Daniel Douster consulted his colleagues on matters large and small. If you need a rectangular logo converted to a square, send a message to one of our designers. If you want to build a new dashboard, you’ll probably set up a meeting with an engineer.
These days, all Deceuster requires is opening Claude or ChatGPT. You can often get what you need within seconds. “We are accomplishing more than ever before,” he told me. But lately, he’s lamented the cost to his productivity.
Now that I’m no longer dependent on my coworkers for this type of work, I estimate that I interact with them about 50% less than I used to.
“It’s sad to see a loss like this,” said Decester, marketing director for the nonprofit Zion HealthShare. “I’m an extrovert. I like to interact with people. I want that kind of interaction.”
Deceuster recognized early on that major changes were afoot as AI tools permeated corporate America. “People are increasingly choosing to work alone,” says Jessica Leaf, an incoming management professor at Wharton who is researching the impact of AI on teamwork.
Signs of tension are already showing. In January, Cisco found that its employees, its most active AI users, trust their teams less than intermittent users. This is probably because power users spend more time on themselves and less time with their colleagues. The company concluded that “AI can unintentionally create isolation when deployed individually rather than in groups.” Coaching platform BetterUp has found that some employees have come to rely on AI for the kind of feedback they previously sought from mentors and managers. These employees were more likely to report less team collaboration, higher rates of burnout, and a greater desire to quit their jobs.
“We are social animals,” says Kate Niederhoffer, chief scientist at BetterUp. Being sociable is “more than fuel; it’s how we survive.”
AI is unraveling the social structure of work. Despite all the monotony and humiliation of our work, what made it bearable and even meaningful was the way it united us with others every day. We found mentors, best friends, work wives, and sometimes real wives. We vented about bad bosses, gossiped after meetings, and united to accomplish things we couldn’t do on our own. Nowadays, such interactions are disappearing.
“If we don’t think about this, we risk making work more isolated and fragmented,” Leaf says. “We just combine our inputs in a way that’s more like an assembly line than a vibrant workplace.”
Early in American history, work was primarily an independent endeavor. The majority of people were farmers and artisans who worked for themselves. The Industrial Revolution changed that. Companies suddenly needed large numbers of people to operate new machinery in their factories. And coordinating all this workforce required another class of workers with increasingly specialized skills: managers, accountants, engineers, clerks, and salespeople. Americans go from having no colleagues. a lot Colleagues – Colleagues with whom you spend most of your waking hours. As a result, work has become more social.
This was especially true for white-collar jobs, where people had to combine different specialties. This required them to constantly communicate, give feedback, negotiate priorities, diagnose problems, and brainstorm solutions. As the world becomes more complex, so does the need for collaboration within the ever-expanding enterprise. Technologies like email, cell phones, Slack, and Zoom have facilitated that collaboration.
For all the humiliation of our work, what made it bearable and even meaningful was the way it brought us together.
Next came AI. As increasingly powerful models improve white-collar productivity, I predicted that companies large and small would soon have fewer employees. This means we will revert to a world with fewer colleagues (AI-driven layoffs in recent months at companies like Meta, Block, and Cloudflare suggest this is happening sooner than I expected). But it’s only recently that I’ve started to wonder if the AI revolution is going to get even worse, not only shrinking companies’ workforces but also weakening the relationships between the employees who still work within them.
If previous office tools made it easy for colleagues to connect with each other, AI appears to be replacing those connections entirely. “ChatGPT and tools like it give us an alternative way to accumulate knowledge that would otherwise be shared privately,” says Reif. “Give people the option to opt out.”
Now that I’ve found myself opting out, I understand why people opt out. Over the past two years, I’ve been experimenting with different ways to incorporate large-scale language models into my work. I still write the manuscript myself. But as I refined what I had already written, I realized that AI could serve as a sounding board before handing off the story to my editor, Zach. I’m going to ask him the same questions I used to ask him. “Does this section explain too much?” Is the conclusion bad? Should I put this sentence here or there? Good. ChatGPT is always available, so you don’t have to wait for Zak to reply. If you disagree, you can discuss it without it getting heated. You don’t have to worry about annoying the other person with endless, persistent questions.
“These interactions used to involve mutual trust and showing vulnerability,” Leaf told me. “ChatGPT is frictionless.”
For a while I was happy with the low friction method. But as I reported this story, I began to wonder what I had to lose in order to lessen Zach’s inconvenience. It’s true that I don’t talk to him anymore. Had I missed an opportunity to learn from him? Had I lost the ability to work through disagreements? Do I feel less close to him? I thought about what being less needy was doing to the professional relationships that matter most to me, and what that meant for how I ultimately felt about my work.
