It is a condition in which a person who is ignorant essentially so foolish that they simply don’t grasp how ignorant they are. A shorter version of the Dunning-Kruger Effect would be the common saying “A little knowledge can be a dangerous thing.”
This is not to be confused with simply being argumentative or contrarily over an issue. It is a condition where a person quite literally cannot see what folly they are spouting, because they imagine that their ability on a particular issue is vastly greater than it actually is. Such as when a person has read a single Layman’s book on quantum physics … and then proceeds to tell the people with degrees on the subject they are wrong.
It is extremely common on the internet, where people wade in with what little knowledge of a topic that they have. Imagining that they know a huge amount, because they are oblivious to what they don’t actually know. And if we are honest, I think we've all done this on some topics. Yes, even me … I’ll pre-empt any “What about you, Bill?”
It can often make all arguments seem impossible … because without loads of supporting documentation it can sometimes be extremely hard for anyone else to determine just who is really an “expert” on a topic and just who is spouting forth from ignorance.
The most novel thing about the D-K effect is that people with real abilities underestimate their competence and understanding. I imagine that type of person wouldn't bother to talk and joining the conversation at all.
Where do we draw the line for when to start expressing our knowledge? After we get degrees on the topic? Even that doesn’t make us experts. So we should just feel free to express and be open to learning from our mistakes and from others.
I think the key is to admit one’s ignorance. We, as a society, need to let people know that it’s OK not to know something, and that saying “I don’t know,” is a perfectly acceptable response (and arguably preferred) than to making up bullshit. Even “experts,” especially experts will be the first to admit when the questions, answers, data fall outside their particular area.