A look at efforts by Google DeepMind, Anthropic, xAI, and other companies to make their chatbots funnier and keep users engaged with them at home and at work
Anthropic, xAI and other artificial intelligence companies see humor as an important feature for their services.
Context & Ripple Effects
The push to make chatbots more entertaining extends earlier coverage of LLMs entering computational humor, moving humor from a niche capability toward a product-engagement feature.
It also follows the conversational-product model associated with Character.ai’s user experimentation with language-model chatbots. The notable shift is that major model developers are treating personality and tone as relevant in both consumer and workplace use.
First-order effects
- Google DeepMind, Anthropic, xAI and peers face a more explicit product-design task: make chatbot interactions engaging, not merely useful or accurate.
- Users encounter assistants whose humor and conversational style are being tuned for longer, more repeatable use at home and at work.
Second-order effects
- Chatbot competitors are pressured to differentiate on voice, personality and interaction quality alongside model capability, particularly where underlying features look similar.
- For workplace deployments, humor becomes part of the trade-off between a more approachable interface and a tone that remains appropriate for professional settings.
Third-order effects
- If engagement-oriented personalities become standard, conversational AI may compete increasingly as a companion-like interface rather than solely as a question-answering tool, raising the relevance of anthropomorphic-AI and companion-governance frameworks.
- The durable differentiator could shift toward the quality of ongoing user relationships with assistants, though usefulness and reliability will still constrain how far personality design can travel into work.
The trend: AI companies are turning chatbot personality—including humor—into a competitive layer for retaining users across consumer and workplace contexts.