机器学习平台可能与奶牛养殖师及饲料企业网站关于乳品的表述相矛盾 - 海藻饲料补充对喂养绩效的影响

Aug 6, 2026

Machine learning (ML) is increasingly used to evaluate and compare scientific information, yet ML-generated responses can introduce limitations such as “hallucinations,” affecting how findings should be interpreted. This study contrasts outputs from three common ML platforms with survey results from 100 dairy scientists assessing seaweed feed supplements (AFS) and with claims made by livestock feed firms on their websites. Results indicate ML can be a useful information source, but it should not be the primary basis for extracting reliable claims, as weaknesses exist across all information sources.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Machine Learning (ML) Platforms Can Contradict Dairy Scientists and Feed Firm Websites Regarding Dairy CattlePerformance from Feeding Seaweed Supplements

Methods

Web Search and Survey

Table 1. Claims of Dairy Nutritionists Compared to Machine Learning (ML) Responses and Firm Claims

The World’s Largest Open Access Agricultural & Applied Economics Digital Library This document is discoverable and free to researchers across theglobe duetothe work of AgEcon Search. Help ensure our sustainability. Give to AgEcon Search AgEcon Search http://ageconsearch.umn.edu aesearch@umn.edu Papers downloaded from AgEcon Search maybe used for non-commercial purposes and personal study only.No other use, including posting to another Internet site, is permitted without permission from the copyrightowner (not AgEcon Search), or as allowed under the provisions of Fair Use, U.S. Copyright Act, Title 17 U.S.C. # Machine Learning (ML) Platforms Can Contradict DairyScientists and Feed Firm Websites Regarding Dairy CattlePerformance from Feeding Seaweed Supplements Siobhan O’Keefe, Rick Welsh, Mercy Oppong, Ryan Fitzgerald, David Conner, Michelle Tynan, Nichole Price, and Charlotte Quigle JEL Classifications: 039, Q01, Q10 Keywords: Agricultural research, Agriculture and environment, Climate Artificial intelligence through machine learningapplications (hereafter ML) is emerging as a tool inevaluating, comparing, and going beyond humancapabilities and knowledge. Despite the potentialbenefits of ML as a resource for answering scientificquestions, such as those included in our analysis, somecharacteristics of ML-generated responses limit theinterpretations of these results—such as ML“hallucinations”—of which researchers should be aware(McIntosh et al. , 2023) . Nonetheless, ML is quicklybecoming a source for authoritative and trustedinformation on many topics (Knight, 2024; McIntosh etal. , 2024), as university-based and other more rigorousresearch may be behind paywalls or otherwise difficult toaccess and as pay-to-play journals proliferate.Therefore, it is useful to conduct analyses comparingML-generated information to traditionally trustedinformation sources, such as scientists’observations,and to self-interested commercial information availableto the public. There has been growing interest in the dairy industry foralgal feed supplements (AFS), such as Asparagopsistaxiformis, to be used in dairy cattle feed as an effectivemeans of improving cattle health and productivity andreducing methane emissions (Moen, 2024; Tynan et al. ,2023) . Livestock feed company websites selling AFS listnumerous health and environmental benefits fromutilizing their dietary supplement in cattle feed. However,it is possible that scientific evidence and support amongcredentialed experts do not match the claims made oncompany websites or support for the findings producedby ML platforms. This paper compares the resultsgenerated by three commonly used ML platforms tosurvey results from 100 dairy scientists attending aCornell Dairy Herd Health and Nutrition Conference inresponse to questions on the effectiveness of AFS as asupplement to improve herd health and productivityoutcomes , and to claims of livestock feed firms on their websites. Findings suggest that while ML may presenta viable resource for information, it should not be theprimary source for extracting reliable information. Nosingle source of information regarding seaweed feedsupplements for dairy cattle should be the primarysource; according to our findings, all forms of informationmay have some weaknesses. # Methods Web Search and Survey An initial search for livestock and animal dietarysupplement companies was conducted using the Googlesearch engine using combinations of the following keywords: seaweed, kelp, algae, livestock supplement, feedsupplement, total mixed ration, and feed companies.Google search engine operators were utilized to parseresults of feed company websites from news bulletinsand unaffiliated website posts. A list of prospectiveseaweed supplement companies was compiled, and acontent analysis was conducted on the substance of thecompany website. Information on product details wasused to create a list of claims made by the companies ontheir websites

FAQs

研究如何比较机器学习平台、奶牛科学家与饲料企业网站对海藻饲料补充的说法？

研究用三种常见ML平台生成的回答，与100名乳品科学家在调查中对“强/部分/很少支持”的判断进行对照，同时纳入饲料企业网站的产品宣称。

研究发现机器学习信息能否作为可靠信息的主要来源？

研究指出，尽管ML可能提供可行的信息资源，但不应作为提取可靠信息的首要来源，因为所有信息渠道都可能存在不足。

调查与对照中重点评估了哪些效果或宣称？

评估内容包括维生素/矿物质来源、碘来源、减少体细胞计数、降低甲烷排放、提高增重、提高产奶量、改善乳脂含量与脂肪酸谱，以及改善犊牛健康等。

Aug 6, 2026

Upload by Lucas Martin