As a person with an extensive background in ML, I feel similarly about CS grads attaching themselves to "AI" enterprises until they convince themselves they're

As a person with an extensive background in ML, I feel similarly about CS grads attaching themselves to "AI" enterprises until they convince themselves they're an "expert" without putting in the due diligence to understand the unifying math under the hood. They end up doing stupid things with their models, completely ruining their business proposition through their lack of understanding in how the ML theory makes the implementation make sense. And then they don't even know how to appropriately judge the quality of the outputs from the reported metrics. It's normies all the way down.