 2025, noted AI researcher Andrej Karpathy coined the term. Karpathy described it as “a new kind of coding… where you fully give in to the vibes, embrace exponentials, and forget that the code even exists”.
It’s an almost punk rock method I’ve also adopted for advertising planning, on occasion. Especially when you are working with great creatives, they don’t need or want loads of graphs and theory to get going. Just prompt them with some basic information, the foundational research and data about the target audience, market conditions, and a proper insight, then let them loose on ideas.
Let them bash it out.
Then tart it up later.
Tarting it up in a planning sense means evaluating ideas using theory, strategic analysis, and marketing principles to find the gold, then executing on those.
Think of the ideal strategy-creative continuum as like the KLF’s ‘Liberation Loophole’. The loophole is a creative mindset that turns constraints into opportunities by deliberately bending or reinterpreting ‘rules’. We are looking for gaps in systems, whether creative brief conventions, category and advertising norms, or self-imposed limits, to escape creative blocks. The goal isn’t to break rules for the sake of it, but to treat ambiguity as a tool to redefine the work on our terms, a strategic subversion.
Back in London in the late 1970s. Nick Lowe, a pub rock luminary-turned-producer, is helping shape the sound of the emerging punk and New Wave. As the in-house producer for the upstart Stiff Records, Lowe was known for a deliberately rough-and-ready studio style. His affectionate nickname was “Basher,” earned by his instruction to bands in the studio.
“Just bash it out and we’ll tart it up later.”
In Lowe’s world, to bash it out meant to get down a song quickly, in as few takes as possible, without obsessing over tiny flaws. The philosophy was to capture the raw feeling of the performance first, snare the excitement of a song while it’s fresh and instinctive and worry about polishing it (with overdubs, effects or corrections) afterwards. This approach was all about vibe over technical perfection.
It’s no surprise it worked a treat in the punky times. Lowe produced Britain’s first punk single - New Rose by The Damned - and twiddled knobs for Elvis Costello, Pretenders, Dr Feegood and many others. In that context, a slightly unhinged, high-energy vibe was more important than pristine production.
Lowe's rock 'n 'roll attitude ensured the music retained its immediacy, rather than being sterilised by endless retakes. It was a conscious antidote to the bloated, perfectionist studio productions of the time to prioritise the spirit of a sound, and accept that a bit of roughness can be fixed in post if needed.
Vibe coding (my ‘vibe strategy’), and Lowe’s ‘bash-it-out’ all share a similar core principle. Keep the creative momentum going and don’t let the insistence on perfection (too soon) kill the groove. All three approaches start from a gut instinct and run with it before the rational, critical brain can step on-the-brakes. It needs a certain faith in the process, a trust that the feel of a thing, if pursued with the right energy, will lead to something worthwhile, even if it’s technically rough at first.
Bash-it-out musicians, devs and planners have a kind of creative pragmatism. Truth (or value) is what works in practice. Why chase an unattainable ideal of a perfect design when you can learn more by doing? The immediate prototype or rough mix provides feedback and keeps inspiration alive. Spending too long at the planning stage can bleed the life out of creative work. By privileging immediacy we ensure that the essence – the soul – of the idea isn’t lost in overthinking.
Because we can always tart it up later.
Pragmatist philosophers have long argued that ideas must be tested in action, not simply thunk. An idea gains meaning and credibility through its practical effects. Vibe and basher methods both embody this pragmatist ethos. The worth of a software idea is proven when you see a rough app actually running and solving a problem (even if held together with cellotape and string). The worth of a tune is proven when that quick take feels good to the ears and makes the listener groove a little, even if the mix is unpolished. Both cases seem to echo the philosophical notion that execution trumps theory (at least in the rough mix) or that becoming and iterating gets things moving better than obsessing over an ideal being of a project, too soon.
In a kinda Deleuzian vibe, we could say these instinct-driven processes embrace the rhizomatic (yes, its a bit PoMo, don’t @ me) nature of creativity – bouncing off in unpredictable directions, free from rigid hierarchy or predetermined form. What I’m now claiming for the lexicon as vibe bashing (geddit) lets a thing take shape organically, each new fix or feature suggestion leading to new emergent structures, much like an improvisational jam session can morph a song’s structure on the fly. This might sound more jazz than punk, but in practice, it simply feels like freedom. The liberation to create first and criticise later, to favour experience over essence. Of course, even Deleuze would agree to the fact that eventually we must impose some structure, some tarting up, but not before squeezing out the possibility juice.
(I’ve not played with them enough but I kinda feel like some of the emerging tools like Springboards.ai and maybe memorised.ai are somewhere in this space. They give you things to play with rather than answers.)
If this feels too loose a logic vibe for the scientific side of your mind, then think of vibe bashing as evolutionary labs for ideas. Both processes allow lots of little mutations (rough prototypes, quick takes) to be tried out in a short time, subjecting them to the environment to see which thrive. It’s effectively natural selection for creative concepts. A vibe-bashed prototype is thrown up as fast as it’s conceived. Isn’t it ironic, dontcha think? For the last few years, we’ve been beaten over the head by the story that AI will replace the creatives. That machines will generate the novels, paint the masterpieces, compose the symphonies and make the ads. And yet, look at where we are. The rise of vibe coding and vibe bashing isn’t really about machines replacing human creativity at all. It’s about machines making it hilariously easy for creative people to bypass the old technical bottlenecks.
Because that’s the real pattern of history, if you bother to peer past the hype. The technology always comes first. The steam press, the electric guitar, Photoshop, Python libraries, diffusion models. But it’s never enough on its own. It only becomes something when creative people start messing around with it, bending it, misusing it, pushing it into places it's not supposed to go.
And that’s when technology always gets interesting. When the tool becomes a toy, a medium for mischief, and a catalyst for beautiful chaos.
AI was supposed to replace the artists, the tinkerers, the gut-led weirdos. Instead, maybe in a few small corners it’s empowering us. Machines have made it easier than ever for someone with a spark to skip the bureaucracy of expertise, invoke the Liberation Loophole, fumble forward vibe bashing on intuition fumes.
And then find a geek to tart it up later.
Eaon Pritchard | Strategy | Consumer psychology