AI agents take over CAE analysis; humans shift to oversight

The demonstration at the Cybernet Systems event illustrates a broader trend in engineering software: CAE tools, once operated through graphical interfaces, are now being driven by code generated by AI agents. The speaker's point that CAE analysis is 'essentially coding' underpins this shift, as AI can handle the scripting and execution, leaving humans to define the physical problem and interpret the results. This mirrors changes in software development, where AI code generation has already begun to alter how work is done, suggesting that CAE analysts may similarly move toward higher-level responsibilities.

The demo also showed the AI's ability to handle ambiguity: when the prompt said 'a load of 10N', the AI checked whether that meant total force, distributed load, or stress, showing a level of understanding beyond simple instruction-following. This capability allows it to adapt, like switching from Ansys MAPDL to CalculiX without changing the workflow, which strengthens the case that AI agents can manage routine analysis tasks. However, the human retains a critical oversight role, especially in defining the analysis plan and reviewing the final results, which the speaker emphasized as essential skills for CAE engineers going forward.