Are you ready to put your machine-learning skills to the test? The latest challenge from Humane Intelligence is calling on researchers to develop two different models to tackle extremism online. The first task, tailored for those with intermediate skills, involves identifying hateful images. The second challenge, considered advanced, requires creating a model that can deceive the first one. It’s a real-world scenario where the good guys devise a solution, only to be thwarted by the bad guys with a new approach.
This initiative aims to inspire innovation in the field of machine learning to combat extremism effectively. However, the core challenge lies in the fact that hate-based propaganda can be deeply rooted in context, making it difficult for an AI model to recognize it without a thorough understanding of the symbols and meanings involved.
Professor Jimmy Lin of the University of Waterloo highlights the importance of training AI models with diverse examples to enhance their detection capabilities. Cultural contexts play a significant role in identifying extremist content, which is why Humane Intelligence has partnered with a Nordic counterterrorism group for this challenge.
While algorithmic advancements are crucial, Professor Lin emphasizes the value of education and literacy efforts in the long run. The battle against fake content and extremism may require a multifaceted approach that goes beyond technological solutions.
The deadline for the challenge is November 7, 2024. Participants stand a chance to win $4,000 for the intermediate challenge and $6,000 for the advanced challenge. Additionally, Revontulet will review the winning models for potential integration into its toolkit for combating extremism.
Join the fight against online extremism and showcase your machine-learning skills to make a difference!