英语轻松读发新版了,欢迎下载、更新

Can we trust AI? | READER COMMENTARY

2025-05-31 17:13:57 英文原文

Recently, there have been reports of not only a MAHA (Make America Healthy Again) report from the Department of Health and Human Services that had artificial intelligence “formatting errors” but reports of AI-generated student papers that have citation errors. The Naval Postgraduate School Citation Guide states, “Generative AI tools can fabricate citations to sources that do not exist” and “create plausible-sounding statements that may not be true or may be biased.” We have been reading about using AI to, for example, find a cure for alzheimer’s disease. How could we trust information about important subjects when the statements in such a report may not be true and/or the citations are to sources that may not exist?

If AI has all the information in the world to draw from, why is it making up citations and fabricating “plausible-sounding statements?” I don’t get it.

— Denise Lutz, White Marsh

Add your voice: Respond to this piece or other Sun content by submitting your own letter.

关于《Can we trust AI? | READER COMMENTARY》的评论


暂无评论

发表评论

摘要

Reports have surfaced of AI-generated errors in a MAHA report from the Department of Health and Human Services and student papers with citation inaccuracies. The Naval Postgraduate School warns that generative AI can fabricate non-existent sources and produce misleading statements, raising questions about trustworthiness in AI-generated information on critical topics. Writer Denise Lutz expresses confusion over why AI makes up citations and false claims despite having access to vast amounts of data.

相关新闻