跳到正文
Hacker News:AI 熱帖·· 24 天前AI 評分49

我用 AI 偽造了一篇科學論文:事情經過如何

I Used AI to Fake a Scientific Paper. Here's What Happened

AI 導讀

作者用 AI 偽造科學論文,兩篇稿件均被自動化學術誠信檢查標記並拒稿。評論指出,第一篇因引用資料虛構被拒,第二篇則因過度依賴預印本被拒;有評論稱,這項實驗事先獲 IEEE Access 編輯及 IEEE 同意,目的是進行紅隊測試。

正文

Image 1Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I Used AI to Fake a Scientific Paper. Here's What Happened (ieee.org) 22 points by pseudolus3 hours ago | hide | past | favorite | 10 comments help Image 2Lerc12 minutes ago | next[–] >When one model refuses or stalls, I switch models, rephrase prompts, and ask what the next step would look like “hypothetically.” I wonder If the hypothetically criterion gives the models carte blanche to make up references. reply Image 3jszymborski1 hour ago | prev | next[–] Sidenote: way to flex your "writing studio". reply Image 4lifeisloving16 minutes ago | parent | next[–] A shed, but a cool shed ig. reply Image 5dist-epoch46 minutes ago | prev | next[–] There is still no paper or notable result in gender studies published by an AI. reply Image 6kirth_gersen51 minutes ago | prev | next[–] Save yourself a slog. The automated checks found that the citations didn't exist and flagged the "papers". A better title would have been "Automated Systems seem to hold back the AI Slop papers. For now." Sweet writing studio, though. reply Image 7scotty7940 minutes ago | prev | next[–] In other news people prompted a model to pick a problem to solve, solve it and write a paper about it. And it did and the paper contains novel contributions. https://www.reddit.com/r/OpenAI/comments/1xgqv27/two_physici... reply Image 8VCFundedGenYer1 hour ago | prev | next[–] Please, even if it is an experiment, don't do this. We don't need more AI slop in the world. reply Image 9pizza23445 minutes ago | parent | next[–] This experiment was agreed with the managing editor of the multidisciplinary and open-access journal IEEE Access, and IEEE itself - it was red teaming, essentially. reply Image 10paxy1 hour ago | prev[–] TLDR - both papers got flagged and rejected by automated integrity checks. reply Image 11red75prime19 minutes ago | parent[–] The first paper (free-tier chatbots) due to fake references, the second paper (a specialized AI-research tool) due to overreliance on preprints. reply Image 12 Consider applying for YC's Winter 2027 batch! Applications are open till November 2. Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact Search:

來源:Hacker News:AI 熱帖 · news.ycombinator.com