Does The Human Brain Have A Finite Memory Capacity?

For most of us, we don’t bother too much about how our brains actually achieve all this, just as we don’t spend that much time thinking about the complex inner workings of our home computers. We just know that we should do things like keeping dust out of the USB ports and emptying the recycle bin from time to time, just as we know that eating healthily and doing activities like puzzles are good for our brains. 

Another thing brains and computers have in common is memory storage. In fact, there are some theories about how our brains store memories that look more like a computer than modern computers do. You can purchase a laptop based on the size of its hard drive – but can we really treat our brains the same way? Does the human brain have a set, finite storage capacity, or is this where the metaphor starts to break down?

How many memories are too many?

The question of just how much memory the human brain can hold is one that has always intrigued. 

One strand of this is the much-debunked idea that we humans typically use only a small fraction of our brainpower, and that if we could just figure out how, we could unlock all this untapped potential – including, presumably, a superhuman memory. 

It’s a trope that pops up a lot in sci-fi, notably in 2014’s Lucy, which was based on the myth that we use only 10 percent of our true brain capacity. In reality, all the evidence suggests that we’re using basically all of our brains all the time. 

Without hope, then, of a miracle pill or device to reveal hidden memory vaults in the depths of our minds, just how much storage space could we be talking about?

It’s not magical; it’s not some hidden dimension of space that’s limitless.

Dr Nikolay Kukushkin

One widely cited estimate was published in 2015. Using synaptic strength as a proxy for information storage, a US-based team of scientists calculated that the brain is capable of storing 4.7 bits of information at each synapse, which pushed the overall estimate of its capacity up to 10 times more than was previously thought. 

Co-senior author Terry Sejnowski described the finding as a “real bombshell” in a statement at the time, adding that the entire capacity of the brain was now thought to be “in the same ballpark as the World Wide Web.”

Synapses are the junctions between nerve cells. Information has to get across these junctions to be transported between cells and onwards, sometimes across vast distances between faraway brain regions. This is achieved with the help of chemical messengers called neurotransmitters, including famous heavy hitters like dopamine and serotonin.

As well as being the site of intercellular communication, synaptic junctions have also long been considered a crucial storage station for memories. Don’t get us wrong – this is a complex area of science with many competing and complementary theories, so we still don’t have a definitive understanding of exactly what memory is and how it works. But many people agree that synapses have some role to play, so using them to measure memory capacity does make sense.

The 2015 study came after years of people believing that limitations on the types and strength of synapses meant the brain’s storage capacity was quite limited too. Since then, others have replicated the findings, including a 2024 paper that put the total capacity at somewhere between 4.1 and 4.59 bits. 

Even more recently, scientists at MIT have put forward a fascinating new theory as to how the brain achieves this impressive capacity. Their new paper proposes a major role for astrocytes – a population of star-shaped brain cells that have often been overlooked, but whose importance is beginning to be recognized. 

“Because a single astrocyte can connect to many neurons, and many synapses, it is tempting to hypothesize that there might exist an information transfer between synapses mediated by this biological cell,” said senior author Dmitry Krotov in a statement. 

What it boils down to is that even the approximately 86 billion neurons the brain contains are not enough to fully explain its memory capacity, so there must be some other cell doing some of the work. Future research will be needed to see whether the astrocyte hypothesis provides the missing piece of the puzzle. 

You might assume we’d be stopping our explorations there. Is the brain’s capacity finite? Studies suggest technically yes, but its size is so great that it’s functionally infinite. No one person is ever going to need enough memories to fill the entirety of the World Wide Web in one lifetime, right?

But when we started digging into this question, it quickly became apparent that it might not be the right question to be asking at all. 

How we misinterpret memory

“Everything is finite. It’s not magical; it’s not some hidden dimension of space that’s limitless.”

This is what Clinical Associate Professor of Life Science at New York University (NYU) Nikolay Kukushkin told us when we sat down to chat about all things memory. There was a moment when we feared this could be a very short interview indeed, but thankfully, Kukushkin quickly turned our conversation in a new and much more fascinating direction. 

“A lot of my students, when we start talking about memory, the first thing they want to know is how they can improve their memory, and what they mean by that is ‘how do I not forget anything?’ But that’s really not what memory does.”

“We think of memory as just a storage vehicle, as just a record button that you press and you hope that everything stays, but if it did that, it would be unusable. If we remembered everything, we would never be able to pull out anything useful from that memory.”

Sometimes I sit down in the end of the day, and I just feel like my head is boiling with words, with images, with things that I’ve read during the day, with people I’ve talked to.

Dr Nikolay Kukushkin

On your computer’s hard drive, no matter how disorganized and chaotic, you can always use the search function to locate exactly the file you need. It would be great if the human brain had something similar, like an elaborate filing cabinet where you could simply think of a query like “birthday party, 2007” and be presented with a Technicolor play-by-play of the whole event, but we all know that’s not how it works.

