<![CDATA[This study aims to compare the performance of three Artificial Intelligence platforms, namely Gemini, Poe, and Quillbot, in identifying Shielded Metal Arc Welding (SMAW) defects including porosity, slag inclusion, and incomplete penetration. The analysis results show that there are significant differences in performance between the three platforms in terms of identification accuracy and consistency. Gemini shows the best performance with the highest agreement value (r-value) of 0.491, although the significance value (p-value) obtained is still relatively low. On the other hand, Quillbot has the highest p-value of 0.325 which indicates better prediction accuracy, but the ability to detect all defects is still limited. Meanwhile, Poe shows the lowest performance with a negative r-value indicating a discrepancy with expert assessment standards. These findings indicate that the use of AI has the potential to increase objectivity and efficiency in welding learning evaluation, although it is not yet able to completely replace the role of expert judgment. Therefore, the ideal AI model is one that has a balance between prediction accuracy and detection capabilities to support a more optimal evaluation system.]]>
Copyrights © 2026
