It started with coding. Generative AI’s aptitude for writing code was the death knell for traditional software development, and companies wanted “vibe” coders. Big Tech execs have been praising the vibes this year: Sundar Pichai is vibe coding a web page, Mark Zuckerberg says AI is coming for mid-level engineering work, and Klarna CEO Sebastian Siemiatkowski says he’s become an amateur coder thanks to vibe coding. Startups are vibe-coding their way into existence.

Now, more of the corporate world is vibing. A small number of companies are seeking applicants for job titles like “Vibe Growth Manager,” who are tasked with experimenting with AI and building marketing prototypes faster. Last month, Microsoft rolled out what it’s calling “vibe working,” which involves using agentic tools in Excel and Word that can generate documents and spreadsheets. It lets people without deep knowledge of spreadsheets “speak Excel,” or “vibe write” in Word by generating, refining, and asking the author questions as they go. Mea’s AI app now has a “vibes” feed for AI-generated video, and Sora’s AI video platform is giving rise to what some are calling “vibe creators” — no longer traditional influencer content, but a new type of influence created by synthetic AI imagery and a few clicks.

Welcome to the vibening, where a lot of white-collar work is being portrayed as just vibes. The term is shorthand for using generative AI to do the tedious and strenuous parts of a project, but it also conveys the idea that work is free-flowing, improvised, and easy. Vibing is a sort of Gen Z take on hygge, slang that was once reserved for chilling with friends or describing a date gone right and has now seeped its way into corporate speak. Managers hold regular “vibe checks” with their direct reports. Some companies have played with a “Chief Vibe Officer” title. Smirnoff announced Troye Sivan into the role as part of a promotional partnership last year, and software company Atlassian nominated a rotating CVO in an attempt to grow bonds between teams.

But vibe working is still work. Working with AI, and doing it well, takes experimentation and expertise. The rise of talk of vibing at work may obfuscate the value of mastering concepts and skills, or the term could be the bat signal of a company seeking energetic workers who want to experiment. If it’s an attempt by companies and executives to convey they’re open to AI and hip, the imprecise language and experimentation can be a recipe for confusion. “Everyone might have a different interpretation of what the vibe is,” says Ben Armstrong, the executive director of MIT’s Industrial Performance Center. “One person’s good vibe could be the other person’s bad vibe.”

So what happens when the vibes are bad?

It’s not surprising to see the idea of vibe work gaining traction, as Gen Z sees the parameters of work with blurrier lines. From the lazy girl job to quiet quitting, there’s less formality in the workplace, and young people are less interested in workplace loyalty and less dependent on the 9 to 5. Workers feel disengaged with corporate culture, so a rebrand to the gentler idea of vibing could be an attempt to attract workers to a less rigid workplace. “I imagine to this particular demographic of people, that’s very appealing: work being about vibing more than it’s about analyzing or synthesizing or reporting, which I don’t think sounds particularly artistic or creative or collaborative or beautiful,” says Emily DeJeu, a professor in Carnegie Mellon University’s Tepper School of Business. But the term “hides the extent to which it is work,” DeJeu says. If executives rebrand work as a “vibe,” it may undercut the necessary expertise needed to do a job. That could become exploitative if bosses rely on workers’ mastery of skills but simultaneously write off the value of work performed alongside AI. DeJeu likens vibe coding or working to jazz. A performance might be largely improvised and seem effortless to the listener, but that vibing on the spot only works because musicians have spent years mastering theory before taking steps to mess it all up. “Labor is labor, and the labor to build expertise is laborious,” she says. The idea that you can vibe and “you don’t have to spend all that time and it’s not hard, is kind of laughable in my opinion.”

Vibe working is still work.

Vibing hasn’t been the silver bullet for coding some expected. There’s been a frenzy for AI-generated code, and OpenAI’s Andrej Karpathy coined the process as vibe coding. The concept has thrown the software developer role into a new era — the nature of work done by many has shifted from writing code to an emphasis on reviewing AI-generated code for bugs, and coders aren’t necessarily saving time.

The trend has caught on, and employers are hungry for employees who know how to use AI. They’re eager to deploy cost savings and find the productivity gains touted by AI evangelists. Even though most companies aren’t training people to use AI, they want workers who get it: 71% percent of business leaders say they would take a job seeker with less experience who has AI skills over a more experienced worker who doesn’t know how to use AI, according to a 2024 report from Microsoft. Two-thirds say they wouldn’t hire someone without AI knowledge. But less than a third of workers have received company training to use AI, according to a survey of workers conducted by Jobs for the Future, a nonprofit focused on transforming the workforce. There’s a big gap in what companies want in terms of AI and how they’re able to convey that and train their workers.

To cope, workers have jumped on AI themselves. Learning about the tech is often happening bottom-up from workers rather than top-down through training. Workers are experimenting in ways that aren’t always tracked, so the best practices are being built just as the tech’s limitations are discovered. “Because a vibe is so open to interpretation, it’s so hard to measure what the outcome of these different tasks might be,” Armstrong says. We’re in a time similar to the early days of the internet, he says, when people were experimenting and developing different types of web interfaces. With AI tools, people are “figuring out when they’re going to be effective, when they’re going to be reliable.” All of that vibing can create vastly different processes with varying degrees of success that also prove hard to replicate.

When people vibework too hard, when they use generative AI without thought, they can produce mounds of workslop, or neatly prepared decks and memos that are often lengthy but lack useful information. “As you have an idea, you should also have your strategy and your objectives, and then you should use AI to help you flush out the idea,” says Emilie DiFranco, vice president of marketing at the firm Marketri. DiFranco says AI is helpful for marketers because it can review and consolidate data, but relying too heavily on AI without the right endgame for a marketing strategy in mind could get messy. “I am a little worried about losing the human aspect of creating that initial strategy and the objectives,” she says. Marketers should be “not just rolling off a vibe, but making sure there’s research, making sure you have those foundational elements before you engage with AI to start helping you put the plan together.”

Vibing is in vogue right now. As companies and execs move quickly to capitalize on the idea, it could turn cringe. But work is still work, and trying to throw a fun twist on how we talk about it or putting generative AI tools in the mix doesn’t mean employers aren’t demanding a lot from their workers.

Amanda Hoover is a senior correspondent at Business Insider covering the tech industry. She writes about the biggest tech companies and trends.