You can still picture a room you stood in fifteen years ago. Yet last Tuesday has already dissolved into a vague impression of work, messages, and dinner.
It feels like a strange design flaw: the distant survives while the recent slips away. But memory does not rank experiences by date alone. What you noticed, what a moment meant, how often you returned to it, and what reminds you of it now all help shape what comes back.
In a wide-ranging conversation with Raj Shamani on Figuring Out, memory researcher Nikolay Kukushkin offers a useful starting point: stop imagining a camera inside your head. Think instead about a living system, changed by experience and still changing when you remember.
Old is not the opposite
of memorable.
An ordinary week contains countless details that receive little attention. A distinctive afternoon can leave many more ways back: a place, a smell, a person, an emotion, a story you have told repeatedly.
Consider two dinners. One is a routine meal eaten while scrolling. The other is a reunion in a room full of familiar voices. The second has more personal associations to draw on. That does not guarantee perfect recall, but it helps explain why age alone tells us so little about a memory’s accessibility.
How long ago something happened and how easily you can retrieve it are different questions.
Kukushkin’s examples make this personal. In one demonstration, an image of a door sticks with the host because their earlier conversation had already made doors relevant. The same image does not arrive in the same mental context for every viewer. We bring a history to what we notice.
A passing moment.
What it connects to.
a way back.
A conceptual guide to the article, not a complete model of the brain.
A vivid memory can still
have moving parts.
Remembering can feel like pressing play. Yet recall is reconstructive: details may be incomplete, and later information or interpretation can affect the account we give. The sense of seeing a moment clearly is not an independent check that every detail is correct.
A study by Jennifer Talarico and David Rubin compared students’ recollections of hearing about September 11 with memories of an everyday event. Both became less consistent with the initial reports, while confidence remained higher for the emotionally significant memories. [1]
The lesson is modest but important. Certainty and consistency can come apart. It does not follow that half your life is invented, or that every act of remembering rewrites everything. The study does not establish either claim.
A memory can feel complete
without being a complete record.
That distinction matters in an argument. Two people can remember the same evening differently without either deliberately lying. If a detail matters, compare accounts with messages, notes, or other records. Leave room for “this is how I remember it” alongside “this is what we can verify.”
Sometimes the way back
is a smell.
You may fail to summon a place on demand, then encounter a smell that seems to reopen it. Research comparing sensory cues has found that odor-evoked autobiographical memories can feel more emotional and more transporting than memories prompted by visual cues. That feeling of return is not proof of greater accuracy. [2]
In the conversation, Kukushkin describes a new carpet’s smell bringing back a childhood computer room. It is his personal example, not a universal recipe. Your own cue might be a cooking aroma, a song, or the texture of a familiar object.
The practical idea is connection. When learning a name or a concept, attach it to something meaningful you already know. A memory palace gives this strategy a spatial structure: imagined locations become prompts for what you want to retrieve. The palace is a tool; the association does the work.
What brings a place back for you?
Think of a familiar place and one specific sensory detail. You do not need to recover a whole scene. Notice how a small cue can give an abstract memory a point of entry.
Give learning another
chance to happen.
Practice matters because the next encounter meets a system that has already been affected by the last one. Kukushkin uses this idea to explain why repetition is never an exact reset. But repeating more intensely is not the only option. Timing matters too.
Research on distributed practice supports spreading learning across separate sessions. The useful gap depends on the material and how long you need to retain it; there is no single interval that fits every task. [3]
Retrieval adds another useful ingredient. In experiments with prose passages, trying to recall material produced better delayed retention than repeatedly studying it under the tested conditions. [4] For an article like this one, a simple application is to close the page, explain the central idea, then reopen it to check what you missed.
What does it mean for a kidney cell to “remember”?
A 2024 study by Kukushkin and colleagues found that non-neural human cell lines showed stronger, more sustained molecular responses to spaced chemical stimulation than to massed stimulation. This is evidence about cellular signaling. It does not mean that kidneys store your childhood scenes, or that a cell-culture timing protocol is a study timetable for people. [5]
What fades is not always
what failed.
We often treat forgetting as memory’s enemy. The conversation asks us to consider a different possibility: retaining every detail would not necessarily help us live, learn, or decide. Some information loses relevance. Some is difficult to access. Some was never strongly encoded in the first place.
Sleep also belongs in this picture. It supports learning and memory consolidation; its role is more complex than simply deleting the day’s unimportant material. A better study plan leaves room for sleep as well as practice. [6]
Letting go is not the same as staying silent.
The transcript discusses stepping away from repetitive rumination. That should not become a rule against talking about painful experiences. Structured psychotherapy can help people process trauma and reduce symptoms. Persistent intrusive memories deserve support, not blame for “remembering too much.” [7]
There is also no established four-year “installation period” during which babies cannot form memories. Childhood amnesia is a real phenomenon, but its explanation remains under investigation. A 2025 study found hippocampal activity associated with later memory-based looking behavior in infants, suggesting that difficulty retrieving early experiences later in life may be part of the story. [8]
Build a relationship
with what you learn.
Kukushkin repeatedly returns to a garden as a metaphor for the mind. Its appeal is the timescale: care happens through repeated encounters, not a single act of control. A garden changes even when you are not standing over it.
For learning, that suggests a manageable approach. Pay attention to one idea. Give it an association. Try to retrieve it later. Check the gaps. Come back again. None of this promises perfect recall, and none requires you to monitor every thought.
Make a connection.
Link new information to an example or idea you already understand.
Return, then retrieve.
Use another session to recall and check, rather than only reread.
Keep certainty flexible.
When details matter, verify them beyond the feeling of remembering.
Why might an old memory be easier to recall than last week?
Pause and answer in your own words before opening the note.
Reveal the idea
Age is only one factor. Attention, personal meaning, associations, repeated retrieval, and the cues available now can all matter. An accessible memory is not necessarily accurate in every detail.
You do not need a perfect record of your life to keep learning from it. You need ways to return, the patience to practise, and enough openness to let the picture change.