Text: Linda A. Thompson – Photography: Saskia Vanderstichele
Many lawmakers, members of the public and even some tech experts are worried about the impact of artificial intelligence on our society. This means that putting such fears at rest will be vital to further development of artificial intelligence technologies, says Dr Ann Nowé. It’s why the AI-Lab, which Nowé oversees, is working to establish a new research centre that will serve as a showcase for the potential of AI, while simultaneously positioning Brussels as the AI hub of Europe.
In 2018, the late Stephen Hawking warned of the dangers of artificial intelligence, or AI, in a posthumously published book. “I fear the consequences of creating something that can match or surpass humans,” he wrote.
In a survey conducted by the European Union the year before, 72% of the polled respondents said that robots steal people’s jobs. And in February of 2017, the European Parliament adopted a resolution urging lawmakers to consider creating a special legal status for robots in order to establish who is liable if they cause damage.
Clearing up misconceptions
Clearly, some scientists, lawmakers and members of the public are concerned about the impact of artificial intelligence and robotisation on our society. According to professor Ann Nowé, assuaging fears and clearing up misperceptions will be vital to further development of artificial intelligence technologies. “One of our missions is to give people a more accurate image of AI, a more realistic one,” says Nowé, who has headed the Artificial Intelligence Laboratory, or AI-Lab since last year.
Clearing up misperceptions will be vital to further development of artificial intelligence technologies.
Ai for the common good
Artificial intelligence refers to the creation of systems and machines that mimic the functioning of the human brain. This can include – but is not limited to – data mining, or using data to find patterns. Like anything invented by humans, artificial intelligence, has the capacity to be used for good and for ill.
But let there be no doubt about it – Nowé and her colleagues use artificial intelligence for good by researching technologies that can help humans perform certain tasks more accurately, more safely and more ergonomically – or remove the need to do them at all. In that sense, AI has the potential to help policymakers address the multiple challenges they are running up against – from ageing populations and global warming to the growing scarcity of natural resources.
Nowé here gives the example of exoskeletons that help factory workers perform physical tasks that take a heavy toll on their bodies. “This exoskeleton is carrying the weight, while the human is handling the fine motor skills. That’s also a beautiful example of collaboration between AI, between robotics and a human,” she says.
That’s all a far cry from the alarmist tone some lawmakers, but also members of the tech community have struck about the likelihood of a future robot supremacy. “This is one of the big misconceptions about AI,” Nowé says. “That we’re going to see robots with a will of their own who are going to turn on humans. I don’t think that is ever going to happen.”
Soon, Nowé and her colleagues will make their goal of developing positive AI technologies even clearer by establishing a multidisciplinary, interuniversity AI for the Common Good research centre together with Collibra, the data intelligence company that spun off from the VUB. An AI Experience Centre, where researchers will be able to showcase and experiment with AI technologies, is moreover also in the works. “The idea is that it will become a meeting place for people who work with AI,” Nowé explains – for researchers as well as companies, policymakers and the general public. “Because you need to bring together many different forms of AI knowledge and expertise to arrive at the next generation of AI systems.”
Continue reading below the picture.
Transforming Brussels into the AI hub of Europe
The end goal for both projects is to ultimately transform Brussels into the AI hub of Europe. “We shouldn’t underestimate ourselves,” says Nowé. “If we bundle our expertise, nationally but preferably also beyond country borders, I don’t think we will be less-than any other city or country,” she says. “We have a very strong degree programme and our PhD students are eagerly recruited by all sorts of institutions the world over.”
The AI-Lab has a long history of weighing in on conversations about the ethical consequences of science and technology. Founded in 1983 by professor emeritus Luc Steels, – at a time when many universities didn’t even have a computer science department – it was interdisciplinary from the get-go.
“Many AI groups or labs tend to work on one aspect of AI, so for instance as part of computer science, engineering or robotics,” Nowé says, pointing out that the AI-Lab includes linguists, engineers, philosophers, psychologists. “The lab’s interdisciplinary approach is something that still sets it apart from most other labs today”, says Nowé, herself a mathematician by training.
The AI-Lab includes linguists, engineers, philosophers, psychologists. This interdisciplinary approach is something that still sets it apart from most other labs today.
In the years to come, the AI-Lab will continue working on human-like AI, so systems that display human-like intelligence, and multi-agent systems, one of the areas in which it has built up considerable expertise. Multi-agent systems are complex systems that require multiple components to interact and to collaborate. For instance, using AI to develop smart grids so that energy is produced and consumed at the most opportune moments.
Another goal, Nowé adds, is to arrive at more transparent AI technologies. To date, researchers at the AI-Lab and elsewhere have made many advances in AI that operates somewhat like a black box – their input and output is clear, but the reason that they came to a decision is a mystery. “It’d be terrific if we could place white box algorithms, which are able to explain their reasoning process and why they came to a particular conclusion, alongside this,” she says, adding that this would allow humans to learn and accrue knowledge from the AI decision-making process.
--
You can find all articles of the festive magazine here.