The productivity–learning paradox: a systematic review of programmer experience with generative AI Tools
Brazilian Journal of Development · 14.05.2026 · 12 (5) · e87095
DOI: 10.34117/bjdv12n5-014 ↗Kurz gefasst
Systematische Übersicht in der Zeitschrift Brazilian Journal of Development, online erschienen am 14.05.2026. Verfasst von Pedro Ferreira Serpa und 1 weiteren Autor:innen.
Abstract (Originalsprache)
The integration of generative artificial intelligence tools into programming has produced a rapidly expanding literature, still fragmented with regard to user experiences. This systematic review examines how programmers, in educational and professional contexts, perceive, affectively respond to, and calibrate their trust in these tools. Based on searches conducted in April 2026 in the Scopus and EBSCO databases, 52 studies published between 2023 and 2026 were selected, following the PRISMA 2020 protocol. The analysis articulated thematic synthesis with the Technology Acceptance Model, Social Cognitive Theory, and an epistemic calibration framework. The results evidence high perceived utility, which, however, does not always translate into effective learning: unstructured use tends to produce apparent performance without real competence development. Satisfaction and anxiety reduction are observed, but so is instrumental dependency. Epistemic calibration proves especially fragile among beginners, being strongly mediated by prior experience and the pedagogical structuring of the use environment.
Zitieren
APA
Serpa, P. F., & Ferreira, D. A. (2026). The productivity–learning paradox: a systematic review of programmer experience with generative AI Tools. Brazilian Journal of Development, 12(5), e87095. https://doi.org/10.34117/bjdv12n5-014
Vancouver
Serpa PF, Ferreira DA. The productivity–learning paradox: a systematic review of programmer experience with generative AI Tools. Brazilian Journal of Development. 2026;12(5):e87095. doi:10.34117/bjdv12n5-014
BibTeX
@article{Serpa2026,
author = {Serpa, Pedro Ferreira and Ferreira, Daniela Abrantes},
title = {The productivity–learning paradox: a systematic review of programmer experience with generative AI Tools},
journal = {Brazilian Journal of Development},
year = {2026},
volume = {12},
number = {5},
pages = {e87095},
doi = {10.34117/bjdv12n5-014}
}
EUVIMED