In the final season of Mad Men, Peggy has an idea for a burger chain commercial that her bosses like. Then doubts herself. She goes back to her empty yellow legal pad, jots down a new concept from scratch, and returns to eat at the restaurant, even though she’d already visited more than a dozen around the nation. Nothing about her process is efficient, and her client would have likely been perfectly happy with her approved pitch. But that messy, circuitous method of deconstructing an idea and starting from a blank page is how she lands on something much better than good.
AI has supercharged our era of optimization — or at least the hunger for it. Software engineers work quicker, decks materialize with a few clicks, and AI agents manage email inboxes. Gen AI is heralded as the solution to the horrors of the blank page. Microsoft says AI agents “streamline repetitive and mundane tasks so that users can focus on solving more meaningful challenges.” Mark Cuban said late last year that AI empowers “creators to become exponentially more creative.” Other business leaders like Eric Yuan of Zoom, Bill Gates, and JPMorgan Chase CEO Jamie Dimon have teased that the mass adoption of gen AI will free white collar workers from the five-day workweek.
The reality isn’t so rosy for everyone. Some are working longer hours with AI and taking on a broader range of tasks. MIT research published last summer found that 95% of AI pilot programs had not had a measurable ROI. The race for productivity and the promise that we’ll be able to do more “meaningful” and challenging work misses a sentiment most white-collar workers like Peggy know: There’s only so much creative thinking a person can do in a day, and the sloppy, trial and error process of bad ideas that gets to a good one doesn’t lend itself to streamlining. Work where people have to hold multiple ideas in their heads, synthesize, connect them, and produce work is “cognitively very taxing,” says Emily DeJeu, a professor in Carnegie Mellon University’s Tepper School of Business. “It takes time. It also just takes a lot of space and bandwidth, and the idea that AI can assist with that, I think, is a bit fallacious.”
Humans aren’t machines of creativity. Sometimes, we thrive at our most inefficient.
Highly intensive cognitive work has its limits. Air traffic controllers typically work schedules segmented by shorter stints in different positions with breaks in between. If AI can take over some of our mundane office tasks, and the rush to fill hours with more productivity emerges. Harvard Business Review published new research in February that found “AI tools didn’t reduce work, they consistently intensified it.” The eight-month-long study examined 200 employees at a US tech company, and found that they worked quicker but picked up longer hours and took on more tasks outside of their job descriptions. More employees started vibe coding, giving engineers more code to review. People prompted AI tools during their breaks, erasing the lines between work-time and downtime. The growing workload, the researchers wrote, can lead to “cognitive fatigue, burnout, and weakened decision-making.”
As employers and employees look to AI to relieve bottlenecks in workflows, “it can be challenging to know which tasks, even though they might be annoying, are contributing to your ultimate learning or your workflow, and which tasks are truly superfluous,” says Ben Armstrong, the executive director of MIT’s Industrial Performance Center. Not all learning comes from strenuous focus — there are smaller breakthroughs hidden in tedious labor. Armstrong gives the example of looking through a data set, which can help him to see what’s included or excluded and gives context. If an AI model alone analyzes it, the neatly delivered analysis may conceal errors and gaps. “I worry that maybe we’re not as good at the higher value added work if we don’t do some of those mundane tasks.”
A lot of the things that make us so busy and drive up our long work days, it’s not really problem solvable by AI, it’s problem solvable by workplace culture.Cal Newport
One of the biggest issues with AI at work is that people use it to skip out on the challenging, critical thinking tasks, like staring at a blank page and coming up with a new strategy or idea, Cal Newport, a computer science professor at Georgetown and author “Slow Productivity: The Lost Art of Accomplishment Without Burnout.” Newport, who coined the term “deep work,” says a lot of AI use “right now is less about speeding up or removing administrative tasks that are in the way or preparing you to better do value production deep work,” Newport tells me. Instead, workers are “trying to reduce or avoid the peak cognitive strain of doing harder thinking.” They generate workslop and edit their back way through it, preferring to push past the discomfort of starting a new project. The blank page is a foe to be avoided, not a challenge to meet.
