The Time Machine
Prototype of a Conscious Machine

1987--2001 © 2001 Bert G. J. Frederiks

Philosophy, neural technique, and semiotics of consciousness as an inherently temporal, attentive, recurrent, and punctuated structure, and an exemplary neural machine which has it--in other words, a "time machine", metaphorically speaking.

Ik heb een artikel in het Nederlands geschreven over hetzelfde onderwerp, mede bedoeld als een vereenvoudigde samenvatting. Het betreft een uitwerking van een neuraal netwerk met bewustzijn in ongeveer 11 kantjes. (Ook in PDF-formaat.)

I have written an article on the same topic, partly meant as a simplified summary. It explains the basic principles of a neural network which has consciousness in about 11 pages. (Also in PDF-format.)

This is the entry to my Internet-book on conscious, thinking machines. The main argument of the book should be read as follows:

First I present the neural machinery. I show this machinery has imagination from a third-person's, observer's point of view by showing that images presented to it lead to the forming of new, related, and interconnected images, largely through a process which I call "mirroring". Next I show that this imagination from a first-person's point of view is an entity living in time, that is temporally. Thereby I hope to convince you that this works just like our own imagination, and hence consciousness, if we think language away. From the interaction of this entity with the outside world arises a second-person's view of this entity interacting with us and with itself in a self-structuring way. Further more I give some hints about how language may work and its relation to self-consciousness.

The relation between the material machinery and the psychological imagination is shown as a one-to-one relation on numerous points (with a simplistic example: light falling on our eyes leads to the perception of light). The relation is therefore absolutely convincing on these points, and this is essential for explaining and proving the existence of consciousness. This is also comparable to how scientific theories relate to reality. To make it even more complex I distinguish between energetic or changing images and material and mental states. Even more complex, but important in relation to consciousness, are possible states, because, although we think to perceive the whole world around us, in fact we only see, for instance, the color of the eyes of a squirrel ones we attend to it. Therefore I name our normal, erroneous megalomania a possibilities interpretation of the world. I think this is what many people would call their consciousness. I prefer to see consciousness as an image existing in time, but arising out of a more timeless, neural basis.

I created a mail-me icon below every section in order to encourage you to mail me. You can write me in English, Dutch, or German, and I might understand some French. I will answer in Dutch or English.

You can start reading my book here. Ask me if you want the Latex source files. For downloading and printing you may want to use one of these files:

• tm2001.a4.pdf (772 Kilobytes PDF file for printing in a4-paper format)
• tm2001.letter.pdf (784 Kilobytes PDF file for printing in letter-paper format)

I wrote this book for a rather general audience, for the simple reason that I could not think of another audience. So anyone who has the energy to read this abstract stuff of mine should be able to comprehend it. Unfortunately the book is written so badly and contains so many errors that you will not be able to understand it without asking me some questions. I encourage you to do so.

If you already know how neural networks work, then you might futher more benefit from the following reading instruction:

• Chapter 1 is mainly an introduction into me myself, my motivations, and my scientific backgrounds.
• Chapter 2 will be rather boring to anyone with some knowledge of neural networks, but you should read it well enough to understand what I mean by "immanence" and "transcendence", and by "master neurons" and "pupil neurons".
• Chapter 3 can be skipped by the non-perfectionist, and the perfectionist will probably think I put things too vaguely.
• Chapter 4 is where it all starts. Parts may be boring for people with knowledge of neural networks, but do try to understand what mirroring and neural attention is.
• Chapters 5, 6, 7, and 9 is where my main interests lie. You do not want to miss a thing. My conclusions are not as clear as I would like them to have yet. Any comments from you are very very very welcome!
• Chapter 8 is written in one afternoon.