We need to stop allowing AI training off stolen datasets

OSU is one of the universities that has been more strident that GPT4 is a commercial product and as such “should” be limited to commercial users, and I agree with that. In my time as a researcher, I don’t think I have ever used any academic version of any software for anything other than my own research. Wildly off base estimate, I’d say over half of free version “academic” licenses are used for personal projects or businesses. <br><br>I don’t know for sure what the OSU data set used for things like GPT-4, and I know our grad students use it for a lot of their research, but if the commercial product will include the use of all that data, then that is clearly not in line with the ethical intents of things like the commons, and I think we should stop allowing that. No commercial products should be allowed to use data from “free” licenses for their commercial products.