Just for fun, and probably just this one time, here’s a copy and paste of the references and sources for the previous GPT 03 post on Chinese approach to AI in Ed.

This is ultra opposite to the distillation of a ton of information into one takeaway graphic such as Kris recently posted. Also I did not use the “Deep Research” function at GPT either…which might access even more sources than the GPT o3.

Of note: There are a lot of sources from .cn webpages cited. This brings up a question about credibility of sources that AI uses for search.

Seems like a CATCH 22…because if research doesn’t use .cn sources, it is likely to miss a lot of crucial information when queries are about that nation, but if it does use those sources then the results become more “iffy” and “questionable”.

This is a problem not just with certain nations, including Russian propaganda attempts to corrupt AI search, but also of just the general social media BS fully within the USA. Information sources are not universally “neutral sources” and that’s seemingly going to be an ongoing problem that could perhaps blow up into some general corruption of AI credibility.

This could occur in similar processes to the way much mendacious social media has severely damaged the idea of online information being actually “truth”. That’s probably a worthy question to ask Sam Altman et al, next time someone might get a chance to do so: “Sam, how does OpenAI avoid training corruption?”