If we use it for data analysis, what about AI detection? Journals or university may not accept our work. Please comment, if somebody faced AI detection issue after using this software.
AI detection isn’t a real thing, no one can really detect AI text and anyone trying sell that as a service is a fraud. Of course there’s a lot of common idiosyncracies in AI writing (see the ‘delve’ controversy) such that if you use it all verbatim you can easily tell when someone is using AI to create writing that isn’t theirs. So while there’s no structured approach that exists, in singular instances where there is sufficient evidence in writing style, it is possible to identify AI work. In any case, we don’t condone cheating in an academic setting using AI tools. We recommend that you have permission from your professor to use AI in your work or that you cite any work that an AI largely produced.
Personally how I use AI for schoolwork is to create outlines for papers, create plots, verify analysis results, check my proofs or talk through approaches etc. and I end up doing the final writeup myself each time. I also ask my prof in each class what the guidelines are around AI in class, and usually they’re pretty supportive of using it like an academic tutor or a smart writing assistant. My brother’s prof this past semester even recommended the entire class to use AI tools to help them for programming assignments.
For a more Julius-centric example, we recently had a user use Julius to help create an entire poster presentation for a Biology graduate student conference. Their work was accepted by the Journal since they cited Julius as a source in their bibliography.
2 Likes
To clarify, do you mean just using AI to run a statistical analysis and then using the results of that analysis in your publication? So you are more or less concerned that the journal itself would not accept the results as it is from AI?