Upper Norwood Presentation

I was invited as a speaker at the Upper Norwood Library Hub in South London on March 4, 2023, as part of a newly funded Minderoo-Oxford challenge project, to talk about Age-Appropriate AI for children. I do not blog much, but I thought it was such a brilliant event and there were so many amazing inputs from the participants in the room and online that I must make a note to myself and share.

My goal for the talk is really twofold: demystifying AI to a possibly very diverse audience, and helping people take a more critical thinking about the good and less good aspects of AI in our life. I started the talk by laying out how AI begins as nothing more complex than a set of instructions that computers can understand, execute and make predictions. Now AI is much more sophisticated and starts to become closer to computers that think like humans thanks to the vast amount of data we have access to and the powerful computers that can carry out computations at a much faster speed.

Having made sure that the audience followed the basics about AI and how modern AI (such as reinforce learning — although not by using words like that) works, I had the basis to then discuss with the audience how AI is not just a science fiction, or exists in remote science labs, but it is basically everywhere in our life. Everything from voice assistants like Alexa or Echo, to online social media platforms or recommender systems, to systems used to help children learn better or support children with learning difficulties or disabilities, the use of AI is pervasive and everywhere. I must add, after my presentation, many audiences came up to me and mentioned how surprised they were to learn this existence and deep embedding of AI in our life.

These observations and reflections from the audiences echoed so much with my previous engagements with other audiences. People tend to think AI as big machines and used for addressing serious problems; not realising that AI can be just around them and involve their personal life and personal data.

I then followed up the story by analysing the various design decisions that could be made during the creation of an AI system, such as the use of personalisation and various behaviour engineering techniques. These design choices may have been motivated by the goal of creating a more usable system for the users at the start, but now they are often associated with some “less good” implications for the users, such as echo chambers or screentime addictions. Although this may be a concept widely recognised in the academia, all feedback we received from the audience showed how much more public engagement is needed to make such a phenomenon much more accessible to the general public.

I closed the talk by giving some pointers to useful books that give further introductions about how AI affects our life and how our Oxford research is addressing Age-Appropriate AI system by including children’s voices and engaging with children of diverse backgrounds.

  • Inputs from the audience The reason I wanted to write this blog was mainly because the amazing inputs we received from the audience.

After the presentation, the room was divided in 3-4 breakout groups to discuss three main questions: 1) what you would like to see AI in the future, 2) what you think about AI for young, and 3) what changes are needed.

I gathered the different reports from different breakout groups into the following three key points.

  • Demand changes to the current approach of educating AI at school The talk was well-attended by many young people from local schools as well as afar. It has been amazing to hear immediately after the presentation how much these young people demand for more discussions at their schools about AI-related practices and impact, which are quite limited at the moment. Although our existing research has always indicated this importance of filling in the gap of children’s digital literacy, which is a crucial factor to affect how much children can make sense of datafication practices around them, or empower them with designs that we are creating for them. It was most rewarding to hear from an entirely different group of young people who immediately recognised this need and shared this urgency with all other children and young people we have engaged with.

  • Demand for more public engagements to demystifying AI A lot of the adult audiences in the room expressed their surprise of how little public engagement there has been related to AI, apart from media articles. They found today’s direct engagement provided them with much more coherent knowledge and understanding that are missing in media articles, which have provided fewer discussions about what AI actually is and how it relates to them personally. This must have been an amazing message for my colleagues who are running this wonderful event and secured this specific funding to support such an initiative, which I hope is a start of many more to come.

  • Demand more engagements and guidance from the UK government for safeguarding citizen’s rights

Finally, many groups raised the questions about what should be the answer to all the big techs who are currently taking such a dominant role in our digital life with their invasive designs and systems, what the role of UK government should play to protect citizens’ right and protect the alternatives for them. Although we recognise that the UK government is making pioneer changes to the legislation space for safeguarding, there is still a lack of a strong message, or coherent strategy, about what the government may provide as an alternative to the solutions provided by the big tech – how citizens can have access to more personalised medical care without worrying how their data may be misused, how children’s learning and development can be supported with AI-based technologies without their digital life being invasively monitored and influenced by the invisible power of personalisation.

It has been a great Saturday that I have learnt massively from the community! Thank you again for having me Upper Norwood!