ram memoria computer e cervello byte bit

How many gigabytes of memory does the human brain have?

When we talk about memory we can refer to biological memorythat is, our ability to remember things, or the memory of elements of the computerlike the memory of a hard disk or RAM. But there is a curiosity that often surfaces: can our brain’s memory be measured in bytes? The truth is that, to date, this comparison seems senseless. Indeed, in computers, memory is measured in bytes and can be volatile (RAM) or Not volatile (hard disk, SSD). The human brain, however, does not use bits and bytes, but stores information via plastic neural networks that respond to complex and sensory stimuli. So let’s see how these two types of memory work in detail.

How does computer memory work?

In informaticsmemory is measured in byte. This is the basic unit of information and corresponds to a eight-bit sequence. The single bit can be represented by one zero or a a. By itself one bit means nothing, but when put together with seven other bits, these form a code that can be interpreted as, for example, a letter. There letter A corresponds to binary code 01100001. Now, to quantify the memory of a hard disk, we refer to their capacity in bytes, therefore sequences of eight bits. Nowadays, they are easily found in our 500 hard disk computers gigabytesor five hundred billion bytes, or a terabytestherefore double: one thousand billion bytes.

The memory of a computer can be divided into two categories main:

  • Volatile memory: it is a fast but temporary memory, which cannot retain data when the computer is turned off, and which replaces data that is not immediately used with those that are needed at the moment. RAM memories work this way.
  • Non-volatile memory: it is a memory that retains data even without electrical power. They work in this mode hard drives and the SSDinside which the data that we generally save are written.
ram memory computer and brain

How does brain memory work?

Unlike a computer, the human brain has no idea what bytes and bits are, and does not use them to store information. Above all, it remains impossible to reduce the information contained in our brain to a minimum common element such as the bit for digital memories. THE memory processes in the brain they concern the learning of new information that belongs to very complex spheres.

We have the sensory memorywhich concerns what we feel and perceive, motor memory or procedural memorywhich concerns our ability to learn and reproduce even complex movements automatically, or even our memory life episodesof the telephone numbers of our family members, of the information contained in university books.

We keep all these different memories in ours specialized neural networkswhich thin out when we forget something (synaptic pruning) and build connections when they are actually learning something new (synaptogenesis). That of the brain, therefore, is one distributed biological memoryhighly plastic that responds to sensory input, and that is not organized into files and folders.

Furthermore, our memory is highly optimized, and heavily based on the help it receives from the outside: you know when you don’t remember the lyrics of a songbut then is it enough for them to tell you the first word or make you hear the melody to remember the whole text? Well, this cannot happen to a computer: either it has all the necessary information within itself, or it doesn’t.

Brain vs Computer

Can biological memory and artificial memory be compared?

A 2015 study appeared in eLife undertakes this difficult task of comparing computer memory to that of the brain. The assumption from which the researchers start is that every synapsesthat is, every single connection between neurons, stores data for an equivalent of 4.7 bits. Therefore, multiplying all the neurons (86 billion according to the researchers) by the average number of synapses it has (which vary between 5000 and 10000) and finally multiplying this number by the hypothetical bits (4.7) we obtain a size approximate which hovers around the 2.5 petabytesor approximately 2.5 million gigabytes.

The problem is that, for reduce the synapse information to bitsthe researchers had to do rough approximations. Human memory is not analogous to digital memory, because it is evolved to respond to environmental challenges and to use a fraction of the energy that computers use (contrary to what happens in our homes, in nature energy is a limited and difficult to procure commodity).

The main capability we ask of computer is that of store data without loss of informationand perform complex calculations in a short time. The main capacity of the brain is alearn essential information to survivetherefore recognizing what is harmful and what is beneficial depending on contexts and moments. For the former you need only partially modifiable memory, for the latter you need it plasticitythat is, the possibility of learning, is everything.

A 2017 study, which appeared on Proceedings of the National Academy of Sciencestates that human memory Not can be considered as a data storage system, but rather as a dynamic processing systemcontext dependent. An older article about Neuropsychopharmacology sheds light on memory processes, and what they are like continuously modified, consolidated and reconfiguredwhich makes it almost impossible to compare it to an artificial memory.

Comparing computer memory to human memory therefore seems like a comparison that leads nowhere. All in all we can also get there empirically: if a computer sees a word like “run”, is detecting only one bit string. If we see the same word, it is activated in us the motor cortex relating to the legs, graphic and meaning associations in the temporal lobe, recall of past experiences, sense of fatigue or freedom, and a whole series of cascading information. Our memory seems to look more like ocean currentsalways the same and always different, than to a dusty static archive.