UZH Magazin

Cleverer Than Us?

Artificial intelligence is getting better and better. We can benefit from this development, say UZH scientists in the latest UZH Magazin. By joining forces with AI, we can grow cleverer.

Roger Nickl, Thomas Gull

Does artificial intelligence make us dumber than we are? Or will it help to solve the problems of mankind? In the video, UZH experts provide answers to these questions. (Video: UZH Kommunikation / Zentrale Informatik, Mels)

 

Artificial intelligence (AI) can already do an impressive number of things: Diagnose diseases, recognize faces, predict climate change, assess whether criminals will reoffend, translate at lightning speed and provide tailor-made purchase offers. Adaptive algorithms and computer programs are developing at an incredible rate. These smart technologies are enhancing our skills and changing the way we work and live. Digital assistants such as Siri and Alexa are already a reality. They’re not all that clever yet – but that can change. In the future, metabots with sophisticated AI applications might organize our entire digital existence and support us in our everyday tasks.

Such advances possible thanks to artificial neural networks that imitate the way our brains work. These ingenious networks are making our digital systems ever smarter and more powerful. This in turn raises expectations and hopes that our lives will become easier and more sustainable. But the rapid development also stokes fears about job loss, manipulation and discrimination. Intelligent programs are increasingly making inroads into areas that were previously reserved for people – and competing with our human intelligence, it seems. An unsettling thought.

Fears and better decisions

Will intelligent algorithms and adaptive AI systems soon outsmart us? Will they work and think for us, or will they expand and enhance our skills?  These and other questions are the focus of the latest issue of the UZH Magazin. We spoke to UZH researchers who are exploring artificial intelligence from a variety of angles. They are all involved in the UZH Digital Society Initiative, which explores matters of digitalization. From our various discussions, we learned that AI might not (yet) be smarter than we are, but by joining forces we can become cleverer. “AI has the potential to make us wiser,” says ethicist Markus Christen. Because it can pool and process huge amounts of data in a very short time, it can help us reach better decisions.

Medical informatics expert Michael Krauthammer dreams of a data warehouse in which doctors can exchange medical data worldwide to improve individualized decision-making for their patients. But there is still a long way to go before this becomes a reality, as it would require patient data to be standardized – a monumental undertaking. The legal questions surrounding the use of AI tools in medicine are no less challenging. Regulating this area is a balancing act between ensuring the greatest possible level of safety while not slowing down progress, says legal scholar Kerstin N. Vokinger. 

Amplifying prejudices

Algorithms such as the ones used by Google sort the internet for us. They determine what we get to see and what not. But nobody knows the exact criteria on which they base their decisions. Information scientist Anikó Hannák is researching their effects. What she has found out so far offers food for thought, since algorithms mirror and even amplify the prejudices in our society. Programmers should be aware of this and make sure that their algorithms don’t cause any harm. And yet, even with all these misgivings, we shouldn’t fear AI, but see it as an opportunity, says information scientist Abraham Bernstein: “Artificial intelligence has skills I don’t have, and together we can do things that I can’t do on my own.”

Roger Nickl und Thomas Gull, Editors UZH Magazin

Write Comment

Die Redaktion behält sich vor, Kommentare nicht zu publizieren. Unberücksichtigt bleiben insbesondere anonyme, ehrverletzende, rassistische, sexistische, unsachliche oder themenfremde Kommentare sowie Beiträge mit Werbeinhalten.

Number of remaining characters: 1000