For more reliable AI, academics should edit Wikipedia

Letter to the Editor

Email Bluesky Facebook LinkedIn Reddit Whatsapp X Access through your institution Buy or subscribe Wikipedia, the “encyclopedia anyone can edit”, is a fundamental training and reference data set for large language models (LLMs). As LLMs grow in popularity, Wikipedia’s content — both high and low quality — is likely to have more impact. Although the long-term effects of the technology remain uncertain, Wikipedia’s importance and reach has clearly grown since the rise of LLMs. Access options Access through your institution Access Nature and 54 other Nature Portfolio journals Get Nature+, our bes

related reading

• Wikipedia:Signs of AI writing - Wikipediaen.wikipedia.org
• The machines are fine. I'm worried about us.ergosphere.blog
• Thoughts on AI in academiatheinfinitesimal.substack.com
• Andrej Karpathy on X: "LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating" / Xx.com
• Opening a conversation on responsible environmental data science in the age of large language models | Environmental Data Science | Cambridge Corecambridge.org
• Writing for LLMs So They Listen · Gwern.netgwern.net
• What We’ve Learned From A Year of Building with LLMs – Applied LLMsapplied-llms.org
• Meet the Humans Building AI Scientists - Asimov Pressasimov.press
• Ai2: Truly open breakthrough AIallenai.org
• Customizing Karpathy's LLM wiki for fighting diseasekamens.com
• When AI Co-Scientists Fail: SPOT-a Benchmark for Automated Verification of Scientific Researchalphaxiv.org
• Latent Space: AI for Science | Latent.Space | Substacklatent.space