Teaching at AI-focused university adapts as advances are made

Classes at world’s first postgraduate AI institution can pivot mid-course to explore breakthroughs, says MBZUAI acting provost

十一月 8, 2023
Timothy Baldwin, at Mohamed bin Zayed University of Artificial Intelligence at Masdar City in Abu Dhabi , United Arab Emirates
Source: Christopher Pike/MBZUAI

Browse the full results of the Arab University Rankings 2023

“AI was somehow a dirty word,” Timothy Baldwin says as he recalls academic conferences from decades ago when no one wanted to be identified as the “AI guy”.

Fast-forward to today and not only has AI become a significant disruptive force whose potential excites and frightens in equal measure, but Baldwin is acting provost at Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI), the world’s first institution dedicated exclusively to postgraduate-level study of AI.

The university in Abu Dhabi, United Arab Emirates, was set up in 2019, and Baldwin, an Australian, has been teaching there for nearly two years.

While many institutions teach AI studies in elective or add-on courses, the subject is the focus of all learning at MBZUAI, he says. Students enrol in either research master’s degrees or PhD programmes and are able to have a “sort of pure AI education”, explains Baldwin.

Of course, teaching AI also means keeping up with AI. “It’s a field that is moving very fast at the moment. And it is only accelerating,” the professor says.

Rapid advances keep faculty on their toes, requiring them to frequently update the curriculum. But, says Baldwin, this is easier to do at MBZUAI than at many other universities because the institution is young and specialised. It is also still small, with just 280 students.

“As a sort of clean-slate university, we really have a unique opportunity to design a modern curriculum and to update our course content on a regular basis to keep pace with all of the advances in the field, to make sure that our students are absolutely getting the state-of-the-art education,” he says.

The core curriculum remains “largely constant” (it is reviewed every year, not more often), while new electives are added as novel sub-areas of AI pop up. These moving parts of the curriculum are informed by academics’ research and international partnerships, he adds.

The university’s small student-to-staff ratio, just 5:1, allows classes to be more organic, Baldwin explains.

“Maybe there’s some breakthrough paper which is published part-way through a particular course, and so then that becomes a centrepiece of the next couple of weeks in the course. The lecturer can pick up on the fact that ‘OK, this is radically going to change the field’,” he says.

AI is a big, growing discipline, and the university has dedicated departments for machine learning, natural language processing and computer vision. Most recently, it added interdisciplinary departments – one for computer science and one for robotics.

Still, there are limitations to such a focused university model. Unlike some traditional tech-centred institutions such as the Massachusetts Institute of Technology, for example, which has a whole range of departments on campus, MBZUAI must seek partnerships off-site in order to collaborate with other disciplines.

“For many of those collaborations, you don’t have them in the neighbouring building or within a stone’s throw. But many of our faculty bring their networks of international collaborations with them,” says Baldwin. Among the faculty of 60 are 20 different nationalities.

Among these is a partnership with the Israel-based Weizmann Institute of Science, which itself specialises in the natural sciences. Baldwin says the two institutions are well matched because one has “incredible strength in the natural sciences” and leads “more of the theory side of things” while the other brings “core AI expertise”.

For example, by applying AI to pathological data collected at Weizmann, the two institutions have been trying to find out why a certain cohort of patients have a particular comorbidity. According to Baldwin, MBZUAI’s expertise has been helpful in providing the data and answering the “what” question, while Weizmann explores the factors behind the “why” question.

MBZUAI, which is named after the president of the UAE, derives its premise directly from the country's National Strategy for AI 2031. In the strategy document, the president’s right-hand man, vice-president and prime minister Mohammed bin Rashid Al Maktoum, writes: “The UAE will build an AI economy, not wait for one.”

The strategy also highlights that the UAE is a young country, keen to build its research capabilities by leveraging its foresight, investment and centralised global location to attract leading academics.

Baldwin compares the intentionality of this strategy with his two decades of working in Australia, where he says AI research appears to take place more “by accident rather than design”. It is driven mostly by talented individuals, rather than by a determined national push.

He describes the strategy-driven mindset in the UAE as a “rarefied atmosphere” and “quite the contrast”. In addition, there is for him also the thrill of helping to build a university from the ground up and the opportunity to revolutionise all kinds of industries with AI research.

“It’s [about] being very much at the centre of making this incredibly powerful narrative come about.”

tiya.alexander@timeshighereducation.com

请先注册再继续

为何要注册?

  • 注册是免费的,而且十分便捷
  • 注册成功后,您每月可免费阅读3篇文章
  • 订阅我们的邮件
注册
Please 登录 or 注册 to read this article.