Thinking “is nothing more than symbol processing as is done in computers” – for transhumanists like Ray Kurzweil it is a foregone conclusion that the thinking can be fully fathomed through a precise study of brain processes. If it were possible technically to emulate the brain we would have a thinking device – an artificial intelligence.

The latter would in turn be capable of continuous technical improvement until it far exceeded the human being. This belief that it is possible to build a general artificial super intelligence is fed by singular successes in computer engineering and, above all, through novels and films.

One of the  most spectacular phenomena in this respect is Dan Brown’s latest book Origin which occupied top place on the bestseller list of the German news magazine Der Spiegel for weeks. An ultra-intelligent machine is involved in the action from the beginning. It reveals itself initially only as a voice in the headset of the novel’s main protagonist. It gives him smart tips and organises sophisticated escapes when he is being pursued. The novel very cleverly combines the promulgation of transhumanist views with gripping action in which it turns out at the end that a super intelligent machine was the driving force behind all significant events.

Old ideas

The idea that we could build an artificial brain is not a new one. As long ago as 1943, when the first computers were being developed, the American neurologist and cyberneticist Warren McCulloch and his colleague Walter Pitts published an essay in which they described an artificial neural network which in a very simply way imitates biological processes in the human brain.

Every nerve cell in the brain is connected with many hundreds of other neurons through synapses. When these neurons are activated, this has an effect on the nerve cell at a certain threshold, it “fires” and then in turn acts on other neurons.

Based on this idea, McCulloch and Pitts constructed an artificial “nerve cell”. Each of these nerve cells is technically interconnected with the others. When electrical pulses come from the other cells, they are registered by the cell and given a value which in turn is passed on as an electrical pulse. Computer programs imitate the logical structures of human thinking independently of any consideration of brain processes. In contrast to programmed computers, artificial neural networks no longer map the laws of human logic but go deeper: they imitate the processes which take place in the human brain when a person thinks.

Neural networks imitate the brain processes which take place during thinking. This happens in that the internal weighting of the connections between the artificial nerve cells independently changes during training. Hence artificial neural networks are no longer programmed by humans but trained. In the course of many millions of training processes, the network adapts increasingly better to the set task. Once training has been successfully concluded,  the weighting in principle no longer has to change.

Although an incredible amount of human intelligence has flowed into a neural network, it is frozen in it. A neural network thus does not possess any independent intelligence. It is neither stupid nor clever nor does it take any decisions; these are concepts that simply do not apply to such a device. A mouse trap is not intelligent either just because it closes precisely at the moment when the mouse eats the cheese. So-called artificial intelligence has been made intelligent through human thinking but it does not think itself. Hence the question once again: what is thinking? Let us try to observe how an idea arises through thinking.

Conceptional thinking

Thinking is invisible for the physical senses. The only person who can observe it is the person who produces it. I might imagine a cube, for example. In my mind I see its twelve edges, six surfaces and eight corners. It requires a certain effort to maintain this mental image. I see the cube as an image before me.

If I now simultaneously attempt to observe the activity of my thinking, I experience a force that is directed away from me towards my mind’s image which has taken form in the idea of “cube”. The mental image itself is finished, oriented towards the past. I have a clear idea of the cube but it is merely an image: I cannot place any physical object on the cube in my imagination. The mental image is “virtual”. The situation is quite different when we try to observe the will.

Volitional action

A wooden cube located on a shelf. I want to pick it up and put it somewhere else. My decision is very consciously taken but as soon as I lift my arm to put it into practice, I note that although I can see and feel my arm moving, I have no awareness of the actual act of will which comes to expression in the movement of my muscles. I am one with the movement, I am the movement.

In the mental image I move towards things, in the act of will I go in the opposite direction: the movement is from the cube towards me. It is the cube which guides my action; it is its presence towards which my movement is directed. The phenomenologist Maurice Merleau-Ponty says: “The grasping movement is, from the beginning, in a magical way at its destination, it starts only in anticipation of its end.” Just as the mental image always points towards the past, so the active volitional act points towards the future.

The end of thinking and its beginning

Human beings grasp the world as it has become with their thinking, they understand the laws of physics and chemistry and so on. A formula is coagulated knowledge. And it is from this that they in turn make their machines up to and including neural networks. Once people have understood the laws of logic, they are also capable of building them into a machine: programmable computers are the result.

With neural networks, in contrast, the human being implements what they have understood about the human brain in technical processes. Technical neural networks imitate brain processes.

But neither computers nor neural networks can think. They are the end point of human thinking, previous thinking crystallised in silicon.

Where, then, does the starting point of human thinking lie? Inner observation shows us that an idea or mental image arises in an area which is inaccessible to ordinary consciousness. When we try to observe how an idea arises, we come to a place at which our eye loses itself in darkness. It clearly requires an extension of consciousness to be able to investigate this “place”.

Steiner answers the question about the origin of ideas in a surprising way: in the idea, the same spiritual process reaches its conclusion which also underlies the formation of the human body. From the point of conception onwards, the spirit forms the body and brain; as it shapes their form, it dies in them. The last residue of this spiritual formative force comes to expression in the formation of an idea. Living thinking turns into abstract thinking in the course of childhood and adolescence and as such becomes increasingly dependent on the physical foundation. At about the age of twelve, when the child enters class 6, their thinking is developed to such an extent that it can gradually grasp logical structures. That is why in Waldorf schools science teaching (physics, mineralogy, astronomy), which refers to logic and evidence, starts from class 6. Then using and understanding digital devices also makes educational sense.

The extra-corporeal living thinking dies in the human body, turning at first into mental images and finally into abstract thinking. When people build machines that imitate this thinking, the corpse dies a second time: the until then still mobile human intelligence crystallises into devices which work purely as machines even if outwardly they might feign life.

The balance

Something new starts with the will, so where does its end lie? Steiner’s answer: in life after death. It is only there that what humans have endeavoured during life unfolds to full reality. Just as every mental image, every abstract thought is something that has reached completion, so every volitional activity is a new beginning, a seed which only finds its full reality outside space and time. Anyone who observes themselves during creative activity will notice that they are particularly active with their will. Every creative process is the seed of a new beginning. When people imaginatively seek opportunities to realise something, then their will is particularly active. It then becomes so strong that it brings the idea to life again, brings it a little closer to its origin again.

The technical artefacts of the digital world make life comfortable for us. We can take it easy because the devices do a lot of things for us. Personal assistants such as Siri or Alexa endeavour to suggest things even before we are aware we want something. Such assistants can lead us through life today.

Self-determined personalities who want to be involved in shaping and changing the world need powers of imagination through which they can develop images of how the world should look in the future. And they require perseverance to carry out their intention even in the face of adversity. For that they need a strong will and a rich feeling life.

Education must therefore challenge the will and feeling of children in particular measure and encourage them to develop intensive feelings and self-determined volition. Particularly at Waldorf schools, the awareness has to be maintained among parents and teachers as to the importance of developing the feelings and will in a time of artificial intelligence. Because only through a rich feeling life and a strong will can humans withstand the grip of devices that have turned independent. Through commitment and strength of will, a forward-looking, creative thinking must be set against the thinking “frozen” in machines. These qualities are best practised in the crafts and art lessons.

About the author: Edwin Hübner is a professor at the Freie Hochschule Stuttgart – Seminar for Waldorf Pedagogy. He is the author of several books on the subject of media education.