Digital technology is much more than just a means to optimise work. We invented it to increase our knowledge and create something new. Artificial intelligence (AI) raises fundamental questions: Is there a soul? How does the mind gain consciousness? Intelligent machines promise to repeat original creation: We want to understand where we come from and we are not willing to leave until we find out. Artificial intelligence and innovative companies are supposed to help.
Artificial intelligence has come to stay. It will bring us closer to our dream of a sustainable life in harmony with nature’s resources. Highly efficient systems for work management will give us more comfort in everyday life. Artificial intelligence will increase global prosperity, tackle health problems and push ahead the self-optimisation of humanity. Artificial intelligence will bring us closer together as a species, which will lead to more regulation in our societies. It will help us to explore space and create new military destruction tools of unprecedented precision.
Artificial intelligence as key technology of the 21st century
“Artificial intelligence could be one of mankind’s most useful inventions,” advertises Google’s research company “Deep Mind“ on its homepage. All over the world, there are pioneering projects of practically applied AI with great social benefit: more precise medical diagnoses through radiomics, digital assistants to control complex large-scale projects, intelligent traffic control through AI-optimised traffic lights, self-learning systems for customised energy management of buildings, sensor platforms for environmental monitoring or food safety and traceability through transparent real-time monitoring of logistics chains.
Artificial intelligence is the key technology that will account for future international competitiveness. National AI strategies are being adopted worldwide to support research and development, know-how, promotion of young researchers and international networking. Investments are running into billions – above all in the USA and China, which are competing for global AI supremacy. However, despite the technology’s enormous potential, not all companies succeed in efficiently integrating artificial intelligence into their strategies and processes. 40% of companies that make significant investments in AI do not achieve measurable success with it.
Why do AI initiatives fail? Apparently, the complexity and risk of AI are underestimated. There is a lack of courage for real change management, too little is invested in training employees adequately. In addition, there are false expectations towards the technology: Homo Deus links the search for collective benefit to the religious belief in intelligent design. Where artificial intelligence should be a tool for individual empowerment and real-time collaboration, it becomes an anonymous monitoring device that dictates work instead. Individualism, autonomy and business traction lose ground. Where does this false belief come from and how can it be fought?
Man is the best machine
Artificial intelligence is a highly complex cluster. According to Alan Turing, artificial intelligence refers to the ability of computers and symbol processing programs to mimic the performance of the human mind. This idea originates from the modern idea of man as a machine, which emerged at the beginning of modern science in the 17th century. Artificial intelligence research concentrates on the automation of intelligent behaviour and machine learning. It is also a scientific theory of the mind that investigates the problem of definition, schematic representability and scalability of intelligence. The discipline is multidimensional and interdisciplinary: new insights include results from different fields such as robotics, neurobiology, psychology and philosophy. Finally, artificial intelligence refers to all technologies and applications that can simulate intelligent behaviour. A distinction is made between weak and strong AI.
Weak AI, also called Artificial Narrow Intelligence (ANI), takes over thought processes for easily schematisable problems. Algorithmic operators process data, which execute pre-programmed tasks automatically. The enormous advantage of weak AI is its procession speed: transistors in modern computers work millions of times faster than the human brain. Thanks to deep learning and artificial neural networks, weak AI can independently derive new rule sets. The performance of weak AI depends on the amount of data – the big data – that is stored for learning. Weak AI does not clearly distinguish between hardware and software – usually it refers to systems or computer programs that are supposed to enable machines to perform intelligence tasks. However, there are some obvious disadvantages of weak AI: Due to its material, it is extremely prone to error. It cannot (yet) make its own experiences, which are elementary for human learning. It does not exist in its own cultural context, as AI-critic Hubert Dreyfus already noted in the 1970s. Therefore it has no intuition that allows it to evaluate facts in consideration of a greater overview.
Strong AI, also called Artifical General Intelligence (AGI), is the notion of an artificial intelligence that can understand or learn any intellectual task as well as a human being. In contrast to weak AI, it is open: it leaves the initial determinacy behind and acts autonomously. When and whether we will ever reach this status is disputed among AI researchers, but some believe – above all futurist Ray Kurzweil – that it will happen by the middle of the 21st century. As soon as this status is reached, the road to a superpower AI, also known as Artificial Super Intelligence (ASI), is only a short one: this artificial intelligence is supposed to surpass humans in cognitive performance. According to Oxford philosopher Nick Bostrom, this super intelligence will confront humans either as oracles, as geniuses or as sovereigns. Some people even fear that the huge increase in human cognitive abilities with the help of machines will lead to an explosion of intelligence that will fundamentally threaten or even destroy human existence – „Skynet” is near. Realistic future scenario or techno fairytales?
