Unfortunately, I don't think teaching logic will get you what you want here. Fake news may be false, bit it is often (mostly) logically consistent - this is part of the power.
(One challenge: though neural networks generalize very well, we still lack a decent theory to describe them, so much of the field proceeds by intuition. This is both cool and extremely bad. “It’s amazing to me that these very vague, intuitive arguments turned out to correspond to what is actually happening,” says Ilya Sutskever, research director at OpenAI., of the move to create ever-deeper neural network architectures. Work needs to be done here. “Theory often follows experiment in machine learning,” says Yoshua Bengio, one of the founders of the field. Modern AI researchers are like people trying to invent flying machines without the formulas of aerodynamics, says Yann Lecun, Facebook’s head of AI.)
One of the most interesting aspects of the field - we don't have robust ways of predicting what will work without trying it.
I would love to see a writing format with something like the following structure:
1) A TLDR of Twitter length
2) A paraphrase of the key concepts and point argued in the article
3) The actual full-length article
This way you could ramp up at each stage and decide if it's valuable to keep reading.
Most of what I've seen is in English but there are some second-level services popping up that promise an API for chatbot translation, e.g. Cyrano (http://cyrano.unbabel.com/).
Dragon Skin (https://en.wikipedia.org/wiki/Dragon_Skin) claimed to withstand multiple hits without loss of performance, although there was some controversy around it.