RIYADH: The Saudi Data and AI Authority participated in the 2023 Conference on Empirical Methods in Natural Language Processing in Singapore, highlighting its groundbreaking advancements in large-language-model development.

Areeb Al-Owisheq, the AI adviser representing the authority, took part in a workshop titled “Arabic Language Models: Challenges and Opportunities” during the five-day conference, which runs until Dec. 10.

She showcased the authority’s ambitious initiatives, including the construction of expansive language models, the curation of extensive Arabic datasets, and the development of linguistic models tailored for the Arabic language.

Al-Owisheq underscored the pivotal role of safety considerations in the development of these AI models and noted that “large language models are sophisticated deep-learning constructs trained on vast datasets. Natural Language Processing, a pivotal aspect of artificial intelligence, empowers computers to comprehend, generate, and process human language, allowing effective querying of data using natural language text or speech.”

SDAIA’s participation in the conference reflects its commitment to showcasing Saudi Arabia’s advancements in data and artificial intelligence. The authority aims to position itself as a global leader in these cutting-edge technologies.

These efforts align with SDAIA’S initiatives to transfer, localize, and generate knowledge in the field of artificial intelligence, and to integrate it into the Kingdom’s digital transformation endeavors outlined in Saudi Vision 2030.

SDAIA also unveiled the Sawtak (Your Voice) system during the event. Sawtak employs generative AI techniques to convert speech into text, and, the authority says, “boasts unparalleled precision in recognizing various local dialects and surpassing existing applications in the market.”