Predictions about the future are a tricky thing, because people have always overestimated the technologies of their time. In the middle of the 19th century, utopians dreamed that automated factories, using a minimum of human labour, would be built within a very short time. Even then, these progressive visions were accompanied by fear of dramatic unemployment and social breakdown. Some experts do not believe in the dramatic image of the intelligence explosion. The American physicist and Nobel Prize winner Frank Wilczek assumes a long-lasting “co-evolution” rather than a so-called technological singularity. According to Wilczek, there will be a whole ecology of different types of powerful intelligence interacting with each other over many decades. This would give people enough time to develop moral rules for digital ethics by learning in practical interaction.
The art of intelligence is the interaction of decentralised units
The term artificial intelligence is problematic. It suggests that machine intelligence can replace human intelligence. Another problem is the term intelligence. In addition to mathematical-logical intelligence, there are other intelligences that are maybe even more decisive for human behaviour: for example the emotional intelligence to use feelings as a means to evaluate the importance of facts and events, and the social intelligence to weigh up decisions in a large radius of interactive relationships between individuals and collectives.
Current AI research and robotics development already explore to which extent emotions and affects enable faster decisions than rational judgement. Nevertheless, the current debate on artificial intelligence is based on an outdated concept of identity that regards the brain as the central command-and-control unit operating top-down. But the brain is not a mathematical supercomputer that controls the body. In fact, it is far less autonomous. Brain performance depends on physical factors such as intestinal bacteria; hormones influence our moods and our cognitive skills. We are easily manipulated by external circumstances and our memories are often flawed.
So why are we still so effective? Maybe because the brain can act in a decentralised way. Connectivism claims that instead of a controlling central ego, many autonomous units make decisions and thus generate complex behaviour. In The Society of Mind, Marvin Minsky describes a society of sub-intelligences that cooperate by negotiating limited and conflicting viewpoints among themselves. The understanding of consciousness as a diverse group of inner agents follows the pragmatic view that humans do not follow logical, programmable rules, but grow and learn by trial and error. If intelligence in postmodernity is no longer a substance, but the constitutive difference of multiple connections, then it makes little sense to explain intelligence from a biological-positivist model.
Excursus: The highest consciousness – all or nothing?
Will increasing mathematical intelligence at some point produce consciousness? Very unlikely. For this to happen, the process of thinking and consciousness would have to be the same. But we know very little about the connection between thinking and consciousness. So what is consciousness? A self-reflection? A “mental pollution created by firing complex neutral networks”, as Yuval Noah Harari claims in Homo Deus? An illusion? Or is it empathy?
How are brain and consciousness connected? Split-brain patients – epilepsy patients whose brains were surgically separated – prove that a coherent personality exists without a physical connection of the brain. Patients with cerebral diseases such as dementia have an intact ego-consciousness, although they have lost most of their memory. Experiments with hypnosis prove that people can be manipulated to suppress brain impulses for a short time and even forget things on purpose. Which power controls consciousness? Or, to rephrase Aristotle, how is the soul shaped to let the body become a perfect possibility?
“I think, therefore I am.” we learned from Descartes. Is that true? Is thinking something elementary or just a cognitive experience? Greek antiquity considered the movement of the soul as the harmonious root of life. Modern philosophy separated mind and sensual perception. The humanistic dualism and the mind-body problem emerged: man, the only being capable of rational thinking and therefore the crown of creation was born. The belief in reason as a godlike power makes SciFi enthusiasts wonder, if the uploading of the brain to a computer is the same as reaching immortality.
There are, however, very contrary views of what makes us human or what life is all about. Asian religions such as Buddhism differ from Western philosophy. The highest consciousness is a dissolution of thinking. The true unity of the spiritual self can only be achieved through meditative contemplation and spiritual emptiness. The dualistic consciousness, on the other hand, which creates mental patterns through the perceptions of the senses, is limited and incomplete: it is separate from the supreme wisdom (Prajñā), which permeates the world and can only be experienced through direct intuition. How can forgetting help us to find our inner self, which longs for progress and innovation?
