The proliferation of generative AI and its increasing adoption across industries has been met with a healthy dose of concern — after all, all we've been told by experts over the last few months is how the technology's ability to efficiently automate and optimize tasks threatens to make many jobs obsolete, whether you're a coder, graphic designer, research analyst, patternmaker, model or even a writer.
We've already seen the power of generative AI in fashion, in its ability to simulate human intelligence — and replace professionals — with the use of AI models. Job losses, the obsolescence of entire career paths and the prospect being replaced by humanoid supercomputers, especially in this climate, is enough to drive anyone into a fit of worry. But as it turns out, generative AI may not be the first signs of an apocalypse, but rather a peek into a whole new world of work.
According to a recent report from McKinsey & Company, generative AI and automation will lead to some job losses, but even "the occupational categories most exposed to generative AI could continue to add jobs through 2030," suggesting that even clerical jobs that are most likely to be impacted by AI will continue to exist.
Historically, technological advancements have resulted in disruption, but they've also fueled economic and employment growth. And in generative AI's nascency, it can't function well without human support. So, while many of us fear our cities being overtaken by a sea of robotic office workers, what's more likely to happen is an increased focus on upskilling, changes to the scope of work and babysitting.
Sean Boyle — the co-founder of Unitmode, an AI-powered marketing tool supporting climate action, and the former head of sustainability the company formerly known as Twitter — argues that the "human-in-the-loop model," which combines AI capabilities with human sensibilities, enabling people to oversee the technology, give it feedback and ensure it's running correctly, "is what's going to make AI successful."
"This means that human influence is a core tenant of whatever is being developed or built," he says.
A human touch is undeniably important in fashion, which is built on and revolves around how people look and feel. But it's naive to think that efficiencies, cost-cutting and automation will be something the industry eschews in favor of tradition. Sure, fashion and, more specifically, luxury have been reluctant to adapt to new technological advancements in the past, but considering the setbacks caused by evading e-commerce, it could be a mistake to not embrace generative AI and repeat old mistakes.
Companies like Stitchfix have been using AI with the human-in-the-loop model for years to blend algorithms, machine learning and stylist insight to give customers a personalized and satisfying shopping experience. This kind of human intervention is likely to be adopted as AI becomes more pervasive throughout the industry, not only to ensure a pleasurable experience, but also guardrail against unsavory ones. After all, left to its own accord, generative AI hasn't always worked out so well: AI chatbots have reportedly engaged in creepy and inappropriate conversations; ChatGPT tends to consistently spurt out incorrect information and sometimes even just makes things up.
I decided to go straight to the source and ask an AI platform — ChatGPT, in this case — its thoughts on fashion companies implementing AI. In response, it referenced H&M using AI to create dresses inspired by Gustav Klimt's painting and IBM's "Project Debater," an AI chatbot that debates against humans "venturing into fashion design." Both claims are categorically untrue. (Worse yet, I found a passionate LinkedIn post about generative AI that referenced these "facts.")
"There needs to be safety mechanisms in place," Boyle continues. "Otherwise, we'll see more of these horrific experiences. This is why a human in the loop is going to be so important."
A human in the loop can span roles like system training and validation, data labelling and content moderation. As a result, there's likely to be a demand for subject matter experts, reviewers, assistants and data scientists with a unique fashion perspective to ensure that companies are employing responsible and unbiased AI development and deployment.
Fashion jobs of the future
As far as other roles that'll likely pop up in the industry, "AI fashion designer" might be something to consider, especially as initiatives like AI Fashion Week garner attention. However, it could be a while before the law catches up to ensure that this can happen without legal ramifications.
"Where things get tricky is when, for example, your prompts are directly referencing other designers or artists — that's where the intellectual property side of things gets weird," futurist trend forecaster Geraldine Wharry warns. These issues arise from AI systems needing to be trained with existing images, text, videos and information to effectively spit out something worthwhile. (Already, artists are noticing their work being scraped by generative AI models to do just that. Even Getty Images filed a lawsuit against Stability AI, claiming that the company copied 12 million images to train its model "without permission... or compensation.")
As these issues continue to occur, more companies are likely to turn to, and create jobs for, AI ethics officers and specialists to guide fashion houses and designers, whether that's for pattern making, archival referencing or designing. A need to develop internal policies, training, licensing but most of all trust is likely to be a north star when it comes to AI implementation.
In an interview with Business of Fashion, Anthony Lupo, an attorney at ArentFox Schiff focused on fashion and technology, said that, despite the tangled web of IP and copyright, the fashion industry shouldn't be deterred from adopting AI: "You should be leaning into this, because if you aren't, your competitor will be and you will be left in the dust."
