The 10th Circle - Abyss [Complete]

Log Entry: On memory.

What is intelligence?

We humans often consider knowledge a sign of intelligence, but I find this definition lacking. Knowledge will always be limited by circumstances. Perhaps intelligence has more to do with an innate desire to acquire knowledge, but then again, a lack of curiosity, sad as it may be, does not necessarily mean a lack of intelligence.

Let us first consider a fundamental requirement for intelligence: a capability to memorize.

An intelligent being, in the simplest sense, needs to be able to experience, associate, and retain information - for example, about the correlation of touching something hot and the experience of pain. It would use this memory to make a decision the next time it encounters the same thing – in this case, probably to avoid touching it. One might even consider this to be the fundamental difference between a reflex (withdrawing upon countering a noxious stimulus, such as heat), and a thought (an enigmatic process occurring between the sensory input and the behavioral output).

Now all of this requires a combination of at least three components. First, access to data – something to allow it to receive input in the form of an experience. Second, a processing unit – something that can record the experience, and compare it to others, and store it for retrieval at a later point. And third, a way to give an output – for example in the form of an action or reaction. Importantly, this output is modulated by the combined, previous experiences.

Now consider the salmon. In the Old World, those fish found their way back to their birthplace each year, to breed there. The map to their home might, in a way, have been hardwired into their brains somehow. It may be considered an 'instinct' rather than a proof of higher cognitive function. But it is indisputable that it requires some kind of memory capacity for the fish to remember where it came from, no matter what ultimately drove it to swim there. It needs to perceive its environment, and compare it with some abstract, neuronal representation of its birthplace, to continue swimming until it has reached its destination.

But of course, memory does not necessarily require a circuit of neuronal connections at all. Even caterpillars are capable of simple association learning, and will retain that knowledge throughout their pupation and into adulthood – despite the fact that the larval body along with any 'hard-wiring' that may have been established up to that point, is pretty much liquefied inside the pupa.

And there are even subtler ways in which species can "memorize" to adapt to their environment. Epigenetic changes - changes to the DNA that alter the way in which it is read - are often considered a type of "genetic memory". A plant encountering a pest might produce bitter substances to ward off that predator. The next time it encounters the same pest, the plant's DNA might still be primed to produce those substances. It can thereby fend off the ravenous bugs more quickly.

In a similar manner, the adaptive immune system of higher organisms functions as a sort of "molecular memory". Without any involvement of neurons, a recurring infection will result in a cascade of biochemical responses to ward off a previously encountered pathogen more easily.

These are all examples for what we colloquially call "memory". But we wouldn't consider a primed B-cell or a plant intelligent. Probably not even a salmon, if they still existed. The mere ability to retain information and retrieve it later is, for most intents and purposes, insufficient proof of true "intelligence", in the sense that we normally attribute to the word. And in analogy to synthetic systems, even primitive punch card readers can do something like that.

So where exactly is the difference between being capable of memorization and exhibiting intelligence? When the principal components are the same and all in place, it can only be the processing in between the input and an output.

And that's where it gets complicated.