Technology and art – two sides of the same innovation coin
Our concept of artificial intelligence is incomplete because it does not include inspiration. To inspire you need chaos, curiosity, an adventurous spirit and spontaneity – games instead of boring daily routine. Plato already knew that you particularly need imagination, – phantasia – if you want to think in new ways. Above all, abilities such as autonomy and qualia – subjective feeling – are central. Technology does not reveal itself, art does. If artificial intelligence is to tell us something about the mystery of being human, it must guard our dreams, fantasies and memories. If artificial intelligence is to become a vessel of trust, then it must be our aura.
What is art? Like man himself, it is contradictory and unpredictable. It does not need to explain itself to achieve its greatest effects. It is as difficult to describe as beauty, but it has much in common with it. It does not care for utility, it aims solely at the aesthetic effect of the overall picture, which uplifts, educates or disturbs. Art loves ambiguity and mystery, it withdraws determination. Art reveals humans in their nakedness: beautiful and horrible, cruel and kind. At the same time art is ambitious, self-confident but selfless. The artist wants to create something unique and yet reach everyone. How does art achieve this? Art is a radical expression of the singular and at the same time aims at generalisation. Art is exchange and dialogue, but it also changes, inspires and challenges. To achieve this, an artist is constantly searching for the unknown, which he makes visible by alienating the familiar. Art plays with our self-perception and makes new offers of identity.
Innovation is often surrounded by the myth of the absolutely new. The absolutely new, however, is only a theoretical thought that epitomises this tempting dream. Practical innovation, on the other hand, is a recombination of known solutions in a field unfamiliar to the discipline. Therefore, real innovation must take diversity and interdisciplinary cooperation into account. Artistic techniques can help to show structural relationships by creating aesthetic images.
AI as artistic stimulus for physical and mental learning
Innovation projects are not a deus ex machina. They successfully realise evolutionary trends, which can be experienced but not formalised rationally. You may call it God, universal soul, Elan Vital or Zoe – there is something that cannot be grasped by algorithmic subdivision, since it is hidden in individual perception. Intelligence cannot be reduced to imitate a holistic experience, but must transsubjectively reveal the flaring up of existence in imagination, as the natural scientist and philosopher Gaston Bachelard describes the poetic image in the book Poetics of Space. According to Bachelard, original creative power and basic ethical attitudes are pre-linguistic:
“The soul gives consecration. It is the elemental force here. It is human dignity.” (Gaston Bachelard)
So how do you bring together cognitive superpower and creative inspiration, which roots in subjective imagination? Let’s look back: our linear, alphabetical culture understands the mastering of memory as practice and training, that is as mastery of technical skills. In antiquity, however, the concept of technology was not yet separated from art. The téchne included both manufacturing and producing as well as the works of the later so-called higher arts: music, poetry and rhetorics. Mastery was not restricted to following given formal rules. Instead of repetition and uniformity creative improvisation was desired, which enabled the creation of individual variations.
Our simultaneous environments of the internet need a new strategy. Exact programming of all factors will not do to understand highly complex problems. Heuristic generalisation by ignoring irrelevant aspects is more effective. Artificial intelligence, which is to master digital space successfully, needs an associative understanding instead of factual analysis. Practices such as mnemonics can help. In this ancient art, memories are thought of as place: Images and complex contexts are better remembered through association, visualisation, chunking and localisation. Mnemonics is similar to the principle of neuro-computing, because dynamic processes and feedback effects are created that can be depicted on many different spatial and temporal scales. In contrast to quantum technology, neuro-computing is based on the description of a highly simplified model.
Conclusion: Added value through AI can only be achieved by investing in creative self-confidence
What do we learn for change management and business innovation? Added value through smart technology can only be created by fostering interdisciplinary approaches, creative self-confidence and the willingness to part with old procedures. Only self-critical distancing from the cognitive process will produce the necessary antifragility, which can unite the apparent paradoxes of artistic innovation: revolution and origin.
Artificial intelligence can only be successful if it serves our playful desire for change, challenges us and helps us to be better than now. In helping people to learn with context-aware computing instead of making decisions for them, AI can show us ways to accept and practically shape our responsibility as a species. AI must not enlarge our abundance, but find ways to end shortages. Technology is not meant to create new smart cages but to increase our freedom.