"Participants preferred answers that took longer to receive, as they believed the AI was 'thinking' or showing 'deliberation.' This perception is invaluable for AI companies."
"Unlike other products, users judge AI responses similarly to human responses, assuming slower answers are more thoughtful, which can lead to misconceptions about AI capabilities."
"The researchers advocate for 'context-aware latency,' suggesting that simple questions should receive quick answers while complex ones should feature slight delays to enhance perceived thoughtfulness."
Research indicates that users prefer slower AI responses, interpreting them as more thoughtful. A study with 240 participants showed that delays of two, nine, or 20 seconds led to higher satisfaction with answers. This perception contrasts with typical product evaluations where speed is favored. The researchers recommend implementing 'context-aware latency' in AI chatbots, providing quick responses for simple queries and slight delays for complex ones, termed 'positive friction' to enhance user experience and perception of AI deliberation.
Read at Computerworld
Unable to calculate read time
Collection
[
|
...
]