Sapere aude assisted (not replaced) by AI.
Neural networks are useless for this task, because they don't allow us to see the chain of reasoning that comes to the conclusion. Neural networks only show the result, but not the reason for it.
The reason is however what is essential, because it is what makes the system trustless ('trustless' in the system theoretic sense of the word https://steemit.com/cryptocurrency/@r... )
and decentral, not reliant on authority. If it only delivers the result without a reason for why it is correct, then we have to trust the flawnessness of the machine, i.e. another appeal to authority: The authority of the designer.Only if it explicitly spells out its reasoning and makes it humanly understandable for us, there will be no need anymore to trust the machine, since the source of the reasoning has no bearing on its soundness, only the reasoning itself, which can be verified by us ourselves if it has been made explicit by the machine.
What does this mean more concretely?
Imagine for instance a computer program that can identify logical fallacies in natural language as it appears in media articles, that has a mechanism to verify basic facts like "Paris is the capital of France" and that gives us the source from where it got this fact, that can double check claims being made in an article with a database of historical knowledge, scientific knowledge, political knowledge, possibly a video and photo database etc., that can remind us of our cognitive biases and identify and alert us to rethoric in language that is likely to trigger these cognitive biases, that can show us statistics on demand etc.
There could be many other features to assist human reason. What's important though is that the program's reasoning needs to be explicit and transparent so that we can verify it at each step.
Like a calculator that instead of merely showing you the result of a calculation also shows you its derivation.
So far the internet has been used by malicious actors to exploit the human weaknesses and spread propaganda to make us believe falsehood and control us like kettle appealing to our base, unconscious instincts.
https://washingtonpost.com/news/posteverything/wp/2018/05/21/fascism-is-back-blame-the-internet/?tid=ss_tw&utm_term=.afa8bdd4ed3b
But the internet cannot be undone and it cannot be censored.
The only way to fight back against the oligarchs who want to create a dystopian anarchocapitalist world ruled by corporations that control humans like kettle and to regain sovereignity is to develop tools that help and strengthen human reason instead of exploiting human irrationality.
This is the proper continuation of the enlightenment project in the age of the internet.