All Intelligence Is Artificial

All Intelligence Is Artificial

Merriam-Webster dictionary defines artificial as:

1. humanly contrived

2. (a) having existence in legal, economic, or political theory (b) caused or produced by a human and especially social or political agency

3. (a) lacking in natural or spontaneous quality (b) IMITATION, SHAM

4. based on differential morphological characters not necessarily indicative of natural relationships

When a long series of evolutionary twists have gifted our ancestors with the capacity to reproduce and recognise complex sounds, another long succession of errors and mishaps has followed, bringing about what we today call language. Parasitising on sensory mechanics that have until then evolved for the mere survival of our bodies, the language has slowly, but surely, become a precious tool used to transmit information, thus turning our brain into what Daniel Dennett defines as a "Virtual machine". This is the point in time, in which the language has become a way to install ideas in human brains, a kin to software programs we install on the multitude of electronic devices we use every day.

This is also a point in time, when intelligence started to emerge – not as a natural phenomena, but rather as a human contrived use of the neurological wiring we are all provided with. Physiologically we're not too different from the humans that first started using languages. Even if it happened 150000 years ago – the earliest some researchers indicate as the moment language skills have settled in within our kind – we're talking about beings that had roughly the same, or very similar capabilities to the ones we leverage in our development and everyday life in modern times.

The emergence of language has allowed for two things to happen: it was an easy way to transmit information, such as "this plant tastes well" or "there's more fish downstream" and it?collectivised the survival effort, allowing to compound knowledge and experience. This has also given an unfair survival advantage to those who were faster to leverage the information they acquired from others and derive new ideas:

"Input 1: Predators hunt under the cover of darkness.

Input 2: Predators are afraid of fire.

Output idea: It's better to have a source of artificial light during the night, to both scare predators and be able to notice them in case they show-up."

This is intelligence, isn't it? And both inputs and outputs in this case are humanly contrived, as humans are the source of both inputs and human is the source of the logical conclusion in the output. Everything we know is no more than such inputs and outputs compound together and filtered through multiple (millions, billions, trillions) of iterations. For example: 150000 years ago humans experienced thunderstorms exactly as we do today. But the way we think about lightning has evolved a great deal through the aforementioned processes of deriving new ideas and filtering them. From the fury of gods we went all the way to discovering electricity and meteorology, which at current time provide us with a neat and agreeable explanation of the phenomena. Alas, the intelligence has no boundaries, so in 150000 years we might have a completely different idea and explanation for the same thing, despite thunderstorms themselves wouldn't change at all. But our ideas, our understanding of the world around us will and shall evolve. So while thunderstorms are a natural phenomena, our understanding of them is entirely humanly contrived.

Language has taken us further on the way of compounding experience and knowledge to create fantasies (to help explain natural phenomena at times) and abstract concepts (to explain the inner-working of our mind). Literature and art are an expression of fantasies or self-inflicted hallucinations that play a pivotal role in training our minds to accept things we can not see or experience first-hand. Love, hatred, guilt, fear are abstract definitions to emotional states that all humans experience and learn to recognise through language, connecting abstract concepts to sensations that our animalesque nervous systems produces for the sake of our survival. These concepts make us even more human, as they give us empathy – an ability deemed indispensable for our survival as a social animal. So fear and love are indeed natural, but our understanding and interpretations of the basal urges that our intelligence converts into what we call emotions is entirely humanly contrived.

The compounding of all the experiences and thoughts that have survived to our days is what constitutes our collective potential intelligence. Potential, because no-one can know or learn everything, but we all share a certain set of concepts that allow us to understand each other and collaborate. This is what culture is and we have the capacity to draw whatever we choose from it in our lifetime, acquiring the bits and pieces of compound knowledge that make it up. Over the centuries we have invented many different ways to achieve this, but the fundamental act of learning has remained the same: whether we read books, listen to podcasts or go to lectures this is all about the same – about developing our intelligence. But while we all might be learning the same things, we draw different conclusions from whatever knowledge we acquire. Some conclusions are better than others, some contribute to our survival as a species, while others put it at a greater risk. Some conclusions – such as scientific discoveries and art – in turn become part of our compound knowledge. Needless to say this is all entirely humanly contrived.

A lot of this applies to what we call "artificial intelligence": digital computers acquiring information and deriving new knowledge. The mechanics and inner-workings vary a great deal, but the principle remains the same and understandably so: while creating this technology we only had one example to imitate - ourselves. And now we are getting to the crux of the problem: "Artificial Intelligence" is a pleonasm. Intelligence is human contrived by definition, whether we like it or not. It is artificial. A more appropriate term would probably be machine intelligence, but I doubt it would ever be adopted. What's even more interesting is that whatever knowledge machine intelligence produces, cycles back into the pool of knowledge that is accessible to us (and to machines too), so perhaps machine intelligence is just another tool we have devised to help us enrich our knowledge and help the survival of our species. So shouldn't machine intelligence then be regarded as part of intelligence in general?