Artificial Intelligence — From Science Fiction to Reality
Strategies, Challenges & Best Practices for Cities
During the fourth edition of the Creative Bureaucracy Festival in September 2021, BABLE had the opportunity to host a session on how Artificial Intelligence (AI) systems are currently used in the urban context, and how cities are tackling the challenges and risks associated with their use.
Our consultants Gretel Schaj and Philipp Riegebauer moderated a very interesting conversation with practitioners from leading cities: Michael Donaldson, Chief Technology Officer at the Barcelona City Council, Imre-Gustav Vellamaa Board Member of R8tech, a TOP50 EU Proptech company, and Johannes Cruz Data Scientist for the Greater Helsingborg Area, Sweden.
To define AI in a simple way, we can say that it is a wide-ranging branch of computer science concerned with building smart machines capable of performing tasks that typically require human intelligence. Imre-Gustav Vellamaa said:
“You maybe remember that in 1997, Deep Blue was the first computer beating the world champion in chess, Garry Kasparov. Since then, artificial intelligence has been developed very strongly. Today artificial intelligence can run factories, cars, buildings, and a lot more.”
A surge in the utilisation of AI has been seen in recent years, with applications such as traffic monitoring which improves the flow of transit in cities and measures the level of activity in city centres. Throughout the past few years, both the public and private sectors have used, developed, and enhanced similar chatbots and applications. At the same time, some of those implementations have been criticised. For example, the use of such tools by the justice system to assess whether a convicted criminal would re-offend, the biases towards people of colour, and an algorithm that was used to recruit people and was found to have some bias against women.
Still, there are several governments around the world — especially local governments — who are taking a stance and moving forward with diverse ways to address those challenges. For instance, Amsterdam implemented a registry providing information about how AI systems are used in different public services, as well as how the data is used, how the risks are assessed while also addressing who is responsible for these systems. Governments are nowadays trying to regulate the use of AI and move forward with projects which create real and positive social impact.
As a data scientist, Johannes Cruz has been involved in many different projects for the city of Helsingborg, like the monitoring of food waste in public schools and a speaking robot named Emily which allows immigrant groups to practice their Swedish for job interviews.
“We hope it will prepare them for real life interviews and increase their chance of landing a job”
The city of Barcelona has also recently started to integrate AI in their projects, with the monitoring of the capacity of beaches during the pandemic, which follows a privacy-by-design approach.
“The cameras were just analysing the density between the people at the beach, so we would be able to send that information to a website for people to decide whether to go down to the beach or if it was too crowded. And at the same time, it gave us information in order to decide whether or not to close the beaches. It was especially important that these cameras were just providing real-time information and not recording people.”
On the other side of those implementations in cities are companies such as R8tech, a digital operator which is “developing and using AI to take commercial buildings’ technical management to the next generation.”
Agreeing with the idea of maximising the social benefits of AI, Johannes Cruz told us about Helsingborg’s open data platform where the city publishes its data:
“We hope that this can facilitate innovation with data for external parties, hopefully improving the lives of our citizens. And the more and the better data we can make public, the more democratic the digital scene will become. Perhaps with hackathons or other forms of competitions, we can find new AI solutions that will lead to societal benefit in our city.”
Unfortunately, as mentioned previously, several challenges and risks associated with the use of this technology go along with the positive impacts of Artificial Intelligence on society. Especially in these very complex times, where lockdowns have favoured a trend of hyper digitalisation, the ethics that surround the use of AI is at the heart of the public debate. Michael Donaldson acknowledged:
“We know there’s the possibility of risks, discrimination, not just from an ethical point of view, but also in the way we use the data and what we do with that data, whether we share it or not.”
As an example of those risks, he reminded us of what happened in England in 2020 — after the government reevaluated exams with a computer-generated score in response to the Coronavirus outbreak, nearly 40% of students saw their grades reduce. The software calculated test scores based on a school’s past performance on tests and on “mock” exams. Officials said the model was intended to make the system fairer by balancing out exaggerated marks given by some teachers. However, students from low-income areas with struggling schools were outraged that their futures had been left to lines of code that favoured students from private schools.
So, what can cities do to implement Artificial Intelligence and to manage the risks associated with it?
For Johannes, “AI is a mean and not an end”, and to allow an innovative culture to thrive, we must keep exploring data-powered technologies.
“I think it’s really important not to look at it as something which stands alone, but to think broader and see if AI could be included in something we already are about to do. It should be a wingman and not the focal point.”
AI thus opens new possibilities for data to be used in transformational ways, but it still requires guidelines. As AI tools become more prevalent, citizens and politicians alike need to understand them to make the most out of them. A prominent ethical consideration in all cases is consent, transparency, and data ownership, particularly the emphasis on citizens regaining control of their own data.
Data ownership continues to be a point of contention — with a general consensus that the data belongs to the citizens, and that their agreement is needed to use it during public projects for benefit. Along with data regulations, cities must take all risks into account when delegating administrative systems to technology, in order to avoid involuntary discrimination.
In addition to these very interesting insights and inputs from our speakers during this session about Artificial Intelligence systems, we recommend you visit the BABLE platform to find more AI-related Use Cases and Solutions to urban challenges for free.
Watch the full dialogue here!