The UAE has been quick to adopt AI. Now it must build and export it

During his visit to Abu Dhabi in June, OpenAI chief executive Sam Altman said the UAE 'has been talking about AI since before it was cool'. AP Photo

Ever since the UAE issued its National Strategy for AI 2031 six years ago, it has been consistently establishing milestones in the area of artificial intelligence.

There is little doubt AI holds immense economic potential for the country. PricewaterhouseCoopers forecasts this field could contribute up to 13.6 per cent of the UAE’s gross domestic product by 2030. More importantly, AI can also help respond to the country’s social challenges – including developing solutions for obesity and heart disease, reducing traffic fatalities and improving air quality and education outcomes.

AI, particularly Generative AI, is at a stage of development where it can accelerate the achievement of the National Strategy’s objectives. With regard to investment in research and development, AI could support the UAE in a number of ways. In terms of human capital development, for instance, AI can be the teacher that helps the national workforce upskill in both AI-related and other domains.

AI-powered information gathering and data analysis can help create relevant commercialisation – that is, bringing products and services to the market. In terms of policy refinements, AI can help overview the landscape of international policy to ensure the establishment of a robust AI infrastructure, including a governance and regulatory framework that establishes the UAE as a leader in ethical technology and responsible data use.

The Mohamed bin Zayed University of Artificial Intelligence in Masdar City is meant exclusively for AI research. MBZUAI

By 2030, the UAE is set to reap commercial benefits worth $5.3 billion from investments in Generative AI alone

The Oxford AI Readiness Index 2021 sheds light on the need for the region to prioritise the development of its human capital and technology sector. Presently, there is a trade deficit when it comes to knowledge-intensive services, which include R&D and technology-based services. Developing this is critical to facilitate the transition from adopting AI to one day building and exporting it. In this light, the Emirates intends to upskill a third of its Stem graduates every year.

In 2019, the Mohamed bin Zayed University for Artificial Intelligence, a graduate-level university exclusively for AI research, was established. The MBZUAI has formed a number of partnerships for AI R&D with organisations such as the Abu Dhabi Health Services Authority, the Technology Innovation Institute in Masdar City, and IBM in New York. Another step the UAE has taken is to train all leading government employees in Generative AI, while more junior public sector employees receive training as needed.

Just last week, a unit of Abu Dhabi AI company G42, MBZUAI and Silicon Valley-based Cerebras Systems launched Jais, an open-source bilingual Arabic-English Large Language Model developed in the UAE.

The potential of Generative AI extends to supporting the region’s R&D by analysing global scientific publications, patent data and funding trends to identify emerging research areas with high potential for returns.

For this purpose, the UAE Digital Economy and Remote Work Applications Office has already published a guide to Generative AI. It could serve the UAE well to extend the mandate of this office to ensure adoption and uptake of these recommendations by researchers and ecosystem developers.

By 2030, the UAE is set to reap commercial benefits worth $5.3 billion from investments in Generative AI alone – a return rate of almost 990 per cent on every dollar spent. This estimate of the outcomes of commercialisation efforts is difficult to ignore.

However, commercialisation itself is much harder to achieve. It requires the establishment of an entire ecosystem of digital upskilling, training, investment, incubation, support, collaboration and market entry – both for research endeavours as well as for commercial and industrial activity. An entirely new data infrastructure needs to be developed. And most crucially, the UAE would need to provide incentives to encourage the development of technologies and commercial enterprises focused on AI within the Emirates.

The design at the Mohammed Bin Rashid Library in Dubai incorporates technology and AI to make it as accessible as possible. AFP

The UAE has been making strides towards AI commercialisation by putting in place the ecosystem for startup development activities described above. It has also, notably, been buying Nvidia chips recently to continue to develop large language model-related applications and cloud services, including its own open-source large language model, Falcon.

The country now has an opportunity to strengthen this framework and infrastructure for commercialisation efforts – including the ability to house and responsibly share national data. Thoughtful big data governance will ensure that commercial AI applications are tailored to local needs rather than imported from elsewhere.

During a visit to Abu Dhabi this year, OpenAI chief executive Sam Altman said that the Emirates “has been talking about AI since before it was cool”. In June, the newly launched Dubai Centre for AI announced that it will use AI to refine policy. This AI will conduct simulations that analyse the potential effects of new policies, make predictions based on different scenarios and assess which interventions could be fruitful. Analysing data will help this centre identify trends and generate insights, thereby providing decision support to policymakers and public servants.

It may be useful at this point for research organisations such as these to centre the UAE’s cultural, historical, social and natural resource context. Learning from leading innovation economies, while keeping in mind strengths of the local ecosystem, the UAE has an opportunity to create policies that are fit for purpose. Both the economy and society will benefit from taking a wider and more contextual view of policy refinement with regard to AI.