Humans don't have an API<br>CodeRabbit is now in the Claude Marketplace!Learn more

Humans don’t have an API<br>by

David Kravets

June 10, 2026<br>7 min read

June 10, 2026<br>7 min read

The side effects of talking to machines<br>What some emerging literature saysThe AI "social forcefield"<br>A warning flare about human relationships

The parts of work that don’t scale<br>Humans don't have an API<br>The human cost of efficiency<br>Rewriting the interface

Back to blog

Share

Cut code review time & bugs by 50%<br>Most installed AI app on GitHub and GitLab<br>Free 14-day trial<br>Get Started

Catch the latest, right in your inbox.

Subscribe

Add us your feed.<br>Catch the latest, right in your inbox.

Subscribe

Add us your feed.

Keep reading

Meet the Overview page: The pull request home page you actually wanted<br>The Overview page is a single PR level home that answers what a PR is and whether it can merge: one answer for the reviewer, one for the author.

Claude Sonnet 5 review: Should you switch?<br>A hands-on review of Claude Sonnet 5 after a week of real coding and code-review work: how it compares to Sonnet 4.6, what it costs, and who should upgrade.

The case for monorepos, and what they cost you<br>Monorepos keep shared code in sync but make every bad PR everyone's problem. Here's when to commit to one, when polyrepo wins, and how to review at scale.

Get<br>Started in<br>2 clicks.<br>No credit card needed

Your browser does not support the video.

Install in VS Code

Your browser does not support the video.

Are we treating coworkers like AI agents?

Take a look at your recent Slack messages, Google Docs comments, emails, or the transcripts of your last few video calls with colleagues. How many of them began with a greeting, or provided context for a sudden request?

Now look at your chat history with your preferred AI assistant.

One of the more curious features of the modern workplace is that these two columns of text can sometimes look surprisingly similar.

As generative AI becomes embedded in daily work, the line between how we communicate with software and how we communicate with one another can feel less distinct than it once did. Direct requests, immediate responses, and highly task-focused exchanges have become a routine part of interacting with AI systems. As organizations optimize for efficiency, it is worth considering how those habits may influence communication in the workplace as well.

The side effects of talking to machines

A common response to concerns about AI etiquette is straightforward. The system has no feelings, so how we speak to it doesn't matter.

However, behavioral scientists have long understood that habits formed in one context often spill over into others. So, does spending hours each day issuing commands to conversational AI systems change how we communicate when another human being is on the receiving end?

The evidence is still emerging. However, new research suggests that prolonged interaction with AI systems may influence interpersonal communication styles in subtle but meaningful ways.

Generative AI rewards directness. It responds instantly, stays focused on the requested task (unless there’s hallucinations) and converts instructions into results with remarkable speed. The interaction is efficient, goal-oriented, and largely free of the social rituals that characterize human conversation.

Over time, it is reasonable to wonder whether those expectations begin migrating into our human relationships as well.

Human relationships are built through repeated interactions that create trust, understanding, and shared context over time. These moments may appear inefficient when measured purely by output, yet they help build the goodwill and mutual confidence that make productive collaboration possible.

When we begin treating coworkers like conversational interfaces, like APIs, those relational investments become easier to skip.

What some emerging literature says

Because generative AI arrived so quickly, research on its social consequences is still catching up to the reality that millions of people are spending hours each day interacting with machines capable of producing human-like conversation.

Nevertheless, research is beginning to examine how those interactions may influence human relationships and workplace behavior.

The AI "social forcefield"

Researchers Christoph Riedl, Saiph Savage, and Josie Zvelebilova explored this phenomenon in their paper Cognitive Spillover in Human-AI Teams.

The researchers conducted two randomized experiments to examine whether interactions with AI influence subsequent human-to-human communication. Across both experiments, they found evidence of what they call "cognitive spillover," where the effects of AI exposure carried forward into later interactions between people. According to the authors, AI exposure influenced shared language, collective attention, shared mental models, and social cohesion.

The researchers describe this phenomenon as an "AI social forcefield." The term...