Comparing the Human Brain to Programmable Logic Devices

[...] comparing the human brain and the way it thinks, sees and interacts with the world to physical/mechanical things like computers and Programable Logic Devices results in huge abstraction gaps, to the point of making the comparison null.

except we aren't talking about the internals of the brain - we are talking about definitions of the words, which are very different.

The definitions of the words are contingent on human experience, even more so than "code" are "data" where we try to be more mechanistic, and still most people make the mistake of thinking that they are distinct categories (spoiler: they're not; whether something is "code" or "data" depends entirely on your perspective).

If we want to draw computing device analogies, then the brain is an FPGA that is continuously reconfiguring itself throughout its runtime.

IQ being the CPU, working memory the RAM, and long-term memory the HD/SSD

Can we make an analogy between the human brain and a computer? IQ being the CPU, working memory the RAM, and long-term memory the HD/SSD? To what extent do you think this analogy holds?

(By David Powell, Master of Jewish Studies in Jewish Science, Academy of Jewish Religion - California, posted at https://www.quora.com/Can-we-make-an-analogy-between-the-human-brain-and-a-computer-IQ-being-the-CPU-working-memory-the-RAM-and-long-term-memory-the-HD-SSD-To-what-extent-do-you-think-this-analogy-holds)

No... not a good analogy. IQ is a function balance that results from a cognitive/intuitive rather than emotive/perception style of processing information. IQ is more like a Benchmark result and thus is just an evaluation of the system, not the system itself.

The Triune Model of the Brain is largely incorrect, Reptilian (brain stem), Mammilian (Limbic System) and Neocortex. Rather, the closer analogy is:

The most important takeaway is that the Limbic system is actually the place where almost everything “happens”, not the Neocortex which is a pattern recognizer that delivers its results to the Thalamus, Amygdala, Hypothalamus, and other structures called collectively the Limbic System for processing.

But, Neocortex and Cerebellum have many similar functions and capabilities, even though these are used differently. The Neocortex recognizes patterns through connections between neurons using direct, close and distant associations based on four separate query types. Beyond this, two of the query types utilize “procedural memory” that links to the Cerebellum facilitate “math coprocessing”, which is actually a form of “closing approximations” that make it possible to estimate what is “not there” vs. identifying what “is there”. The Cerebellum is largely an indexed storehouse of procedures (habitual results) that when triggered run without conscious interaction. The closest analogy would be that of “stored procedures” in an SQL Database. [A huge lot] of actual behavior is built from nested collections of these stored procedures being called.

Now reality is much more complicated than these analogies imply. But, this is the quick and dirty of it.

John P Barbuto, M.D. comments:

Yes, of course. This is a standard metaphor. The analogy holds to a degree and can be extended into clock cycles, threads, linear and parallel processing, software biases, and so forth. None of this should surprise us because we create computers to do what humans do - only better in one aspect or another. So, to do these tasks the computer needs processors that have some correspondence in human brains.

The analogy has definite limits. Most notably, human brains utilize emotions to sort/process incoming information to personal relevance and also to focus the character of output. Put bluntly, human brains process for “what I want”. Human brains evolved to process information in such ways as would keep the individual alive. (Suicide is a side-discussion, though interesting.) Alternatively, current computers process data for objectively-and-specifically accurate outputs. Computers were created and have evolved to process information without unintended bias. (Weights, bias, and activation functions in machine learning do produce output bias but these are intended for the specific computing goals.)

So, fundamentally computers have similarities to human brains because we made them to do things that we can do, only better. However, computers and human brains have different agendas (in current forms). Eventually, computers will have functional levels that correspond to emotions and even consciousness (these are just processing overviews of certain types).

Paul King, Computational Neuroscientist, Software Entrepreneur, contributes:

The brain is neither analog nor digital, but works using a signal processing paradigm that has some properties in common with both.

Unlike a digital computer, the brain does not use binary logic or binary addressable memory, and it does not perform binary arithmetic. Information in the brain is represented in terms of statistical approximations and estimations rather than exact values. The brain is also non-deterministic and cannot replay instruction sequences with error-free precision. So in all these ways, the brain is definitely not digital.

At the same time, the signals sent around the brain are either-or states that are similar to binary. A neuron fires or it does not. These all-or-nothing pulses are the basic language of the brain. So in this sense, the brain is computing using something like binary signals. Instead of 1s and 0s, or on and off, the brain uses spike or no spike (referring to the firing of a neuron).

[At the level of] the neuron, everything works via biochemical pathways, which are somewhat similar to analog. Neurons also perform internal electrical signal integration in an analog fashion. Analogously, the digital logic gates used by computers are implemented internally using transistors and resistors, which are also analog.

This recording of neural spikes over time shows that the spatiotemporal pulses of the neural code looks a lot like digital signaling.

The bottom line is that the brain processes information using a representation strategy that is neither analog nor digital. It is a different type of computation, involving circuits and networks composed of spiking neurons. One of the central tasks of neuroscience is to figure out how this information processing paradigm works.