Going On Vibes – Where Software Sits Today, Fashion Could Follow

We write a lot about software and fashion here at The Interline – usually together! But as much as these two sectors, and the places they intersect, are our dual passions, they only rarely follow in each other’s direct tracks. What happens in fashion to drive the day-to-day direction of the industry is usually pretty distinct to what’s setting the pulse of software, and vice versa.

Sometimes, though, those two sets of stars align, and fashion and technology find themselves wrestling with the same questions at the same time. And as AI has carried on climbing the ranks of industry-agnostic “big things to care about,” the two tracks have wound up running in parallel more often than usual.

This week, there are a couple of indicators that fashion and software are not just tackle a common, monolithic concern from separate angles, but that both sectors are experiencing an acute version of a narrower concern: what will it mean if the next big company is started by people with no fashion or software experience at all?

And what if, instead of hiring people with specialist skills and experience, or identifying a partner to help them turn an idea into a workable product or business… they use AI and essentially just wing it?

The former of those approaches, at least, is nothing new in either sector. 

In fashion, celebrities and entrepreneurs have, for a long time, seen apparel and accessories (less so footwear, usually) as industries they can essentially “borrow” their way into. They accomplish this by either partnering with existing brands and making use of their designers, developers, supply chain partners, and distribution networks as scaffolding to bootstrap new capsules or subbrands, or, if they have the money to invest, by recruiting people from those domains to port pre-existing fashion knowledge into new company vehicles.

(We don’t mean to trivialise the work that goes into collaborations, but it’s safe to say that most celebrity partnerships are lopsided when it comes to who brings the industry knowledge to the table.)

In software, too, it’s become a running joke that everyone, deep down, believes they’re a budding tech billionaire who just needs to get someone else to do the troublesome coding work to turn their transformative idea into a killer app. They even made a film about it.

The latter approach, though, is something new for both industries – even if software is a little further down the road. To carry on the tortuous analogy, if the Winklevoss twins were transported from turn-of-the-millennium Harvard and wanted to prototype a revolution app idea, they’d no longer need to turn to a fellow student in a computer science class. They’d open an IDE like Cursor, invoke one of the language models best-tuned for coding, and simply let the AI handle the development for them.

And if they still needed more in the way of operations and architectural support, they might have turned to a platform like Base44, a small start up that managed to raise a fairly unremarkable amount of money in “big tech” terms this week ($80million in cash from self-serve web development platform Wix), but that raised a much larger question about how concepts are being developed and commercialised – one that could have some real parallels in fashion.

Base44 describes itself as “an AI platform to build fully functioning apps,” which places it squarely in the realm of “vibe coding” – the practice of conceiving the idea for a piece of software, and then collaborating to some extent with an AI model to build it. There are a lot of different definitions of what really constitutes coding based on vibes or feelings, from seasoned developers using the approach to allow themselves a more strategic outlook while an AI model handles the minutiae, to entirely novice programmers (i.e. people with no software engineering skills or experience) showing up with an idea, letting an AI create it for them, and then either using the resulting app themselves, or working to commercialise it.

For some career coders, the idea of complete neophytes deploying software they let an AI create, and that they do not understand even the basic inner workings of, is treated as either a joke (it’s hardly the most secure way to develop and roll out software) or an existential threat to livelihoods they’ve spent decades building. And the latter is an understandable position: it’s not exactly the easiest pill to swallow to have an entirely new audience appear and effectively try to usurp your position without actually acquiring any of the necessary skills.

But the word “necessary” is doing a lot of heavy lifting there. Up until relatively recently, if you wanted to create an application for people to run in their browsers or on their desktops, laptops, or phones, you needed developer accounts, knowledge of languages like JavaScript or TypeScript, and libraries like React, and either the ability to spin up development environments (local or cloud) yourself, or the willingness to pay major cloud infrastructure providers to do it for you. 

Without those skills – or without partnering with, or hiring someone who has them – you simply could not develop software. Those skills were de facto essential to the task at hand: bringing an idea to life. Vibe coding solutions, models, and platforms collapse that skill acquisition process to essentially nothing, meaning that anyone with a couple of hundred dollars to spend on Claude Code tokens can at least approximate (if not outright replicate) outcomes that would have otherwise taken years to work towards.

