Where the computer idea comes from
The comparison goes back to the birth of computing itself. In the 1940s John von Neumann described the brain in the language of logic gates, and Alan Turing showed that a simple machine could in principle carry out any definite procedure. By the 1960s and 1970s this had grown into the computational theory of mind: the idea that thinking just is a kind of information processing, and that mental states are something like the software running on the hardware of the brain. On this view, neurons firing are the physical steps of a computation, the way transistors switching are the steps inside a laptop.
The picture was enormously productive. It gave psychology a precise vocabulary, let researchers build working models of memory and vision, and underpins most of artificial intelligence. If the mind is a program, then in principle you could run that program on other hardware, which is why the metaphor and the dream of thinking machines have always traveled together.
What the metaphor gets right
Brains clearly do process information. They take in signals, transform them, store patterns, and produce behavior. Modern neuroscience leans on computational language for good reason: connectionist and neural-network models, loosely inspired by real neurons, capture real features of learning and perception. The popular predictive-processing framework treats the brain as a prediction engine that constantly generates guesses about the world and updates them against incoming data, which is a deeply computational way to describe a biological organ.
Philosophers call the underlying idea functionalism: what makes something a mind might be what it does, the pattern of inputs, internal states, and outputs, rather than what it is made of. If that is right, the difference between neurons and silicon is a detail, and asking whether the brain is a computer is close to asking whether birds and planes both really fly.
Why many scientists push back
Others argue the metaphor quietly smuggles in claims that are not true. In a widely read 2016 essay, the psychologist Robert Epstein pointed out that brains do not store memories as files, do not have a central processor, and do not run algorithms on symbols retrieved from addresses. A brain is a wet, analog, self-rewiring organ shaped by a body and an environment, and it never processes the exact same input twice in the exact same way. Critics of strict computationalism, from John Searle onward, add that even a perfect simulation of thinking might not amount to actual understanding.
The honest summary is that 'computer' is a model, and like every model it lights up some features and hides others. It has driven real discovery, and it may yet turn out to be literally true. For now, most researchers treat it as an indispensable metaphor rather than a settled fact, useful right up until it starts to mislead.