<![CDATA[This community service program aimed to strengthen physics teachers’ capacity to design deep learning by integrating artificial intelligence (AI) and wetland contexts. The partner was the Physics Subject Teachers’ Association (MGMP) of Tanah Laut Regency, South Kalimantan. The activity used a participatory, practice-oriented training format consisting of needs mapping, conceptual strengthening, guided practice in pedagogical prompting and AI-output verification, contextual physics learning design, and evaluation. Knowledge achievement was measured using 10 multiple-choice items, while participant satisfaction was collected using a 10-item five-point questionnaire. Descriptive analysis showed an average knowledge mastery of 97.27%; seven items reached 100% correct responses, while three items—deep learning principles, effective prompting, and AI ethics—reached 90.91%. Satisfaction averaged 4.59 out of 5 (91.78% of the maximum score). Overall satisfaction received the highest mean (4.78), whereas understanding of AI utilization obtained the lowest, but still positive, mean (4.33). These findings indicate high post-activity mastery and strong participant acceptance while highlighting the need for continued hands-on mentoring in pedagogical AI use, prompt design, output verification, and responsible AI practices.]]>
Copyrights © 2026
