Artificial intelligence learns to reason
Science · 21.03.2025 · 387 (6740)
DOI: 10.1126/science.adw5211 ↗Kurz gefasst
Editorial in der Zeitschrift Science, erschienen am 21.03.2025. Verfasst von Melanie Mitchell. Beteiligte Einrichtung: Melanie Mitchell is a professor at the Santa Fe Institute, Santa Fe, NM, USA. Hauptthemen laut MEDLINE: Generative Artificial Intelligence, Problem Solving, Thinking, Large Language Models. Die Arbeit wurde bisher 21-mal zitiert (Crossref).
Abstract (Originalsprache)
Julia has two sisters and one brother. How many sisters does her brother Martin have? Solving this tiny puzzle requires a bit of thinking. You might mentally picture the family of three girls and one boy and then realize that the boy has three sisters. Or you might figure out a more general rule: Any boy in the family will have one more sister than any girl. In other words, the answer to such a puzzle isn’t something you immediately know, like Paris is the capital of France; it requires reasoning, a central feature of human intelligence, and one that large language models (LLMs) like GPT-4, for all their impressive behavior, struggle with.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Melanie Mitchell is a professor at the Santa Fe Institute, Santa Fe, NM, USA.
ORCID-Profile: Melanie Mitchell
Zitieren
APA
Mitchell, M. (2025). Artificial intelligence learns to reason. Science, 387(6740). https://doi.org/10.1126/science.adw5211
Vancouver
Mitchell M. Artificial intelligence learns to reason. Science. 2025;387(6740). doi:10.1126/science.adw5211
BibTeX
@article{Mitchell2025,
author = {Mitchell, Melanie},
title = {Artificial intelligence learns to reason},
journal = {Science},
year = {2025},
volume = {387},
number = {6740},
doi = {10.1126/science.adw5211}
}
EUVIMED