Ethics and the power of AI – having a chat with Minister Bart Somers

More and more cities, countries and organisations want to commit to the digitalisation of their governments or administrations. Mechelen, a city in the Flemish part of Belgium, is one of them. Coincidentally, Arinti – known for its AI solutions – is located on the outskirts of this vibrant city. On the 10th of February 2022, the mayor of Mechelen – who is also the Vice-Minister-President of the Flemish Government – met up with Arinti and Microsoft to discuss the possibilities of AI and its ethical consequences.

Municipalities without town halls

According to Bart Somers, who is the mayor of Mechelen, Vice-Minister-President of the Government of Flanders and Flemish Minister for Home affairs, Administrative affairs, Civic integration and Equal opportunities, Mechelen as well as the Government of Flanders should take steps to enter the new digital era. “We should be able to move to a situation in which people are able to do everything that has to do with government services from home, work or wherever they are, without having to physically go to a town hall”, he explains. To make this happen, the Government of Flanders launched a project for local administrations to make their services digital proof. “We must encourage and inspire our governments to incorporate the digital options that are already available. We do, however, realise that some people have yet to enter this digital reality”, Somers adds.

We must encourage and inspire our governments to incorporate the digital tools that are already available.

– Minister Bart Somers

Therefore, an important task for those who want to take their organisation, government or administration to the next level is to make sure that their users have sufficient skills and knowledge to follow in their footsteps. Moreover, there are and will always be people who prefer not to use the internet and instead prefer to go, for example, to the town hall themselves. “And this should always remain possible”, Somers clarifies. “By realising more and more things digitally, we free up more capacity for personal, physical, customised service provision, which will always be important.”

The power of AI

Innovation happens very quickly. That is why municipalities and governments need to keep up in this digital age, so they won’t get too far behind. But the Government of Flanders does not have all the knowhow and creativity in-house to create the necessary digital tools and environments, which is why they are working together with Arinti. “Arinti is an interesting company. We want to work with them and learn from them. What are they doing today? What will they be able to do in the future? By learning from them, we can optimise our digital future”, Somers concludes.

Innovation happens very quickly. That is why we have to facilitate municipalities and governments, so they don’t get too far behind. Arinti can help them accomplish that.

– Minister Bart Somers

In most cases today, AI saves time and money by automating repetitive processes and tasks. “But artificial intelligence can do a lot more than just that”, says Sammy Deprez, Managing Partner and Senior Consultant at Arinti. “For example, it can reduce stress, put you at ease, provide new insights, and prevent frustration. These things are not easy to demonstrate: you can’t feel thirty per cent more at ease. But the effect is still there.”

An example will clarify this. The city of Mechelen has a chatbot called Yana. It asks young people going out at night where they are going, and when they think they will be back. By gathering this information, the chatbot is able to inform someone when something goes wrong. Furthermore, Yana can simulate phone calls. By doing all of this, it puts young people at ease.

The power of a chatbot is that it helps people out immediately. Whether someone wants to ask an organisation for information, complain to a company about a product, or just feel less alone: anyone can talk or chat with a bot whenever, wherever. Another advantage of these bots is that they are multilingual. The bot doesn’t need to speak the same language to be able to understand someone. It will, for instance, recognise a component in a specific language, translate it into Dutch – which it understands –, and translate it back to the original language to answer. That, again, is the power of AI.

Pre-built blocks make it possible for everyone to start working with AI right away.

– Sammy Deprez, Arinti

“The best thing about all of this is the fact that you can just invest in pre-built blocks, for example a pre-built chatbot, and get started with AI right away. These blocks are readily available and easy to use”, Deprez explains. “Furthermore, Arinti can help you optimise a pre-built block and adapt it according to your specific wants and needs.”

Addressing concerns about AI

By now, it should be clear that artificial intelligence is needed to join the digital future. But that is easier said than done. Many people know AI from movies or TV-shows and think that it is something to be afraid of. When it comes to AI, they are concerned about matters related to privacy, governmental control, and transparency. “With artificial intelligence, people quickly think in doomsday scenarios”, Deprez says. “But computers are still just computers. They are not the same as actual human beings. And they will probably never be.”

