First of all, the human brain doesnt store information like computers (8-bit)
It stores only small pieces of information it uses as references to create more information, but that information is not stored, but it is manipulated in what you can call the brain's random access memory.
For example, the alphabet:
ABCDEFGHIJKLMNOPQRSTUVWXYZ
How do we know them? How does our brain keep them?
The truth is: It doesn't
The brain only stores extremely simple information and simple "commands/processes" (the term is incorrect, but since no scientist has coined it yet, i cant state it

)
The thing with words is similar.
Let's say your brain's "storage" has many "levels":
I will explain how your brain can also "store" an outstanding amount of information using a special kind of compression and an extremely complicated algorithm.
- Minor R.A.M (Things someone just told you for example)
- Random Access Memory (What you just remembered)
- Major R.A.M (What you barely studied for a test)
- Memory (This level includes information that is integral and it is similar to the actual RAM on a computer but it is actually stored much much longer)
- Storage (This is where the alphabet goes)
- Processing Queque Minor (After figuring something out and trying to remember it again, this is the "processing" level)
- Processing Queque (What you're trying to figure out)
- Compiler (Compile raw data)
- Major Compiler (This is the level of the brain that expands and translates raw data)
- Referencer (You differenciate and compile everything according to this referencer. It allocates data)
- Data (We cannot "understand" any of this since it is an algorith that can be decoded/compile by the Major Compiler, and recompiled closer to what we percieve when it reaches the memory)
- Raw Data2 (Simplified data)
- Raw Data (The simplest form of data)
Basicly, the "Raw Data" level refers to the neurons that can store multiple references to data with each connection.
Data like (1,0) in the brain for example can be compiled to create a certain image or video depending on the neuron, the connection, the region, the reference, and the allocated process information that corresponds to it.
1,0 ====> j3r034jf4c093ggjq04t493f93jfw9skcwneuwfefkwlf4 (It's an unknown analog so i'm inserting random characters)
j3r034jf4c093ggjq04t493f93jfw9skcwneuwfefkwlf4 ====> fn438hg89fn4oigoijlgfekn32ihflh3ih3li4hfoinciej84h8th58h3ifo34gh53hf4ilfn3ijf85hvlknrlkjgojlkrklgfjlgfiuch4lwe.hnfvnrueihglnvsd.joe89y5hj 5gnc4 5oc3hwoirfhq3;w5uto54un45h;jsf;;3poipj43pjt33b34gn5ogno.nfgldihh78ghie47hgildrhkgnlrgshrjeigjivnueoi5jenlvoijmigjtgjjryhtwneiejfoj4theetheneoifjoiho4hoirefmrjo5ih4glrnkl.g
And that data is further compiled with the compilers and processors until it ends up with a result that fits our Perception.
The brain can store data refering to a 9.3 trillion images and videos just by storing data like 1,0 in exactly (taking into consideration the word trillion)
This calculation is correct: 10 neurons
Why?
Each connection doesnt refer to one piece of information.
Networks are what compile information and put togethor the algorithms to output more and more data until it is fully compiled to fit our perception.
The amount of information (by quantity) that we can store is equal to this formula:
# neurons!/ Radical 2
That's 100,000,000,000! / Radical 2 (the exclamation mark is factorial)
If 70! = a number with 100 digits, imagine how huge the above number would be.


This formula refers to the number of possible neuron networks in the brain.
I divided it by radical 2 to make it look kewl


The factorial part are used to find ALL cases ans possibilities when it comes to connection.
--------------
Now that you have an insight on how much information the brain can actually store, i think you should correct the term "exabyte" with something like:
(100,000,000,000!^2 / Radical 2) Megabytes