When we think about memory, we often assume that a superior memory means remembering more things for a longer time. That’s how we think of computer memory, too: the more, the better. But biologically speaking, memory did not evolve simply to store information. It selects what information to store and what to discard — a form of pattern recognition. So, superior memory means not just more information successfully stored but the most relevant information successfully selected.
I first confronted this tension a few years ago in the unlikeliest corner of science — while studying insulin-like hormones in Aplysia californica, a species of sea slug, in Tom Carew’s lab at New York University. The hormones were released after the slugs enjoyed a meal and did several things at the same time:
- They promoted sugar clearance, just like the hormone insulin does in humans.
- They enhanced neuroplasticity and neuronal growth.
- They slowed the animals down.
The first effect is the canonical response to food, evidently preserved across the tree of life in a strikingly intact way. The insulin-like hormones told the cells of the sea slug to absorb the nutrients it had consumed. In the lab, we could even cause the release of these hormones in the sea slug with an anti-diabetic drug used in humans. It is one of those examples that exposes the unity of life on Earth.
The second effect is about memory enhancement, which fits with the first effect. Since food was successfully found, the sea slug presumably wanted to form a long-term memory, and the insulin-like hormones helped with that. The combination of these effects made sense.
The last effect was odd and a bit ironic given that sea slugs are already extremely slow. Still, the difference was clearly seen when they were recorded on camera for many hours. Watching and analyzing those videos, even at 20 times the original speed, was probably the most boring experiment of my life. And since we could block this effect by using inhibitors of insulin-like receptors, it wasn’t just about digestion or the meal physically weighing the sea slug down. It must be said that these creatures can consume an impressive amount of nori per body weight — but that was not the issue.
Instead, the effect was a coordinated food coma induced by dedicated hormones. Evolution had built the response into the slugs. But why did the animal need to become unresponsive, reduce its neuronal excitability, and stop moving after a meal? Why would the animals have evolved this seemingly counterproductive response to food?
The only explanation that I could think of was that memory is expensive. It takes energy — as does moving around and exploring your environment. However, once food is found and consumed, memory must be prioritized over active exploration. The sea slugs didn’t need to experience new things; they needed to preserve the preceding experience for future gain. And so they shifted energy away from motion to memory.
This was back in 2019 when we didn’t know about GLP-1 agonists like Ozempic — hormones related to insulin that cause food comas while also promoting neuroplasticity and memory. I wonder if the entire combination of effects would have made sense right away if we had done the study today.
When we were working on the study, we debated with my boss, Tom Carew, about this “expensive memory” explanation. He questioned the idea that learning requires so much energy as to syphon it from behavior. I saw his point: At least on the surface, it seems that tinkering with the fine configuration of neurons and synapses — that’s how learning creates memory — should cost exponentially less energy than, say, muscle movement, like how a computer uses less power than an electric car.
But computers can, in fact, consume copious amounts of energy. Similarly, there is reason to believe that biological memory can be “expensive” enough to require energy appropriation from other sources. For example, fruit flies have two forms of long-term memory: One is more stable, while the other is cheaper and requires less energy. If the first one wasn’t too expensive to use all the time, why have two?
We also know that organisms make trade-offs related to information processing based on energy use. Yeast, for example, balances their energy budget with the fidelity of cellular signaling. When they run low on food, their interpretation of environmental signals becomes “looser.” If information processing was cheap, there would be no need to scale it down in such situations.
Maybe memory is also a luxury, and reduced motion provides a way to afford memory when it matters most.
The mnemonic economy
The bizarre food coma in sea slugs really affected how I think of memory. Memory may prove so expensive that it requires, in certain situations, an organism-wide appropriation of energy. In that case, brains must face similar economic considerations in other, less extreme cases. The cost-benefit of forming a memory is constantly evaluated against an organism’s energy budget.
For us humans, the choice of some patterns over others is even more critical than for a sea slug. We process vastly more information per unit of time than a mollusk does. Too many things pass through our brains at any given moment, and to make sense of them, they must be generalized and abstracted, meaning other things must be forgotten. Remembering everything would not only be too expensive. It would be useless.
I think we forget about this because we don’t perceive that our memory is limited. We don’t run out of memory like computers do — suddenly, you can’t remember anything new. However, we do remember different things to a different extent, which is our brain’s way of triaging the endless flow of information. Even if we treat our memory as a limitless hard drive, our brains quietly sort our mental files into “to be memorized” and “to be forgotten.”
Is there anything practical to be gained from this insight?
The learning we usually attempt — like trying to cram an entire textbook the weekend before finals — is unnatural. The brain’s goal is to pick out a salient pattern, and the textbook is not what it considers salient.
Instead, the best way to “improve memory” is to listen to yourself. Use your brain’s powers — curiosity, inspiration, even boredom — as tools. When curious, learn as much as you can. When inspired, drop other activities. When bored, change your focus to something else. And don’t forget to take breaks — just as in the sea slug, taking time to rest after a study session allows your mind to digest the new information, enhancing memory and slowing down forgetting.
On the other hand, if memory is expensive and selective, then perhaps we should think more about what we allow to enter our brains. I suspect that in the future, we will develop stricter societal standards for mental hygiene and learn to treat junk memory like junk food.
For example, at least a third of my memory is occupied by TV commercials from the 1990s. Could this space have been used for something else? I will never find out, and it may not matter much. But if someone back then told my parents the commercials would actually outcompete other information from my brain, I bet they would never let me see a TV again.
This article What food comas in sea slugs teach us about memory is featured on Big Think.
The post “What food comas in sea slugs teach us about memory” by Nikolay Kukushkin was published on 03/10/2025 by bigthink.com