arXiv ·
A study investigated language shift from Tibetan to Arabic among Tibetan families who migrated to Saudi Arabia 70 years ago. Data from 96 participants across three age groups revealed significant intergenerational differences in language use. Younger members rarely used Tibetan, while older members used it slightly more, with a p-value of .001 indicating statistical significance.
arXiv ·
This study introduces a Probabilistic Graphical Model (PGM) framework utilizing Pearl's do-operator to causally audit LLM safety mechanisms, specifically isolating the effect of injecting cultural demographics into prompts. A large-scale empirical analysis was conducted across seven instruction-tuned models from diverse origins, including the UAE's Falcon3-7B, as well as models from the US, Europe, China, and India, using ToxiGen and BOLD datasets. The findings revealed a disparity between observational and interventional bias, demonstrating that standard fairness metrics can overestimate demographic bias. Western models exhibited higher causal refusal rates for specific demographic groups, while Eastern models showed low overall intervention rates with targeted sensitivities toward regional demographics. Why it matters: This research highlights the geopolitical nuances of LLM safety alignment and the potential for demographic-sensitive over-triggering to restrict benign discourse, which is particularly relevant for diverse regions like the Middle East in developing culturally-aware AI.
MBZUAI ·
An AI model from the University of New South Wales (UNSW) won the AI Eurovision Song Contest in 2020. Following this, UNSW researchers posed philosophical questions to an AI language model and found that respondents preferred some machine-generated answers over those from philosophers like the Dalai Lama. This raises the question of whether AI can outthink human philosophers, a topic explored through projects like Philosopher AI and attempts to emulate the human brain with neural networks. Why it matters: The exploration of AI's capacity for philosophical thought could revolutionize our understanding of intelligence and consciousness, with potential implications for AI ethics and the future of human-machine collaboration in intellectual fields within the Middle East and abroad.
TII ·
Technology Innovation Institute (TII) and Mila, the Quebec AI Institute, announced a strategic partnership to collaborate on AI safety and next-generation machine learning models. TII will establish a research lab at Mila in Montreal, enabling collaboration between UAE-based researchers and Mila's AI specialists. The partnership aims to translate scientific advances into real-world impact and strengthen the global research ecosystem. Why it matters: This collaboration enhances UAE-Canada scientific ties, positioning both communities for breakthroughs in areas like LLMs and AI safety, aligning with the UAE's vision to become a global AI research hub.