Computers are still just that: computers. They are not the same as actual human beings. And they will probably never be.

– Sammy Deprez, Arinti

To deal with these concerns successfully while implementing AI, governments and administrations need to think about more than just what it can do for them and their users. They need to think about privacy regulations and how to properly inform their users about it. “That is why you need an ethical framework”, Deprez continues. “This already exists on a European and even on a global level. Within this framework, different cities and countries share their knowledge about and experiences with AI so they can learn from each other and won’t make the same mistakes.” Imagine, for example, the following scenario: a local government wants to start using facial recognition. It knows, however, that it is quite controversial to use, and that opinions about it are divided. To learn more about it without having to try it out immediately, the ethical framework provides all the information others have gathered about it. Therefore, the local government can learn from their successes – and mistakes.

Perks and problems of deep learning

AI brings many advantages – but people need to be aware of what’s behind it. The problem is that you don’t always know what’s going on with AI. This is especially true when talking about deep learning. This is a technique AI can use to learn continuously. The more information it gets, the more it knows, and the more accurate its answers or predictions will be. “With deep learning, you can actually go so deep that you no longer know how something was calculated”, Koen Everaert, Managing Partner at Arinti, explains. “You no longer have any insights into how AI got to a certain answer or what data is behind an outcome. This can be problematic.”

If you are going to make predictions or start grouping people based on historical data, you have to make sure the data is right.

– Koen Everaert, Arinti

Imagine, for example, that someone is building a machine that predicts whether a person could ever become a doctor. “If we were to do that in the 1960s, the odds for a man would have been much higher, simply because almost all doctors were men”, Everaert says. “You have to take that into account. If you are going to make predictions or start grouping people based on historical data, you have to make sure the data is right. Sometimes you need to add or generate more data to make sure it’s correct. It’s important to be aware of that.”

Responsible AI

Koen Van Tolhuyzen works as a Business Manager Local Government and Smart Cities at Microsoft. He specialises in ethics and AI. “Microsoft is working closely with the EU and with other governments to put together a framework for responsible AI”, he says. “Our chief legal officer, Brad Smith, has written a book about it that contains an important statement: the more powerful the tool, the greater the benefit – and the greater the damage it could do. That is definitely true for AI.”

The more powerful the tool, the greater the benefit – and the greater the damage it could do. That is definitely true for AI.

– Koen Van Tolhuyzen, Microsoft

Van Tolhuyzen is referring to the fact that more and more hackers – and certain governments – are trying to abuse AI for their own benefit. The intelligence that comes with AI could be very dangerous, and we have to take that into account. “But we have to keep in mind that this is about innovation, that we don’t need to slow down this process”, Van Tolhuyzen adds. “The fact that AI could be dangerous, should never be a reason to stop us from using it. Governments often seem to be afraid of it and that is the real problem here.”

Therefore, Microsoft defined a framework which makes it possible to join the digital era without having to be afraid. “We want to innovate responsibly. It has to be safe. Our job is to make sure that governments and businesses continue to work and put our technology to good use”, Van Tolhuyzen concludes. Microsoft’s framework is based on six key principles: fairness, reliability, privacy, inclusivity, transparency, and accountability. These principles – and the framework as a whole – can be a great guide for everyone who wants to work with AI.

The fact that AI could be dangerous, should never be a reason to stop you from using it

– Koen Van Tolhuyzen, Microsoft

“Being scared of the power of AI should never be a reason to stop using it”, Van Tolhuyzen concludes. “Governments often seem to be afraid, and that is the real problem here. As a government, you need a framework, you don’t need to slow down. Otherwise, you will fall behind other countries where innovation is still booming.”

Looking for such a framework, or are you interested in participating in the conversation? Come chat with us over an informal coffee, we’re always have to trade thoughts!