Imagine having an LLM do a translation of daily news into "simple english", much like wikipedia has: https://simple.wikipedia.org/wiki/Simple_English_Wikipedia

Imagine having an LLM do a translation of daily news into "simple english", much like wikipedia has: https://simple.wikipedia.org/wiki/Simple_English_Wikipedia https://simple.wikipedia.org/wiki/Simple_English_Wikipedia the results are not free of political bias, but may well highlight it in a starkly hilarious way. you might do human training at that level but then you've only created a newly biased model.