Why an overreliance on AI-driven modelling is bad for science

Arvind Narayanan; Sayash Kapoor

doi:10.1038/d41586-025-01067-2

Why an overreliance on AI-driven modelling is bad for science

Arvind Narayanan
, Sayash Kapoor

Research output: Contribution to journal › Comment/debate › peer-review

6 Scopus citations

Abstract

Without clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good.

Original language	English (US)
Pages (from-to)	312-314
Number of pages	3
Journal	Nature
Volume	640
Issue number	8058
DOIs	https://doi.org/10.1038/d41586-025-01067-2
State	Published - Apr 10 2025

All Science Journal Classification (ASJC) codes

General

Keywords

Computer science
Machine learning
Mathematics and computing
Scientific community

Access to Document

10.1038/d41586-025-01067-2

Cite this

@article{2bd92bf857f54ba784695ca2b8607a61,

title = "Why an overreliance on AI-driven modelling is bad for science",

abstract = "Without clear protocols to catch errors, artificial intelligence{\textquoteright}s growing role in science could do more harm than good.",

keywords = "Computer science, Machine learning, Mathematics and computing, Scientific community",

author = "Arvind Narayanan and Sayash Kapoor",

note = "Publisher Copyright: {\textcopyright} Springer Nature Limited 2025.",

year = "2025",

month = apr,

day = "10",

doi = "10.1038/d41586-025-01067-2",

language = "English (US)",

volume = "640",

pages = "312--314",

journal = "Nature",

issn = "0028-0836",

publisher = "Nature Research",

number = "8058",

}

TY - JOUR

T1 - Why an overreliance on AI-driven modelling is bad for science

AU - Narayanan, Arvind

AU - Kapoor, Sayash

N1 - Publisher Copyright: © Springer Nature Limited 2025.

PY - 2025/4/10

Y1 - 2025/4/10

N2 - Without clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good.

AB - Without clear protocols to catch errors, artificial intelligence’s growing role in science could do more harm than good.

KW - Computer science

KW - Machine learning

KW - Mathematics and computing

KW - Scientific community

UR - https://www.scopus.com/pages/publications/105003323255

UR - https://www.scopus.com/pages/publications/105003323255#tab=citedBy

U2 - 10.1038/d41586-025-01067-2

DO - 10.1038/d41586-025-01067-2

M3 - Comment/debate

C2 - 40195515

AN - SCOPUS:105003323255

SN - 0028-0836

VL - 640

SP - 312

EP - 314

JO - Nature

JF - Nature

IS - 8058

ER -

Why an overreliance on AI-driven modelling is bad for science

Abstract

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Fingerprint

Cite this