The Hidden Cost of Not Asking

Across industry conversations, a particular phrase keeps surfacing when people discuss AI adoption: “Now I don’t have to bug [someone].” Product designers skip researchers thanks to retrieval-augmented generation (RAG) tools surfacing insights instantly. Product managers generate mockups without designers. Engineers rely on automated scanners instead of accessibility teams.

The framing is positive — liberation from waiting, unblocked progress, independent problem-solving. And in many ways, it’s accurate. Yet the quick questions, small talk, and organic exchanges that AI automates away form the scaffolding of healthy team dynamics. The “bugs” being eliminated are often the very interactions that build trust, belonging, and culture.

The inefficiencies of interpersonal communication and daily interaction build the larger organism known as work culture. When AI disrupts these interactions, what is lost?

What disappears isn't just a 2-minute Slack exchange, but the whiteboarding session it sparks; not just a “quick question,” but the revelation of a fundamental misalignment; not just an accessibility review, but a mentorship moment. These information-exchange interactions double as the building blocks of workplace connection.

Evidence From Three Decades of Research

MIT’s Human Dynamics Lab found in 2012 that the strongest predictor of team productivity wasn’t formal meetings but the “energy” of informal communication — hallway chats, coffee breaks, quick questions. Teams with the most informal interaction achieved 35% more successful outcomes.

Google’s Project Aristotle (2015) studied over 180 teams and identified psychological safety — cultivated through frequent, low-stakes interactions — as the number one predictor of high performance. Not intelligence, not resources. Trust built through micro-moments, the exact micro-moments that AI usage erodes.

More recently, a 2025 study from Harvard, Columbia, and Yeshiva University examined AI’s effect on team coordination. The authors found that AI-driven automation decreased overall team performance and increased coordination failures, particularly in the short term and among low- and medium-skilled teams. Automation also reduced team trust.

When Connection Breaks, So Does Retention and Innovation

The downstream effects of disconnection are measurable. McKinsey’s Great Attrition research identified lack of belonging as a frequently cited reason for leaving. When informal micro-interactions vanish, belonging erodes — and people depart.

“Employee disengagement and attrition could cost a median-size S&P 500 company between $228 million and $355 million a year in lost productivity.”

— McKinsey

Innovation suffers too. A 2024 study by Korean researchers analyzing private-sector innovation concluded that weak ties — the bridging conversations with people you interact with occasionally — sustain innovative performance in companies actively pursuing technological breakthroughs. Breakthroughs often emerge not from your core team but from the people you would have “bugged” in the past. Eliminating those interactions to favor AI can narrow both the depth and breadth of innovation across design, coding, and content.

AI’s seduction is that it feels like pure gain until the team realizes they’ve become strangers who happen to work on the same project.

Practical Strategies for Balance

The solution isn’t rejecting AI — it’s deliberate integration. Use AI for dull, repetitive, high-volume tasks while reserving human cognition for higher-level problem solving, and actively design workspaces and online interactions to preserve human connection.

Use AI to Remove Toil, Not Interaction

Research published in Harvard Business Review (March 2026) studied 1,488 full-time U.S.-based workers and identified “AI Brain Fry”: acute mental fatigue and cognitive exhaustion from excessive AI use. The cognitive strain carries business costs — decision fatigue, error-prone work — and 34% of workers experiencing brain fry intended to quit.

But AI isn’t inherently harmful. Participants who used AI specifically to eliminate toil — the repetitive, unenjoyable parts of their jobs — showed not only 15% lower burnout rates but also reported “a higher degree of social connection with peers…because they had more time to spend ‘off keyboard.’” In this scenario, AI didn't disrupt team connections; it removed busywork that prevented colleagues from solving problems together.

Institutionalize Productive Friction

Steve Jobs understood the value of serendipitous collision. He designed Pixar’s studios so employees would have to bump into each other. Brad Bird later reflected: “Steve realized that when people run into each other, when they make eye contact, things happen.”

That design philosophy translates to the AI era through intentional practices:

  • Build AI tools that connect the team. Attach the names of original creators to internal agent outputs and direct seekers to those creators, connecting people to institutional knowledge — and to each other.
  • Spotlight successful team AI uses publicly. Highlight in town halls how teams have used AI to work effectively together, establishing a narrative that AI brings people together rather than pushing them apart.
  • Establish rotation programs. Even if AI lets product managers prototype, have them shadow designers anyway. Direct dialogue creates holistic understanding of each other’s craft beyond AI outputs.
  • Hold panel discussions on the evolution of work. Regularly gather cross-functional partners to discuss how work is changing, keeping intentional change open and top of mind.

