A recent study highlighted by Ars Technica has shed light on the behavior of AI-powered search engines, revealing that they tend to leverage a broader spectrum of sources compared to traditional search algorithms like Google. The study, titled ‘Characterizing Web Search in The Age of Generative AI,’ was conducted by scholars from Ruhr University and the Max Planck Institute.
The researchers found that unlike conventional search engines that typically prioritize popular websites, the AI-driven counterparts exhibited a tendency to reference less mainstream sources that often fall beyond the top domains tracked by tools like Tranco. For instance, the Gemini search engine notably cited sources outside the top 1,000 domains, marking a departure from the norm observed in traditional search outcomes.
To gauge the behavior of these AI search engines, the researchers extracted queries from various sources such as ChatGPT, AllSides, and Amazon’s top-searched products list, covering diverse content categories. The findings suggest that while AI search engines may not always align with top Google search results, they offer a more varied selection of sources, potentially enriching the search experience for users seeking unconventional or niche information.
Source: Ars Technica