← NewsAll
Wikipedia at 25: academic neglect may now pose a new risk
Summary
As Wikipedia turns 25, the article argues that many academics have long resisted contributing to the site while relying on it privately, and warns that large language models trained on Wikipedia risk displacing the volunteer-driven, transparent knowledge resource.
Content
Wikipedia reaches its 25th anniversary amid renewed debate about its relationship with academia. The article argues that many scholars have treated Wikipedia with scepticism while students and the public use it widely. It notes a 2005 Nature finding of comparable accuracy with Encyclopaedia Britannica and subsequent research showing many specialist pages can be reasonable resources. The author warns that large language models trained heavily on Wikipedia now surface answers without attribution, which could reduce incentives for volunteer contributors.
Key points:
- A 2005 Nature investigation reported Wikipedia's accuracy was comparable to Encyclopaedia Britannica, and follow-up studies find specialist articles (for example in health and psychology) are often reasonable alternatives to professional sources.
- The article says large language models trained on Wikipedia can provide instant, unattributed answers, creating a parasitic dynamic that may undermine contributions to the encyclopedia.
- Some institutions and projects already involve students and scholars in editing Wikipedia (examples cited include UCSF's medical-school editing courses and the GeneWiki project, which contains over 11,000 pages and is viewed tens of millions of times annually), but broader institutional recognition of such work is not settled.
Summary:
The article contends that academic reluctance to engage with Wikipedia has left a widely used, transparent knowledge commons vulnerable to displacement by AI systems that recycle its content without attribution. The author calls for greater scholarly engagement and institutional recognition of contributions; whether such changes will be adopted more broadly is undetermined at this time.
