In short when a program runs, it allocates memory for the data it is using. Then the garbage collector which you can think of a ‘program’ (though not in a strict sense, just a part of the runtime), takes care of freeing the memory which is no longer needed. In languages with manual memory management, like C or C++ it is up to the programmer to properly (de)allocate memory, which might result in issues like dangling pointers (references to already freed memory, which might cause unpredictable behavior and compromise your whole system’s security) or memory leaks (gradually increasing memory usage over time in absence of user action that would prompt that, like e.g. opening more and more browser tabs, which is also partly why in the past you’ve often had to restart your PC every once in a while to free some trashed memory. In Go most of the stuff is done by GC and only code that uses the unsafe package or relies on external resources, like reading a file or database connection needs to have programmatic termination of access to these (usually by calling something like defer db.Close() ).

Also Go is a nice balance between low and high level, one of such examples is the use of pointers. They might be complicated for most beginner coders, but in reality there is seldom any use of uintptr, double pointers and even slice (list) pointers, but it has a tremendous performance amount since when you do some operation, especially loops where without a pointer you would copy the data you’re iterating through by value instead of just it’s memory address.

Which is especially important with large sets of data, where 160 bytes might seem miniscule with one call, but if you loop over 1 million records, that’s 160 MB (for example in some database used by municipal authorities of a large city). That’s one of the reasons some databases like CockroachDB were written in Go.