Wikipedia says its traffic is declining due to AI summaries
What's the story
Wikipedia, the world's largest online encyclopedia, has witnessed an 8% decline in human traffic over the past few months. The drop was noted by Marshall Miller of the Wikimedia Foundation in a recent blog post. The foundation tracks both human and bot traffic to better understand its audience and curb data scraping by third-party bots for commercial search and artificial intelligence (AI) purposes.
Traffic trends
Search engines provide direct answers to users
Miller attributed the decline to "the impact of generative AI and social media on how people seek information." He noted that search engines are now providing direct answers to users, often using information from Wikipedia. This trend, coupled with younger generations preferring platforms like YouTube and TikTok for information, has contributed to the decline in human traffic.
Volunteer impact
Fewer visits may lead to less content enrichment
Miller warned that the decline in traffic could have a negative impact on Wikipedia's volunteer base. He said, "With fewer visits to Wikipedia, fewer volunteers may grow and enrich the content, and fewer individual donors may support this work." The situation is ironic as most large language models (LLMs) use Wikipedia's datasets for training, but may end up harming one of their most trusted sources of reliable information.
Traffic boost
Wikimedia calls for help in driving traffic back
In light of these developments, Wikimedia is urging LLMs, AI chatbots, search engines, and social platforms that use Wikipedia content to help drive more traffic back to the site. The non-profit is also working on ensuring third parties can access and reuse Wikipedia content responsibly by enforcing its policies and developing clearer attribution standards. It is also exploring new ways to reach younger audiences on platforms like YouTube and TikTok.
Information
New project to help AI systems recently
Just recently, Wikimedia Foundation launched Wikidata Embedding Project. The initiative converted around 120 million open data points in Wikidata into a format that's easier for LLMs to use. The goal is to give AI systems access to free, high-quality data and improve their answer accuracy.