Are these tools shortcuts or sacrilege? The answer really depends on where you stand, and on how planted you’re feeling against the coming storm. Are the best engineers people who have built traditional skills and are now using AI to support them in scaling their abilities? Probably. Are beginner engineers able to achieve a rough stab at similar results? More so now than ever, yes.

And it would be entirely possible to level the same kinds of criticisms pointed at vibe coding at digital art. Are the best digital artists people who learnt and honed their craft in other media first? That was definitely the case a decade ago, but now it’s hard to say. And it’s evidently true that digital-only artists, across 2D and 3D, are extremely talented people upon whose work a new era of creation, content, and tools has been built.

In the long run, it’s neither sustainable nor inclusive for entire industries to remain tech and skill-gated, excluding potentially revolutionary ideas from people who simply lack the resources, time, or traditional pathways to acquire specific skills. From a software development point of view: the next transformative platform or service might have been lying dormant in someone’s mind, unable to escape because its creator was priced out of accessing the scarce resource of technical expertise – and lowering those walls could be all it takes to release it.

And it doesn’t take much to extrapolate that same idea to fashion, which, at an industry average level, has built some very deep moats that have served to gatekeep who is permitted to design, develop, make, and distribute clothing.

To be clear: The Interline expects the first decent-scale (i.e. non-hobby) “vibe fashion” brand to already be out there, and we’re fully predicting an influx of more. After all, there is nothing fundamentally unique about design, patternmaking, merchandising, marketing, or other fashion disciplines. Outside of the unanswered question of “taste,” each of these could conceivably be automated to some extent, and then that automation could be offered up as a turnkey product (with a supply chain as a service partnership or a print-on-demand network backing it up) to people who have an aesthetic vision and are willing to let an AI turn it into tech specs, calculate material yields, game what-if cost scenarios and so on.

Quite how fashion designers, technical designers, and other teams will feel about this is another matter. The likelihood is that we’ll see the same spread of reactions that the software industry is currently demonstrating: some people will find the idea insulting; others will welcome the creative competition that comes from a levelling-out of skills.

But make no mistake: once working on vibes becomes more commonplace in hands-on creative and technical skills, there’s no reason to assume it will stop there. This week, Marvin Ellison, the CEO of Lowe’s (the American home and hardware giant) went on record to say that essentially every back office, knowledge-based role is apt to be challenged by AI in the near future, and that hands-on skills and customer service are the least likely to be replaced soon. 

(There’s a charitable reading of this statement that sees it as self-aware, in the sense that executives are also included in that expendable cohort, and a more cynical one that says he probably assumes CEOs are excluded.)

For the middle tier that Ellison is alluding to – in fashion that would be the creative designers, product managers, material developers, merchandisers, marketers and so on who contribute a slice to the go-to-market journey – the fear will soon be tangible. These, like essentially all knowledge roles, are positions that have felt secure, and have had high walls and onramps placed around them. But it’s entirely possible that an AI model could someday soon produce more than half the value at a fraction of the cost – and once that happens, it’s basically guaranteed that someone else will package and sell that tool to make that entire go-to-market process accessible to anyone.

There’s also reason to believe, based on more news from this week, that fashion insiders are no longer guaranteed the industry’s top positions, with Kering bringing in a seasoned executive who comes from the automotive sector to turn its flagship brands around. The tacit admission here is that time-in-industry may matter less than business acumen and attitude – both of which can be brought to bear on essentially any industry.

So, with this kind of cross-industry optimisation happening at the very top, and software tools already demonstrating just how comprehensively previously-ringfenced work can be done by AI (Microsoft estimates that a third of all its code is now written by AI), it seems inevitable not just that a ‘vibe fashion’ brand will make it big, but that brand will be started by an industry outsider who sees the complexity and the fragmentation of typical fashion value chains as pure inefficiency, not as a craft to be preserved.

This week, and for the coming year, fashion will do well to watch what’s happening in software. Because fashion’s turn at having its moats and skill gates knocked down might be coming sooner than people like to think.