Utilization of Artificial Intelligence Based on Deep Learning to Improve Problem-Solving Abilities (Research in Science Learning for Fourth Grade Students of State Elementary School 2 Suci Kaler, Karangpawitan District, Garut Regency)

Tati Suryati (Institut Pendidikan Indonesia, Garut, Indonesia)
Uman Suherman (Institut Pendidikan Indonesia, Garut, Indonesia)
Asep Rohayat (Institut Pendidikan Indonesia, Garut, Indonesia)

Jurnal Locus Penelitian dan Pengabdian
Vol. 5 No. 6 (2026): JURNAL LOCUS: Penelitian dan Pengabdian

The development of Artificial Intelligence (AI) technology has brought major changes in the world of education, including in science learning in elementary schools. This study aims to determine the effect of the use of Deep Learning-based AI on the problem-solving abilities of fourth-grade students at State Elementary School 2 Suci Kaler, Karangpawitan District, Garut Regency. The study used a quantitative approach with a quasi-experimental design method and a Non-equivalent Control Group Design model. The research sample consisted of an experimental class and a control class. Data collection techniques were carried out through problem-solving ability tests, observation, and documentation. Data analysis used normality tests, homogeneity tests, and t-tests. The results showed that the use of Deep Learning-based AI had a significant effect on improving students' problem-solving abilities. The average posttest score of the experimental class was higher than that of the control class. AI-based learning also increased students' active involvement in the science learning process. This study concluded that Deep Learning-based AI is effectively used as a 21st-century learning innovation in elementary schools.