Some people may not think that AI’s antisocial tendencies are such a big deal. White-collar jobs may have become too collaborative in recent years. With non-stop meetings and constant Slack messages, many of us have a hard time getting any real work done. In any case, those interactions were rarely deep or meaningful.
Getty Images; Tyler Lee/BI
In some cases, fewer interactions can make your workplace more comfortable. Peter Pan, co-founder and chief technology officer of Creao AI, an AI agent platform, says that overhauling his company’s workflow to offload most tasks to agents has reduced conflict among employees. He used to spend 60% of his time managing employees, but about 10%.
“My relationship with my co-founders and engineers has actually improved because we don’t have to constantly argue with each other,” he told me. “Arguing with each other is not a very constructive way to build relationships.”
However, these points overlook something important. Small interactions that used to be mandatory at work helped teams work better together, often without anyone even noticing. They built the goodwill needed to help co-workers overcome the disagreements that inevitably arise. Equally important, it helped employees keep track of what others were working on and prevent duplicating each other’s work.
“AI gives us this incredible speed,” Leaf tells me. “But there is still a lot of interaction and strategic coordination required to ensure velocity is directed to the right targets.”
Perhaps this is not the case for solopreneurs who have found ways to leverage AI to stay in business. But in almost any other business, people will need to coordinate effectively and trust each other, even if the day-to-day tasks become more isolating.
So what could be the solution here? According to Betterup’s Niederhoffer, one option is to use these tools in the following ways: strengthen Relationships with colleagues, not replacing them. Some people are already doing this, asking Claude to help them draft a confidential email to send to a colleague or role-playing a difficult conversation they’re about to have with a mean boss. BetterUp found that people who use AI in this way are actually interacting more with their direct reports and with colleagues from other parts of the company than they did a year ago. “You’re learning new ways to interact with people,” Niederhoffer says.
We may have to work a little harder to find these connections and find meaning in the relationships we have.
Another solution is to reclaim the social time taken away by AI. Even if you have fewer opportunities to connect through your daily work, you can still create other ways to build bonds. Carol-Lyn Jardine, who advises marketing executives on AI through her consultancy Clarity & Motion Collective, says she can do most of her work without having to rely on two business partners, thanks to a suite of AI systems supporting her. But she decides to keep in touch anyway.
“If I don’t stop and talk to them about what they’re seeing in their clients, we don’t really gain insights and learnings that can be applied to all of our clients,” she tells me. “We’re going to have to work a little harder to find those connections and find meaning in the relationships we have.”
The challenge for all of us is to navigate these new trade-offs. This means that while less interdependence in the workplace can quickly increase productivity, you also need teams that work well together and work that allows you to feel connected. DeCoaster, an extroverted marketing director, says it’s a difficult balance to strike. He makes it a habit to walk up to his colleagues’ desks and talk to them, rather than skipping over them. That has allowed him to regain some social time, but he knows that such efforts are limited by the constraints of his workplace. “It’s tough because everyone is trying to get the job done,” he tells me. “What you don’t want to do is be seen as a distraction and lose your job.”
That’s why it’s ultimately up to our companies, and the people within them, to find systematic ways to preserve the social aspects of our work. And we actually know how. Post-pandemic, companies realized how important daily employee interaction was, and many began bringing employees back to the office. Some people have tried something else. They continued to work remotely and figured out how to replicate interactions with a distributed workforce. They rolled out mentoring programs, sent employees off-site, and mandated more frequent one-on-one meetings between managers and employees. They intentionally designed interactions that once happened by chance.
Big changes often happen this way. We embrace something new, discover its unintended consequences, and redesign our habits and institutions to correct them. Deceuster says the current AI boom is reminiscent of the early days of social media. It took me years to realize how isolating these platforms are, especially for children. Only now are schools, parents and governments finally taking steps to protect them. How long will it take for our workplaces to figure out how to benefit from AI’s productivity gains without alienating us from each other? “We still don’t know what we’ve unleashed, or even how to use it effectively,” Decaster says.
That’s especially important now, when our lives are already so lonely. We stopped going to church. We stopped going to the union hall. We stopped going to rotary clubs, sports leagues, shopping malls and movie theaters. But we still go to work. Work kept us connected to others even as all other community institutions declined. If even that is taken over by AI, we will become more efficient and lonelier than ever before.
Aki Ito I’m Business Insider’s chief correspondent.
Business Insider’s Discourse articles provide perspectives on the most pressing issues of the day, powered by analysis, reporting and expertise.
