RIYADH: Saudi Arabia's King Abdullah University of Science and Technology is accelerating research and development into generative AI through a model aligned with the Kingdom's Research, Development and Innovation Authority's Vision 2030, entering the global artificial intelligence race.
“Generative AI (GenAI) is on its way to transforming every aspect of our civilization, and has already begun to do so. Its applications in many sectors, including healthcare, industry, energy, sustainability and entertainment, will be central to Saudi Arabia’s future development,” Bernard Ghanem, director of KAUST’s Generative AI Center of Excellence, told Arab News.
KAUST announced the launch of a Center of Excellence (CoE) on Generative AI on July 1, which aims to become a premier hub for research, development and innovation in pioneering generative AI technologies aimed at solving some of the most pressing challenges facing Saudi Arabia and the world.
“The KAUST GenAI CoE will push the boundaries of what is possible with GenAI in terms of technical capabilities, applications and real-world impact,” Ghanem said.
“We believe the CoE will play a major role in advancing and accelerating the GenAI environment in Saudi Arabia and across the world, leading to an explosion of new models with real-world applications in the four national research and development priority areas that Saudi Arabia has identified.”
KAUST's mission is to enable GenAI research and development through the GenAI model and find solutions that align with the four national priority areas outlined in Saudi Arabia's RDIA Vision 2030: health and wellness, sustainability and basic needs, energy and industry, and future economy.
“Throughout its lifespan, the GenAI CoE will work with Saudi Arabian and global partners to identify specific challenges to address within each of the four pillars of RDI,” Ghanem said.
He outlined the KAUST GenAI CoE's strategy to leverage GenAI in the Kingdom's priority research and development areas.
The center aims to develop GenAI multimodal foundational models designed for clinical image analysis for health and wellness, and establish a GenAI-based drug design and development pipeline for Arab populations.
In line with sustainability, the KAUST GenAI CoE is working on developing GenAI foundational models for Earth observation data from satellite inputs and using the set foundational models for Earth observation insights, with a focus on specific use cases such as agroinformatics, ecosystem assessment, and weather forecasting and prediction.
Regarding energy and industry, Ghanem explained that the centre of excellence is working on developing and specialising GenAI foundational models in the field of chemistry.
The center uses “foundational models for chemical reaction optimization (finding the best chemical setup for a reaction to produce the best outcome) and advanced materials discovery and synthesis (combining GenAI models with an automated robotic chemistry lab to dramatically speed real-world discovery).”
Finally, with a mission to build the economy of the future, the GenAI Center of Excellence is developing and specializing multi-modal GenAI models for business and government transformation, with the goal of creating GenAI models for education, including intelligent tutoring for students and assistance for teachers.
Ghanem said the work at GenAI also extends to establishing “more expressive and efficient GenAI models for visual content creation to support growing creative industries such as social media, gaming and entertainment in general.”
“GenAI's enormous value potential is supported by recent reports which forecast the technology to add hundreds of billions of dollars to the global economy by 2030 and contribute significantly to Saudi Arabia's GDP,” Ghanem said.
This section contains relevant reference points arranged in the (Opinion column).
Ghanem explained that the mission will be carried out through three main pillars: “Innovating general-purpose GenAI models with the characteristics required for ubiquitous, efficient and reliable deployment; specializing these models for solutions across all four pillars of RDIA; and realizing the Kingdom's ambition to accelerate the adoption of GenAI in the Kingdom by focusing on translational research and talent development.”
Advances in Gen AI have raised new concerns about the technology’s negative impacts on society, including data privacy, environmental sustainability, and regional and cultural disparities in quality and coverage.
The KAUST CoE plans to address these concerns through research projects on GenAI’s reliability, efficient training and inference, and the development of Arabic language models.
Ghanem emphasized that the mission of these projects is to “usher in the next stage of the evolution of GenAI technology, with a focus on reliability, internationalization, open access and reduced environmental impact.”
The GenAI CoE will also focus on making a positive impact through GenAI training and skills development programs for KAUST researchers, partners, and the general public. Through its training outreach initiatives, the CoE hopes to address the GenAI talent shortage in Saudi Arabia.
In a press statement, the center said it recognised that more training was needed, particularly at the national level, “to really make a big impact in this regard.”
When asked what scientific, technological and upskilling challenges need to be addressed to advance Saudi Arabia's GenAI sector, Ghanem spoke of the importance of “access to big data, talent development, GenAI hardware infrastructure and GenAI investment.”
“The GenAI ecosystem in Saudi Arabia is young and thriving, and great progress has been made so far. However, several challenges remain,” Ghanem said.
“Perhaps one of the main reasons why current popular GenAI tools perform so well is the access to large-scale data for training and fine-tuning. Access to such large amounts of data is crucial for the future development of GenAI in the Kingdom. There is work being done in this regard within Saudi Arabia, but more can be done to open-source data from various organizations and entities.”
Ghanem said developing a proper GenAI environment in Saudi Arabia will require “a large-scale talent development program (i.e. GenAI for the public), which will include access to higher education in the field, but more importantly, be based on crash courses that teach non-experts the basics of GenAI development.”
Ghanem believes that access to large-scale data and significant local talent is not enough for a thriving GenAI ecosystem.
“Access to specialized hardware accelerators (such as high-end GPUs) is paramount for GenAI's large-scale training and mass inference. Unfortunately, without sufficient access to this hardware infrastructure, progress slows and the ecosystem stagnates to deliver timely impact,” he said.
Speaking about investments in GenAI, Ghanem explained that sound investments in the sector in homegrown, yet internationally competitive technologies and commercial solutions are essential for a thriving and self-reliant GenAI ecosystem.
“Efforts are ongoing in this regard, but in such a rapidly changing and constantly evolving field, there needs to be a more concerted effort to address this challenge,” he said.
“Through the CoE, new GenAI models will be developed and deployed to address the most pressing national and global challenges. We will do this while upholding the highest standards of AI ethics by strengthening key values (e.g. fairness, safety, and reliability) in our R&D pipeline.”