Build Team Cohesion Through AI-inspired Laughter

Positive workplace humor is well documented as a bonding mechanism. AI can facilitate it through absurdity:

  • Bad UX Vibecoding Competitions. Give the team a silly prompt (“Design the worst volume control”) and 30 minutes to vibe-code a horrible solution. The team learns new AI tools, gets creative juices flowing, and laughs together.
  • Hyper-specific AI Creations. Generate an absurd image or AI-composed song about a funny work moment to spark smiles during workshops.

Eliminating toil, institutionalizing productive friction, and building cohesion through humor demonstrate how to integrate the best of human connection with AI’s capabilities.

The question isn’t whether to use AI. Contemporary workers have less and less choice. The question is: what kind of team do you want to become when AI is the newest teammate?

The Hidden Cost of Optimized Teams

Artificial intelligence promises a frictionless workplace: fewer errors, faster output, and predictable delivery. Yet the very efficiency that makes AI attractive can quietly erode the informal interactions that hold teams together. When colleagues stop asking each other for quick clarifications because an assistant supplies the answer, they lose the low-stakes exchanges that build trust.

Leaders who deploy AI with deliberate emotional intelligence can capture the productivity gains while protecting the relational fabric that makes teams resilient. The goal is not to resist automation but to recognize that some human friction is productive — it creates the repetition of contact that turns coworkers into collaborators.

Why Interruption-Free Work Isn’t Always Healthy

Research on psychological safety consistently shows that teams perform best when members feel free to ask “dumb” questions, admit mistakes, and challenge assumptions. AI removes many of those moments. An engineer who never needs help troubleshooting a build, or a designer who never asks a developer to clarify a constraint, loses a chance to build a working relationship. Over time, the team becomes a collection of individuals who share a dashboard but not a sense of mutual obligation.

This pattern is measurable. Studies on collaboration show that communication density — who talks to whom, how often, and through which channels — predicts team performance better than individual skill. Alexandra Pentland’s work at MIT found that teams with high “energy” and “exploration” in their communication patterns outperformed those that optimized for efficiency. AI-driven workflows that route questions to bots instead of humans can starve that energy.

The Trust Deficit in Automated Workflows

Crisis is where trust shows up. When an unexpected bug halts a release or a client pivots midway through a project, teams that have invested in casual, unstructured interaction handle the pressure with fewer breakdowns. Those moments of “unnecessary” conversation — sharing a frustration, telling a bad joke, asking for an opinion on a non-critical decision — are rehearsals for coordination under stress.

Fabrizio Dell’Acqua’s experimental research on teams using AI in a Super Mario task illustrates the tradeoff. Automation boosted individual performance and coordination among players, but the effect on team cohesion was not uniformly positive. Teams that leaned heavily on automated suggestions reported less interpersonal interaction, and their coordination gains flattened as task complexity rose. The findings suggest that while AI can handle well-defined subtasks, it struggles to substitute for the tacit knowledge that emerges from human back-and-forth.

Designing AI Adoption That Preserves Culture

Leaders can take concrete steps to keep AI from hollowing out team culture:

  • Audit for communication loss. Track whether AI tools are replacing inter-team messages. If a shared channel goes quiet, consider whether the automation is solving a problem or simply cutting a conversation.
  • Keep humans in the loop on judgment calls. Use AI for drafting, summarizing, or code suggestions — but require a human sign-off that includes a brief explanation shared with the team. This forces the knowledge transfer that builds competence across the group.
  • Protect unstructured time. Calendar blocks for brainstorming, retro discussions, or even coffee chats can offset the natural reduction in hallway conversations that remote and automated work already compressed.
  • Reward vulnerability, not just output. If a culture only celebrates the fastest, cleanest delivery, employees will hide the struggles that AI could have helped with — and they will stop revealing those struggles to each other.

The McKinsey research on value-building employees reinforces this: the top performers in any organization are not necessarily the most productive individuals, but those who amplify the effectiveness of others. AI that lets someone produce alone, without asking for help or giving it, can train teams away from that amplification behavior.

Emotional Intelligence as the Missing Variable

The technology is not the problem. AI tools do not inherently degrade culture; unexamined adoption does. Leaders who experiment, gather feedback, and adjust their AI rollout based on team outcomes — not just output metrics — can steer around the most corrosive effects.

When teams report feeling less connected, leaders should treat that as a critical, not cosmetic, issue. The stages of psychological safety developed by Timothy Clark show that teams move from inclusion to learner safety to contributor safety and finally to challenger safety. Each stage requires practice. AI that removes the need to practice small acts of asking and helping can leave teams stuck at the first stage — where they feel safe to exist, but not to challenge or grow together.

In the moment of crisis — the botched deploy, the sudden market shift, the lost client — the team with intact culture will respond faster and more intelligently than the one that optimized every interaction into a transaction. Efficiency is a tool, not a strategy. The organizations that thrive will be the ones that pair machined precision with the most human of abilities: knowing when “just asking” is not inefficiency, but an investment.