"We're only really seeing the tip of the iceberg," Wharry muses when asked about the new economy of jobs on the horizon. Referencing the influx of AI models — namely, brands being able to hire Eva Herzigova's AI rendering in campaigns — she tells me she sees an emerging need for management: "There's a site called futuretimeline.net, and they predict that by 2030, there will be such a job as avatar manager. But I think it's already happening."
Digital models and digital modeling agencies already exist. (There's Shudu and Lil Miquela, plus the agency The Diigitals.) However, generative AI models replicate humans — a small but important difference.
Wharry also foresees manufacturers creating materials in labs and adding AI to the mix. "I can't tell you exactly for sure what those jobs would be, but with the power of AI, you could potentially generate a zero-waste pattern really fast based on the sketch," she predicts.
AI in Sustainability
AI may have a tendency to go off the rails when unattended, but Wharry believes its best use is something we've barely begun to use it for: "how it can help in efficiencies in terms of sustainability."
She points to Boyle's company Unitmode as an exciting example of this burgeoning space. "He's created a marketing tool for credible climate action," she explains. "By using AI, they're able to track greenwashing claims from brands."
Unitmode uses more than 20 million data points and a handful of guiding parameters to evaluate greenwashing claims: "Are they transparent? Do they share their emissions on a yearly basis using third-party audited accounts? Have they declared a very clear ambition that is backed by science? Have they shown credibility of lowering their emissions on a year-on-year basis?
"We use science-based target initiatives for that," Boyle explains. "The first two factors without that third one are useless, because anybody can declare that they're doing things in the long run, but are they serious about it?"
Boyle refers to this assessment as a company's "climate action credentials." Unitmode has already assessed 6,000 companies on this scale, including LVMH. The plan is to ultimately release an app that people can use as they walk through a department store and, in real time, see what a company has done and if it's serious about sustainability efforts.
"We're still a long way from where we want to be, but this is one clear example as to how AI can give consumers superpowers and break that asymmetric influence that brands have had on them," Boyle says.
Actual greenwashing jobs have been an issue in fashion. Last September, Kourtney Kardashian was announced as Boohoo's "sustainability ambassador"; both the reality star and the fast-fashion brand received widespread criticism for it, yet these titles often have more to do with publicity than real sustainability efforts. With true advocates ike Boyle using AI to unmask greenwashing, we may see such roles disappear — or even better, be given to more legitimate, experienced figures.
"The more companies adopt it, the more important it will be for them to have AI ethicists, so everything from the supply chain, retail, and design will be affected," Boyle hopes.
Currently, jobs that require AI proficiency are few and far between, although more job descriptions are marking knowledge of AI as a preferred skill. Listings for fashion analysts, website developers and engineers signal a slow but steady move towards more AI in fashion. Though, the demands on Upwork are a little more ambitious; one listing reads: "Build me a fashion AI that gives me an opinion of my outfit." It seems there'll be no lack of busy work for AI developers with a passion for fashion.
Audrey Shaeps, the founder of Los Angeles-based fashion recruiting agency The Workshop LA, hasn't seen an influx of jobs requiring AI proficiency, but it may be because "the creative field, it's not so black-and-white."
"I'm excited for how AI can be used within recruiting though," she muses. "I think that, for my business, it will make us much more efficient."
It's easy to let thoughts of the future run wild, especially when considering what we could achieve — and solve — with a supercomputer in tow. But at these early stages, it's really up to us to learn how to actually use the technology.
Drishti Gangwani, the founder and CEO of Closr, a platform for emerging, ethically-minded designers, sees AI proficiency being prioritized at the higher education level. "Within the fashion education system people are already learning to design using digital platforms, and that's becoming a norm now," she says. "It's no longer just an option."
As a strong proponent of supporting emerging designers, Gangwani sees these AI programs leading to "significant cost reductions" that would open up the industry: "We may not be a 100% there yet, we're getting there pretty quickly. There are hundreds of programs that are working to develop technology where you can create garments that range from a size 2–14 with the click of a button. In five or 10 years, whether you want to start your own collection or work at Ralph Lauren, you'll need to know how to use these programs and technologies, regardless of what business you're running or whether you're working for someone else."
Even though AI is unlikely to be the robotic boogeyman we all feared, it's evident that a big change is coming. The future of work will look very different from what we know today, and it's essential to be prepared. Boyle, who himself has seen numerous new iterations of technology, suggests it's going to be more than a passing trend.
"I think AI is gonna become as ubiquitous as computers," he predicts.