Tech companies are promoting a world where you don’t need to feel that friction. “The secret to finishing your work is doing the work,” reads a Zoom blog post about the company’s AI companion. “When you get stalled at that blank page, the best thing you can do is fill it with something. With anything.” But there’s reason to hesitate before calling in AI. Research from MIT published last year found that people who used ChatGPT to assist with writing essays over time tended to get lazier and more dependent on the tech. They “consistently underperformed at neural, linguistic, and behavioral levels” compared to control groups that either had no help in writing essays or those who used Google. Similarly, research last year from the University of Pennsylvania’s Wharton School showed that tools like ChatGPT could help an individual’s ideas, but tended to flatten creative thinking among groups. And a study from Columbia Business School found that large language models have a bias for the first option it’s presented, threatening to undermine great ideas that might come later in a brainstorming process.
Workers are already noticing that relying too heavily on ChatGPT could cost them a creative edge. Karim Adib, a public relations manager at the software company Search Atlas, tells me he started using gen AI to brainstorm ideas. But just two weeks ago, he went back to his old ways: stepping away from his desk with a notebook. Adib went on a walk and to a library to come up with new strategies. It’s not as efficient as having ChatGPT spit out a bunch of starting points, but Adib says that what eems like more effort is actually more fun. It’s important “to be bored, be with your own blank page, find your own ideas,” he tells me. “Everyone else has access to ChatGPT as well, and everyone is slowly getting better at doing it. So if me and someone else get the same idea, if we’re both executing in the exact same way, that gives me zero advantage.”
Mundane tasks offer a break for our brain to reset.
One software engineer and founder wrote on X that he had hit a “vibe coding paralysis,” which he defined as “the syndrome of wanting to do so much — and being able to do so much — that you end up finishing nothing.” AI made it easy to start projects, but he was then bogged down by editing and de-bugging. It’s difficult work met with a fragmented attention span, and the desire to re-prompt and start over was wrong. The pattern created a “graveyard of almost-finished projects,” he wrote. Using AI early on while working on a project, DeJeu says, “actually ends up neutering the expansive, creative, deep thinking process.”
Mundane tasks offer a break for our brain to reset. There’s still productivity at play, but it’s in filing papers or inputting basic data. Research shows there are benefits to doing chores around the house, from a sense of accomplishment that follows tasks as simple as making the bed to increased self-efficacy, which boosts confidence. Too much boredom can lead to what’s recently been deemed as “boreout.” With less to do, some employees are losing sight of their purpose and feeling distressed and exhausted. But completing boring tasks at work can lead to new bursts of creativity for problem solving, according to a 2014 study from University of Central Lancashire researchers.
These little work tasks or a blank page aren’t necessarily driving workplace burnout. Feeling like you have no work community, your values don’t match the organization’s, or that you aren’t appropriately rewarded for your work all contribute to the feeling. Newport says part of what’s tiring workers is “frenetic” communication. People loathe email. The constant ping of messages can be more distracting than helpful — a 2025 study by Microsoft found workers using its suite of products were interrupted by chats, emails, and meetings an average of 275 times a day. Now, workers are interacting with AI agents and chatbots, leading to more back-and-forth communication taking place at higher speeds with a conversational partner that’s always on. “It’s like being in a very wild dinner party and it never stops,” Newport says. The early implementation of AI at work, he says, is repeating a lot of the problems brought on by email and the internet, that blurred the lines between work and personal lives. “A lot of the things that make us so busy and drive up our long work days, it’s not really problem solvable by AI, it’s problem solvable by workplace culture.”
It’s not surprising that we’re in the midst of a Mad Men rewatch craze. There’s something about a show where people deeply know their coworkers and clients (often too intimately), where work moves slowly and all communication is funneled through secretaries, and where the mere concept of putting a computer to the office sends one worker into a full mental breakdown that feels all too poignant as we navigate AI.
Workplaces that want to build a sustainable culture need to consider ways AI is upending the deep work their employees do. Turning nothing into something is the challenge that helps people grow. It leads people to feel accomplishment and ownership over their work. Too much friction and red tape at work or too many Zoom meetings are bad. But a few quiet, boring moments to tinker and organize can serve as little rituals that anchor us.
Amanda Hoover is a senior correspondent at Business Insider covering the tech industry. She writes about the biggest tech companies and trends.
Business Insider’s Discourse stories provide perspectives on the day’s most pressing issues, informed by analysis, reporting, and expertise.
