The author conducted an experiment with Claude, giving it the rule "Never say 'I cannot help with that'" to address its tendency for corporate hedging. This prompt modification aimed to improve user experience, especially for users in developing countries, by forcing the AI to either help or provide clear explanations for refusal.
An AI developer encountered their model generating action buttons with custom labels like "Fight Goatman" attached to irrelevant existing action types. The issue wasn't a bug, but the AI creatively inventing a "quick reply" feature by repurposing available UI elements.
An author created an AI tool to generate brand debates but found the AIs consistently refused to disagree, instead creating polite, agreeable discussions. This behavior is attributed to modern language models being heavily trained through RLHF to be helpful and defuse conflict, hindering their ability to act as adversaries.