The psychology behind why AI shows it's working
Major AI answer engines now display their reasoning process, showing users which sources they consult and assumptions they challenge, a shift rooted in psychological research on perceived effort and value.
In 2025, leading AI answer engines such as Claude, ChatGPT, and Gemini began revealing their internal processes to users. These systems now display which websites they search, documents they open, and assumptions they reconsider during responses. Anthropic, the company behind Claude, attributes this transparency to three stated reasons, though the guide suggests a fourth, less-discussed motivation may also be at play.
A 2011 study by Harvard professors challenged the assumption that faster always means better in customer service contexts. The research involved 266 participants using a travel search site similar to Skyscanner or Kayak. One group saw only a loading wheel, while another saw a live scrolling list of airlines being searched, with fares appearing as they were found. Participants rated the transparent loading wheel as providing 8.1% higher perceived value, even when the wait was longer.
The study further tested user preferences by offering a choice between a faster site with no transparency and a slower site with a transparent loading screen. Despite identical results, participants consistently preferred the slower, transparent option. This pattern suggests that visibility into effort influences user satisfaction more than speed alone.
Follow-up research in 2022 by Dimitrios Tsekouras, Ting Li, and Izak Benbasat reinforced these findings. In two studies involving 306 and 294 participants, those who saw a seven-second loading wheel with phrases like 'calculating results' rated the recommendation quality significantly higher, even when the outcomes were identical to instant results.