Human Effort for Human Attention: The Etiquette of AI at Work

As AI-generated text becomes more common in the workplace, a new etiquette is required to avoid burdening colleagues with low-effort content. The author suggests that if you expect someone to spend time reading your work, you must prove you put effort into it first. This means reviewing, labeling, and adding context to AI outputs rather than simply forwarding them.

Key Points

• AI-generated content is becoming ubiquitous in software engineering, leading to 'AI fatigue' among readers.
• Sending unread AI output to teammates is disrespectful because it devalues the recipient's time and attention.
• The core principle for AI collaboration is to demonstrate human effort whenever requesting human attention.
• Practical steps include labeling AI content, adding human commentary, and pre-reviewing AI-generated code.

Sentiment

The overall sentiment is cautiously supportive of the article. Hacker News largely agrees that asking for human attention requires ownership, context, and some visible care, especially when AI makes it easy to generate long or polished-looking output with little personal investment. The disagreement centers on whether effort is the right ethical standard: a substantial minority argues that quality and accountability are what matter, while others worry that effort-based norms can become performative or exclusionary.

In Agreement

• The article's principle is a familiar reciprocity norm: the effort invested in a request should be proportional to the attention expected from the recipient.
• Low-effort questions and unreviewed AI output shift cognitive labor onto helpers, reviewers, and teammates, which can damage team dynamics and create bottlenecks.
• Preparing a clear, contextual request is itself useful work because it often reveals the answer or clarifies the real problem before anyone else is interrupted.
• AI-generated communication is especially frustrating when it is verbose, generic, inaccurate, or pasted from a trivial prompt the recipient could have run themselves.
• Using AI is not the core offense; failing to edit, understand, personalize, and take responsibility for the output is the real breach of etiquette.
• Automated systems can let powerful parties cheaply impose work on others, encouraging an escalating cycle of generated requests and generated responses.

Opposed

• Respect should be measured by usefulness, clarity, and correctness, not by how much effort the sender visibly spent producing the message.
• AI is a human tool, so AI-assisted output can still be treated as human output when the sender stands behind it.
• Visible effort can become performance theater, rewarding people who make work look hard or legible instead of people who create actual value.
• Some commenters expect people to care less about whether text is AI-generated if the result becomes clearer, kinder, and more effective than ordinary communication.
• Strict norms around demonstrating effort may be less inclusive for people who are anxious, inexperienced, blocked, or unsure how to begin asking for help.
• Recipients can also use AI to summarize or simplify incoming text, so the burden of dealing with generated content may not always fall entirely on humans.