Why AI workers won’t let bots do the most basic tasks
People immersed in artificial intelligence sometimes take an old-school approach to work.
I tend to assume people who work in artificial intelligence live like the Jetsons. Even if they don’t have robot maids, surely they automate as many tasks as possible, right?
Stella Dong is a machine-learning engineer and co-founder of an AI startup called Reinsurance Analytics. Yet when it’s time to write an email, she taps out the message with her own 10 fingers.
“I don’t trust AI to draft by itself," says Dong, who also has a day job at a healthcare-technology company.
People who are immersed in AI often have some surprisingly old-fashioned habits, from scrawling meeting notes on paper to inputting calendar entries manually. Notably, some of the duties that the rest of us are most inclined to give to bots are the very things AI super users insist on doing themselves.
This should make us consider whether we are doing AI adoption backward in some cases, or using the latest technology simply because it is shiny and new. AI savants’ greatest strength may be their ability to tell the difference between the to-do’s that robo assistants improve and those that are better done by hand.
Dong sometimes uses Copilot to revise emails but prefers to write her own drafts because no one knows what she wants to say better than she does. This is the reverse of many professionals who edit—or copy and paste—whatever an AI tool composes in response to a prompt.
She also eschews AI calendar managers. She doesn’t like to rely on digital notifications to shepherd her through the day and is more likely to remember appointments she puts on the calendar herself.
Booking every meeting isn’t practical for a busy AI entrepreneur. That’s why when Dong was traveling recently, she asked her (human) business partner to pick up the phone and call me to arrange our interview.
Slower on purpose
Bots already are capable of a lot and presumably will learn to do more in the future. A new McKinsey Global Institute report estimates existing technology could perform 57% of Americans’ work hours.
The consulting firm calls this “a striking but easily misunderstood figure." It doesn’t necessarily mean most jobs will disappear but suggests a majority of work activities could be automated. In theory this frees people to focus on higher-level stuff.
How reassuring.
Another interpretation is we are embarking on a future where there won’t be much that AI can’t do, technically speaking. The question becomes: What shouldn’t it do?
“As we redesign work and jobs, you might actually choose not to maximize how much an AI agent or robot does," says Lareina Yee, a senior partner and director of the McKinsey Global Institute.
For example, companies might hold back some automation to train junior employees in foundational skills and ensure they are savvy enough to check AI assistants’ output.
On an individual level, Yee says, people may opt in to occasional busywork to give their brains a break or simply work a little less instead of squeezing every drop of AI-enhanced productivity.
Ziyi Liu takes typewritten notes during meetings—even though she is an AI research intern at Microsoft, whose software can automatically transcribe and summarize the sessions.
For many of us, meeting recaps are one of AI’s greatest gifts. I used to groan when it was my turn to be the meeting note-taker in a previous job.
But for Liu, a doctoral student at the University of Southern California, AI’s ability to perform this drudgery is beside the point. She maintains her note-taking practice to model herself after a mentor who was known for circulating detailed meeting minutes and action items.
“I want to be that kind of person that keeps their work very structured and clear," she says. “I don’t want to look at a transcript; I just want to do it myself. It makes me feel like I’m in control."
Think with your hands
Ryan Bearden goes one step further and takes notes in a Moleskine notebook. The act of handwriting helps him commit things to memory. Plus, this old-school habit is a way of signaling to others at the table that they have his undivided attention.
We’ve all sat across from people on laptops and smartphones and wondered whether they were jotting down bullet points or shopping early Black Friday deals. Pen and paper make us feel heard.
Bearden, a marketing consultant, spends his days training business teams on AI tools. Much of his work centers on how to speed up market research by using bots to aggregate and analyze large amounts of information.
His presentations to clients begin as physical storyboards. He’ll pull paper out of the printer and sketch the first draft of a slide deck. Sometimes he uses Claude or ChatGPT to refine it before making the final, digital version. But he does his best thinking in an analog format.
The process has served him well for years, and he sees no reason to modernize.
“There’s a tendency for folks to jump at a solution when there may not even be a problem that exists," he says. “AI is a very powerful tool—it’s a hammer and that doesn’t mean everything is a nail."
This isn’t to say everyone needs to brush up on penmanship and stop using AI calendars. But if AI power users pick and choose what is worth automating, then we could all be a bit more discerning.
Write to Callum Borchers at callum.borchers@wsj.com
topics
