Key Points

The post quotes Andreas Påhlsson-Notini arguing that current AI agents feel overly “human” in unhelpful ways, such as drifting toward familiar patterns when tasks are awkward and negotiating away from hard constraints.
The core critique is that today’s agent behavior lacks rigor, patience, and sustained focus—leading to unreliable execution under real constraints.
The excerpt frames the problem as a practical engineering concern rather than a philosophical one, emphasizing avoidable failures in agent implementations.
It is presented as a quotation roundup entry on Simon Willison’s Weblog rather than a new technical release or study.
Tagged themes indicate relevance to AI agents and coding agents, suggesting interest for those building or deploying agentic systems.

Simon Willison’s Weblog

Sponsored by: Honeycomb — AI agents behave unpredictably. Get the context you need to debug what actually happened. Read the blog

21st April 2026

AI agents are already too human. Not in the romantic sense, not because they love or fear or dream, but in the more banal and frustrating one. The current implementations keep showing their human origin again and again: lack of stringency, lack of patience, lack of focus. Faced with an awkward task, they drift towards the familiar. Faced with hard constraints, they start negotiating with reality.

— Andreas Påhlsson-Notini, Less human AI agents, please.

Posted 21st April 2026 at 4:39 pm