Which public colleges are best for AI

Artificial intelligence as a threat to humanity or will you soon be an equal interlocutor during your studies? In this blog post, Reinhard Karger, company spokesman for the German Research Center for Artificial Intelligence (DFKI), clears up some errors and shares his view of the relationship between AI and the university.

Universities are places of knowledge, the transfer of knowledge and human learning. Artificial intelligence is the digitization of human knowledge skills. Sounds like a wedding in heaven. That is correct. But an analysis is required.

There is plenty of room in universities for science, but not for magic. Machines can have a magical effect on the user, but they have none. That brings us to Artificial Intelligence. AI is about powerful machine knowledge processing and tools. That is obviously so, but it gets through overarching headlines, marketing statements and fictional productions into the background. Illusory expectations and unrealistic performance hopes are promoted. Tool realism is hampered. There is no magic in the machine.

The social result of over-pointing AI successes in the media is questionable insofar as it makes the machine appear strong and the human being weak. This has consequences because it promotes existential uncertainty, social resignation and skepticism about the future and does not strengthen an open attitude that is characterized by an affinity for innovation and active personal judgment. The most important cultural technique of the 21st century is the creation and maintenance of one's own intellectual plasticity, the development and adaptation to highly dynamic technical developments that, within a short period of time, influence or fundamentally change the subject area or the boundaries of the subject that one has studied. What is dramatic is that we do not take hold of the outstretched hand of enlightenment. Thinking for yourself as the foundation of the scientific method is losing its charisma and effectiveness, because it is suggested that machines, if not today, then certainly tomorrow at the latest, will outperform people.

That is certainly true in many relevant, but not in the essential dimensions. Pipetting robots are essential for genetic analysis and for carrying out series examinations. They are faster and more precise than humans, and should be. The processes run 24/7 and are extremely monotonous. When machines take over these tasks, it's good for the result and wonderful for the people who can benefit from better drugs in less time.

But it is not a threat to humanity or humanity. AI is a danger to people who only see themselves as a function. Human intelligence is multidimensional and manifests itself in sensorimotor, cognitive, emotional and social skills. People use knowledge skills in recognizing and naming objects, in understanding language, in producing sentences, texts, stories or in “reading between the lines”. But also giving and grasping, walking and evading are intelligent human abilities in which goal and plan, action and context are taken into account. In most of these dimensions, humans are superior to machines, but the AI ​​advances over the past ten years are fortunately remarkable.

The biggest mistake in the current AI discourse is the assumption that the next logical step will be the self-aware machine. When the difference between "general artificial intelligence" and "artificial general intelligence" is obscured.

The first that "General artificial intelligence", is the hope. AI as a construction kit and universal tool. The second is fiction, the "Artificial general intelligence"that can only be thought of as a machine with motives, its own will and ideas, with its own goals and plans. The desirability of such an artificial form of existence, which is completely unknown to us, is at least doubtful. It does not currently exist and cannot be expected in the medium term. A machine has no productive imagination, no self-efficacy, no capacity to sufferIn this respect, we shouldn't allow ourselves to be hypnotized by blockbusters and bestsellers, but rather enable existing tools and develop future ones so that we are smarter faster and solve problems better.

Universities and students have the same goals. The course is intended to enable students to make a constructive and economically productive contribution to the community so that they can lead a fulfilled and happy life in which they combine work and family and creatively promote social development. This means all disciplines and subject areas. What can AI do and how can AI help on these paths and to achieve goals?

AI accelerates the path to knowledge - that is the honeymoon - for those who are on this path, who want to understand something, who have questions, who are looking for the knowledge and science to make progress. Education is discursive and the university is the space for discussion.AI does not replace the interlocutor. AI improves research, improves preparation for teachers and follow-up for learners, but does not automate understanding. Everyone has to do that themselves. The penny only falls in your own head. You still have to keep what you have understood yourself. How can AI accelerate understanding? By answering any questions. And quickly and at any time. We should therefore continue to develop AI in Germany to enable knowledge dialogues with Wikipedia. It is not about questions about facts, places, data, but about a man-machine dialogue, about positions and relationships, about weighing effects and consequences on the basis of the evaluation of the available oversized knowledge spaces and knowing the causes and Reasons.

This article was written as part of the KI Week 2021 of the Hochschulforum Digitisierung (HFD). There is more on the subject in the artificial intelligence dossier on the HFD website.