what is using namespace std exactly?

A namespace is like a container for distinct names (originally from C). Essentially, when you use libraries of functions, you have the problem that common names may overlap. For example, if you have a library that executes the maximum value of a vector, it may be called max(). Well, since that name is so short and generic, there's a high possibility that some other library already has a function named max(), so it's "enclosed" in a namespace.

Your namespace can be called whatever you want. If you call it let's say "math_lib", then you can access your functions within that namespace by specifying the namespace then the function using the namespace::function notation (and yes, you can have several layers of this). As you might imagine, within a source file, you may use the functions within a namespace over and over again, and rather than having to specify the namespace every single time, you can use the using namespace directive, which essentially tells your compiler that you're using the specified namespace for the whole function, method, or file (depending on the scope of your statement).

So when you're using namespace std, you're literally using the namespace named std. Meaning you don't have to say std::vector (for example) every time you use a vector and can simply omit the std:: portion.

In summary, a namespace is made to contain a number of functions. Usually libraries use them to make sure their functions don't overlap with other functions having the same name/prototype. It was a bigger problem in C because of the lack of classes to act as containers (and because of the way prototypes were handled), but it's still around in C++ essentially for the same reason of creating containers.