(A follow up to this post. )
The development of AI tools for performing statistical analysis and training machine learning models is concerning to me. Will the tools hit a point where a user can simply state an analysis (like a prompt in a LLM), provide some data, and the AI will take over? Can those results be trusted?
If feels like this is already a concern in statistics. Pre software a statistician would have to use tables and perform hand calculations, thwarting all but the most committed lay person. Those methods acted as a gate. With the PC and software, anyone could push some buttons and get some p values. That is how things are now: I am not confidence that the typical person with a course in statistics really knows how to use statistical methods correctly.
Would be interesting to find commentary from the 1970s or 1980s on how statisticians felt about non experts doing statistics.