2019: „Vision: Artificial Intelligence“

Looking back over the last few centuries, it seems that every major surge in technical innovation has been accompanied by collective and individual fears, but also by great hopes. Mankind’s great leaps forward in development cannot be evaluated without looking at the many individual fates they entail. Hardly any innovation has spared large circles from disadvantages, because technical progress divides at least all those who participate in it from those who are excluded. In the last century, mechanization and automation in the world of work were technological revolutions that upended entire societies and exacerbated the economic divide between entire nations, even continents. At the beginning of this century, digitization brought similar effects, extending from work life to the private sphere. Now, humanity is on the cusp of a new and incredible revolution, as a new instance of thinking appears on the horizon. Artificial intelligence (abbreviated A.I.) or in German “Künstliche Intelligenz” (K.I.) has now hit the headlines.

What happens when computers become smarter than humans? Every day we are surprised by new success stories about the capabilities of computers and software. Entire fields of science are experiencing successes in their research on the basis of “A.I.”, because especially where huge amounts of data are available, appropriate programs can now perform evaluations and make forecasts with the support of fast computers. So-called “weak” A.I. has long since become part of everyday life in science and industry. Another dimension of the development is close at hand:

What has been the subject of dedicated science fiction literature and commercial film productions for several decades is now beginning to become reality. Computers have begun to learn on their own, surpassing the performance of human intelligence, long ago in chess or the game of Go, and now in many applications of image and speech recognition or autonomous driving. The dramatic effects in so-called “deep learning” prove that computers are performing in ways that only a few optimists predicted years ago. However, these capabilities, e.g. for predicting moves of complex games, are still related to statistical evaluations of all computable possibilities and their effects on the game. We are looking in this area at improving mathematically based evaluations of the best from an immense number of variables. However, computers are beginning to work in networks and share information, creating effects that are no longer transparent to most people. “Big Data” captures the entire world and everything that lives there to make the interactions between all these variables and constants more predictable. This applies not only to weather forecasts, but also and especially to the predictability of people in their actions and reactions. What began many years ago as comparatively innocuous market research, combined with the design of computer-driven neural networks, has evolved into far-reaching tools that place disturbing possibilities in the hands of those who control them. If the power imbalance between those who don’t even use computers and those who control computer networks is already reaching worrisome proportions, the next stage of development is for these systems, which “learn” faster than humans, to achieve performance gains in relentless cascades of self-improvement that make futurologists dream – or shudder. The very idea that a kind of “superintelligence” could emerge from these developments in the all-too-near future, beyond the control of humanity, drives many professionals and self-professionals to ask about the limits of meaningful forms of development. Calls for “computer ethics” are being voiced, and the question of whether computers should be encouraged to develop “feelings” is driving some people into a frenzy.

Regardless of how one may view these new possibilities, some aspects of the students’ work remain striking. What many contributions have in common is that the sometimes bizarre concerns seem to be fed more from past film productions than from the state of current development, which, fortunately, is still far from implied options.

Likewise, it is striking that in the confrontation with “artificial intelligence” the question of the quality of natural intelligence becomes interesting. In this confrontation, an understanding of the special achievements and possibilities of human intelligence increases, including the question of the imageability of the same. Thus, a skeptical but also amused mood runs through the students’ pictures. Neither the participants of the project “Vision – Artificial Intelligence” let themselves be blinded by the prognoses about the “last invention of mankind” nor by the euphorias about the “end of stupidity”. Rather, they all defined their own reflections in their pictures and also revealed the information from which their thoughts were fed. It also became clear to all participants that the variety of ideas circulating on the Internet made an objectively credible and non-panicked evaluation of “artificial intelligence” rather difficult. Technically complex information from experts was juxtaposed with seemingly irrational opinions, and no conclusive interpretative authority was tangible. Even the statements of high-ranking scientists from the ranks of the most modern technology corporations still paint an ambivalent picture of A.I. at the moment.
Apart from all fears and hopes connected with the A.I., the whole group was united in the thought that the big and abstract questions about being human cannot be solved by software developers or algorithms, but by each individual for himself and in the cooperation of the peaceful community of all people, if at all.

We wish our friends and guests a lot of inspiration when dealing with visualizations of our students’ ideas!