15. May 2023Insight

Report from Nordic AI Summit: The future of ethics in Artificial intelligence

At GoodRequest, we take the future seriously. With the rise of AI, we are fully aware of its possible implications. To better understand this technology, we attended this year's Nordic AI in Media Summit in Copenhagen. Experts from all over Europe and Nordic media gathered here to discuss AI's current state in media. Here is a summary of three particularly interesting talks and questions that we should all consider.

Beata BiskupicovaGraphics Designer

Mikkel Flyvebom is a professor at Copenhagen Business School. In his talk, he discusses recommendation systems. He explains that recommendation systems are information feeds, such as those seen on Netflix or Facebook. These systems have become a critical part of our daily lives, influencing what we read, watch, and buy. However, they are not neutral and can have negative ethical, cultural, and political implications.

ChatGPT reached 100 million users in just two months.

For instance, why is being an astronaut the most desirable job for a 14-year-old in China, while in the USA it is being an influencer? How is it that a child's TikTok feed in China differs significantly from one in the USA? And why does a child in China see primarily content related to child's age, such as explainer videos, scientific experiments, while a child in the USA has a feed that is mostly unfiltered with entirely different content?

Flyvebom argues that we need to have meaningful conversations about these systems and consider their context of use. For instance, we should prioritise creating recommender systems that are aligned with our values and provide us with meaningful content - especially for those who will shape the next generation.

New SnapChat feature giving recommendations to a 13 years old on how to plan a special date

Bo Bergstedt, a representative from Danish TV2, gave a fascinating talk on the future of deepfakes and AI-generated content. He highlighted the power of AI to create hyper-realistic fake images and videos, which can be used to spread disinformation and undermine trust in the media. For example, how can celebrities like Matthew McConaughey, Nicki Minaj, Greta Thunberg, and Stormzy be featured in the same TV show without ever appearing together in real life?

Members of the Danish parliament turned into superheroes.

With the rise of deepfakes, it is crucial to ensure that we can trust the authenticity of the content we consume. Additionally, we need to develop future technology to detect and combat deepfakes effectively.

It only takes 12 pictures for Stable Diffusion and 30 minutes of training to create fake images of anyone.

Agnes Stenbom's talk focused on imagining possible futures. She emphasized the need for media content to adapt rapidly. According to her, over 32% of people in Sweden avoid news due to its stressful and negative nature, and this number is even higher globally. This raises questions about the importance of education and how to prevent the spread of disinformation. Stenbom argued that AI can provide new ways of consuming media with a clear focus on users, but the industry needs to quickly transform and reconsider its role in this new landscape. It's important to imagine all possible futures, not just one.

Overall, the Nordic AI in Media Summit 2023 provided a thought-provoking glimpse into the future of AI in media. The conference included two fully-packed days of new information and ideas, including some difficult ethical, political, and cultural questions. It's essential to keep asking these questions and have meaningful discussions on these topics, as artificial intelligence will inevitably become a part of our lives sooner than we may have imagined.

Beata BiskupicovaGraphics Designer