The design of the electronic mind

2021-10-19 | aprates.dev

[1] Leia este post em português

On a cold winter morning in Germany, still sleepy and after a good dose of coffee, during a class on artificial intelligence, I started to have some interesting thoughts. I decided to open the notebook, and I jotted down some crazy ideas that came to my mind… That's how the electronic mind project began. Now, a decade later, as I reread those notes, these are my tidbits on the subject.

          __
  _(\    |@@|
 (__/\__ \--/ __
    \___|----|  |   
        \ }{ /\ )_/)_
        /\__/\ \__\__)
       (--/\--)   
       _)(  )(_
      `---''---`
According to Alan Turing, creator of computing, a machine could only be considered provided with human intelligence if it were capable of:

Being kind, industrious, beautiful, friendly, having initiative, sense of humor, distinguishing right from wrong, making mistakes, falling in love, liking strawberries with cream, making someone fall in love with her, learning from experience, using words correctly, being the subject of her own thinking, having as much diversity of behavior as man, and doing something really new.

Don't get me wrong, I believe that today's famous virtual assistants such as Siri (Apple), Cortana (Microsoft), Google Assistant, Alexa (Amazon) etc. are impressive, but fall far short of achieving this degree of autonomy and sophistication.

Even if these features seems difficult to implement, the questions that remain are:
However, there is an almost scientific truth that all technologies that man has today were imagined as a delusion in the past and for future utopias (or dystopias) it is only a matter of time for these daydreams to become reality too.

Definitions

First, let's get to the misconceptions. The first confusion that can occur on this topic is to believe that brain and mind are the same thing, when in fact the brain makes up the physical and mechanical part that supports the mind. The mind is something intangible made up of ideas and emotions.

We can then compare our human composition with computers:
In this context, here are my definitions:

Modular structure

The mind appears to be a system that processes ideas associatively through independent intelligent modules that collaboratively generate awareness. The mind learns and reacts through perceptions, having its own universe made up of a collection of flows and contexts of ideas.

The environment and the way in which new impressions are captured may be related to the following phenomena of human experience.

Subconscious

Sigmund Freud, who lived between 1856-1938, built his famous model of the mind composed of Ego, Id and Superego. The Ego is the system responsible for resolving conflicts of the irrational needs of the Id and the moral impositions of the Superego. He considered that the mind works on a conscious and unconscious level.

Mental model

In 1943, Kenneth Craik states: "If the organism carries a scaled-down model of the external reality and of its own possible actions inside its head, it is able to experiment with various alternatives, decide which is the best one, react to future situations before they arise, use the knowledge of past events to deal with the present and the future and, in every way, react in a much more complete, safe and competent way to the emergencies it faces."

Intuition

More than 3,000 b.C., in the Indus Valley, the Harappian civilization developed a theoretical philosophy called Sámkhya. According to this philosophy, human beings can reach different levels of consciousness. The evolutionary stages of consciousness are: emotional, mental, intuitional and Púrusha, which means the absolute essence or spark of life. In general, man lives on the emotional and mental level. However, he is eventually able to glimpse the last stage, a deep state of self-knowledge, through his intuition.

Humanizing integrated circuits

Why do we think? Well, I don't know why we think. Perhaps, because we were born that way and we cannot avoid it… The whole story of the evolution of life ultimately cannot satisfy all the 'whys', but it may try to justify the 'hows'.

I believe that to recreate a mind it is necessary to study it. Realize the rules that govern the orchestra of thoughts. We need to delve inside ourselves seeking to get almost to the root of our synapses, not necessarily on how our brain works (which seems to have been the focus of modern research fronts), but trying to understand why and how ideas come about.

An interesting feature of human intelligence, for example, is understanding how we forget things. Forgetting is not exactly a flaw, and perhaps it contains the secret sauce for a more sophisticated artificial intelligence.

Yes, our memories are unstable: people forget details especially when a detail is not used. Our mind subtly or completely distorts memories when we remember something. We seem to have an emotional safety device: we forget or adapt all information that is not useful or pleasant to us.

Sometimes we need to hear information over and over to keep it. If it's not something we are looking for or when we cannot make an association with some other concept that is already familiar to us, the tendency is for such a piece of information to go unnoticed.

The mind loop

Curiosity is the engine of the mind that keeps it in an insatiable gluttony for new ideas and relationships it can establish between them. It keeps asking itself 'what do you mean?', 'why?', 'what is this?', 'but what if…?', and is never satisfied.

And so, based on guessism really, I believe that our mind operates through a kind of infinite game-loop that is always trying to satiate this curiosity that has no end. This procedure can be described in pseudo-code, more or less as follows:

while (is alive)
     add new perceptions to mind
     discard irrelevant ideas
     analyze ideas (based on memory references)
     if (possible new relationship)
         if (possible to do something about it now)
             take concrete action
         else
             keep interesting analysis in memory

Final Thoughts

Thinking rationally is just an attribution of a logical method, but it is not the natural process of human thought. Although logic is the basis for integrated circuits it is not the basis for our mind. Human beings are not logical, so I imagine that emulating a mind will have less to do with rules and more to do with representations of contexts of ideas.

Having worked in game development in the past, I've learned that what matters is looking real. It doesn't matter how you produce the visualization of scenes or the intelligence of a character in a virtual universe, the important thing is that it causes the illusion in the audience and can be reproduced in real time. It is with an open mind to creative solutions that are less concerned with reproducing how our brain works and more focused on reproducing our behavior that I believe the greatest technological advances will be observed.

And you, what do you think about this?

See also

[2] Capsule Archives
[3] Capsule Home

Want more?

Comment on one of my posts, talk to me, say: hello@aprates.dev
or /msg aprates on irc.libera.chat

[4] Subscribe to the Capsule's Feed
[5] Checkout my projects on GitHub
[6] Checkout my projects on SourceHut

Join Geminispace

Gemini is a new Internet protocol introduced in 2019, as an alternative to http(s) or gopher, for lightweight text contents and better privacy.

Not sure how, but want to be part of the club? See:
[7] Gemini quick start guide

Already have a Gemini client?
[8] Navigate this capsule via Gemini


© aprates.dev, 2021 - content on this site is licensed under
[9] Creative Commons BY-NC-SA 4.0 License
[10] Proudly built with GemPress
[11] Privacy Policy