The Municipality of Amsterdam has a Civic AI Lab since December 10, 2020. It is a group of researchers developing artificial intelligence (AI) technologies for, for example, education, welfare and resident mobility. In the development of AI models, particular attention is paid to impartiality and equal treatment. Scientific director and initiator Chenne Ghebreb (47) explains why this is necessary.
‘Municipalities and governments are increasingly relying on AI systems to make decisions about citizens. The effects of those systems can have far-reaching consequences. Look at the problems with the issue of benefits in tax and customs administration’, outlines Chenne Ghebrebe. A neuroinformatician and AI expert, he is the scientific director and founder of the Civic AI Lab. According to de Volkskrant, he belonged to the country The two hundred most influential Dutch people of 2020.
Kebreb, ranked 195th, is praised for his role as a neuro-informatics expert, his professorship at the University of Amsterdam (UvA) and his expertise in AI, a ‘highly topical field’ according to the newspaper. The profile ends with the sentence ‘Get three million euros from the government to fight racism with AI’.
As one of two hundred influential Dutch people, the Eritrean by birth responded to the election somewhat hypocritically: ‘Well…, I read that too, you weren’t informed in advance, so it was a surprise. I am honoured, but I am very pleased with the attention being paid to my department and my laboratory.’ As a scientist, he finds the explanation of ‘fighting racism with AI’ somewhat simplistic. A detailed description of the objectives of the laboratory follows. For example, he wants to bridge the gap between basic and applied research and uses the word ‘inclusion’ as it has a more positive connotation than the word ‘racism’.
‘High emphasis on technology’
In the current focus on AI, the emphasis is on AI’s technological promises, Ghebreab says. Programmers and ICT experts do not pay enough attention to basic human rights and social consequences of system outcomes. According to the AI expert, discovery of consequences and protection of minorities often comes second.
He gives many examples. ‘Think about the benefits issue in the Tax and Customs Administration, where thousands of parents with migrant backgrounds are being unfairly left behind. It is about racial profiling via algorithms, The Dutch established a data protection authority earlier this year.’ He also cites a recent example in the United Kingdom where an algorithm was used to predict final exam results for students who could not take the exam due to the coronavirus. Kebreb: ‘People thought it would be a quick and cost-effective way to solve final exam problems. However, the algorithm was partly based on location rather than individual performance, which put students from socio-economically disadvantaged areas at a disadvantage.
As a third example, he mentions an algorithm that the largest health care companies and health insurance companies in the United States apply to two hundred million American citizens. ‘That mechanism leaves blacks with less health risk than whites, meaning blacks have less access to health care.’
VU and UvA
In the current rise of AI technology, more attention is paid to algorithms and less to the effects on people. The Civic AI Lab, a collaboration between the Municipality of Amsterdam, VU University Amsterdam (VU) and the University of Amsterdam (UvA), aims to change that.
Ghebraab initially wanted to launch the lab by the end of 2019, but was unable to secure the necessary funding. One of the financiers backed out at the last minute, meaning he could not meet the condition of starting with five PhD students. This maxim comes from ICAI, an AI network organization in the Netherlands. The minimum number of PhD applicants is set at five to make an impact. The Ministry of Interior and Kingdom Relations decided to partially fund the lab and later, on World Human Rights Day (December 10, 2020). Civic AI Lab officially opened During an online session.
According to Ghebraab, the three million euros collected will be enough to conduct research with five PhD students over the next four years. In the lab’s first series of studies, five researchers are starting studies on the use of AI in the municipality of Amsterdam. They study applications for education, welfare, environment, mobility and health. The lab will also serve as an information hub for residents and organizations with questions about AI applications and their consequences, for example, ethics, equality and potential discrimination against minorities.
Everyone is doing something with AI
“Almost everyone is working on AI these days”
Gebreb: ‘Today, AI applications are being developed and deployed outside of tech companies and universities. Nowadays almost everyone is working with AI. It doesn’t always go well. That’s why it’s important to be open about uses and consequences and get cooperation. Only in this way can we gradually improve AI and make it inclusive.
The Civic AI Lab is part of the Innovation Center for Artificial Intelligence (ICAI) is a network organization currently consisting of seventeen laboratories in seven cities in the Netherlands. Labs explore public-private and public-public collaborations around various AI applications.
Amsterdam is currently an important location for AI developments and applications that require a strong focus on ethics and equity. A laboratory is also attached AI technology for the people. It was recently Featured in New Scientist. It has a comprehensive focus on research, education and innovation in the field of AI in the Amsterdam region.
The lab is currently located in the main building of the Science Park in Amsterdam, but will move to a new AI center to be built on the campus in spring 2022.
Five Themes of the Civic AI Lab
The Civic AI Lab researches and develops social AI technologies that have fairness, equity, and inclusion as core values and that the outcome of its algorithms must be traceable. Also known as ‘Responsible and Explainable AI’.
It occurs in the following areas:
Care: Development of smart algorithms integrating data for birth and youth health care and extracting prognostic factors in and around Amsterdam, respecting fundamental human rights such as non-discrimination. https://news.google.com/rss/articles/CBMigQFodHRwczovL3d3dy5jb21wdXRhYmxlLm5sL2FydGlrZWwvYWNodGVyZ3JvbmQvb3ZlcmhlaWQvNzExMTM5Ny8xNDQ0NjkxL2luLW5pZXV3LWFpLWxhYi1rb210LWRlLW1lbnMtYWx0aWpkLW9wLWRlLWVlcnN0ZS1wbGVrLmh0bWzSAQA?oc=5Equality and Privacy.
In this field Education Mechanisms are explored to ensure greater equality in citizens’ opportunities for education and training.
basically welfare AI technology is being developed to provide insight into inequalities in overweight and obesity and the impact of social and geographic environmental factors, knowledge transfer and opportunities for sport and exercise.
In the project environment Based on image data of streets and neighborhoods, algorithms are being developed that provide a new perspective on a city’s inequality based on health, education level, crime, income and other factors.
Movement. Based on information about public transport in the city, traffic movements and the social background of citizens, a picture of mobility poverty should emerge among certain groups in certain areas of the city.