This is not a failing, though. It’s actually the system working as intended, as Kukushkin explained.

“Memory is not meant to be an accurate representation, a snapshot of an event that you’ve attended,” he told us. “It’s meant to be a fluid, updatable strategy for the future.”

We like to think that certain memories are sacrosanct, like some password-protected, non-overwritable file. Everyone remembers where they were on 9/11, right? Wrong. Studies have actually shown that we often overestimate the fidelity of our memories.

“What do we even call ‘remembering’?” Kukushkin asked. “You do remember. You remember very well. It’s just not accurate. […] It always comes through the prism of our perception.”

Famous experiments have demonstrated how different people observing an identical event can come away with radically different recollections. It’s one of the reasons why witness testimonies throw up so many problems for law enforcement. 

The 1974 Loftus and Palmer car crash experiment is one example. Participants were shown an identical video of a traffic collision. By changing the way they asked questions after the viewing, psychologists Elizabeth Loftus and John Palmer were able to extract very different testimonies from them about how fast they thought the cars had been traveling. 

Over time, as well, our memories fade, though we may not consciously realize it. If we asked you about a day 20 years ago, you might remember that you were wearing a blue shirt. But if we’d asked you the same question 10 years ago, you might just as confidently have remembered that you were wearing a red shirt. 

Pretty much no one is immune to this – just look at the Mandela effect, where whole groups of people can be convinced that they remember Pikachu with a black-tipped tail, or the infamous Fruit of the Loom cornucopia, despite the fact these things were never real. 

It’s one thing to accept this as the reality of our memories; it’s another thing to be comfortable with it. “It’s scary to think, how many of these layers of our personality are just, you know, hallucinated,” Kukushkin said, triggering an existential crisis in the interviewer that persisted long after our call was over.

But, he suggests, how our memories work is not necessarily at issue – they’ve served us well for millennia, after all. It’s more our comparatively modern obsession with keeping perfect records of the past that is at fault. 

“I think that the more natural state of being is this fluid memory, fluid history, but it’s just not compatible with how our society is structured. We have to preserve precise, specific information about the past, and that kind of conflicts with how our memories are set up to be.”

What is memory even for?

“We forget that memory is evolutionarily there to parse important things from unimportant things. And when we lose track of what’s important and unimportant, that is the decline of memory,” Kukushkin told IFLScience.

Our memories evolved, just like much of our physiology, to optimize us for an early Homo sapiens hunter-gatherer lifestyle. In that ancient environment, it was vital to our survival to recognize recurring patterns, to correctly separate friends from foes within the natural world – large, fanged mammals = scary; that brown mushroom = tasty; that red berry = deadly. 

But modern life is not like that at all. Now, we’re exposed to so much more information on a minute-by-minute basis than our ancestors would have ever had to cope with – and according to Kukushkin, it’s becoming more than our memories can handle. 

“Sometimes I sit down in the end of the day, and I just feel like my head is boiling with words, with images, with things that I’ve read during the day, with people I’ve talked to. And I realized that, you know, this is unnatural.” 

We couldn’t tell you with certainty what we had for dinner last night, but need a perfect four-part harmony rendition of a commercial jingle from 1997? We got you.

Kukushkin believes that it’s not a lack of absolute capacity that is the problem – rather, it’s the rate at which our modern information ecosystem requires us to overwrite our old memories with new ones. We’ve established that our brains have an incredible capacity, but even that may no longer be enough – and you can’t just plug an external hard drive in or start uploading to the cloud, unfortunately.

“I think if our problem is that our minds are still packed with memories, we have nowhere to write new ones, that’s probably still going to be a problem, even if you overwrite them,” he explained. “I really think that the only way to combat that is just to reduce the amount of inputs into your brain and be mindful of that. I think that as a society, we will have to do that.”

That means things like our 24-hour rolling news cycle, and constant scrolling through short-form social media content. Kukushkin would like to see more of a focus on what he terms “mental hygiene”, where people would be as mindful about what they’re exposing their brains to as they are the food they put into their bodies.

“It’s okay to get dopamine from social media. You just have to be aware that you’re going to pay for this later,” he suggests. “I think of dopamine and memory in general as a trade-off: yeah, you’re allowing something into your brain, but that will also have consequences, because it will stay in your brain and it will make you do things later.”

Perhaps one day, Kukushkin muses, society will approach regulation of things like advertising and children’s social media use with a more neuroscience-based mindset. It’s not necessarily about outright bans, but more about finding a “safe” level for consumption of attention-grabbing tech.

And in a small way, he’s already started experimenting with this in his own teaching at NYU.

“I banned all screens from the classroom, and it was the single best thing that I’ve ever done in my classes,” he told us. “It was night and day.”