‘AI race’ leads to concerns among experts, who are also in the dark

The NOS spoke to four AI experts, who together have decades of knowledge in this field. They all think that regulations should be introduced quickly. The fact that the European Parliament took a step in that direction this weekthey think is important. They are all concerned about what is happening now, although they differ on what deserves the most attention.

US President Biden urged the bosses of major AI companies last week that the technology has “huge potential”, but also a “huge danger”.

“There are now only nine countries that have nuclear weapons,” says Wim Nuijten, scientific director at Eindhoven University of Technology’s AI institute EAISI. “Now is the time to look at similar measures for AI.”

Of the experts NOS spoke to, he is most concerned about the long-term consequences. “My main concern is that AI can do incredibly violent things.”

Without a ‘moral compass’, AI could even wipe out life on earth in the long term, says Nuijten. “I really believe that.” The hypothesis is that AI is at that moment more intelligent and powerful than humanity and makes a decision that intentionally or unintentionally ends life on earth.

The other experts nuance these concerns quite a bit. It shows that science is in full swing. Jelle Zuidema, associate professor of explainable AI at the UvA, thinks that the alarmism that is now being heard is not justified. “Of course we are a step further. But the future in which intelligence other than humans is in charge is still very far away.”

He is accompanied by Virginia Dignum, Professor of Responsible AI at the Swedish Umeå University, and Maarten de Rijke, Professor of Artificial Intelligence at the UvA. All three mainly warn about what is going on right now.

“It has become very easy to build applications on top of the AI ​​systems,” says Dignum. “That’s what I’m most worried about. We don’t let people drive a car without a driver’s license, do we?”

There is also considerable criticism of the speed with which everything is rolled out, including by the tech giants. So there is a battle going on between Google and Microsoft in particular to present new AI applications as soon as possible.

“We have arrived in a world where that technology is developed with society as a research lab,” says De Rijke. “It is too crazy that we only leave these kinds of technological developments to a handful of tech companies.” Moreover, says senior lecturer Zuidema, it is not yet clear what can go wrong with the technology.

At the same time, it is also clear that AI will offer many opportunities and will in any case lead to changes in society. Zuidema is also enthusiastic. “It’s a giant leap forward. It’s a combination of admiration and criticism.”

Despite its doomsday scenarios, it also sees great advantages in generative AI in particular, which can generate texts or, for example, images based on questions. He then thinks of new ways to combat cancer or Alzheimer’s or to grow crops.