While artificial intelligence (AI) can be a powerful technology for addressing
complex grand challenges such as climate change, public health, food
and water security, and economic prosperity, there is still no cohesive
way to organize all the methodologies used by AI in generating practical
solutions and to properly evaluate the science (and scientists) of AI.
Bringing academic rigor, reproducibility, and standardization to the field is,
in part, the mission of a professor who heads up the world’s first AI-focused
graduate university.In 2019, the government of the United Arab Emirates (UAE) made a
historic decision to launch Mohamed Bin Zayed University of Artificial
Intelligence (MBZUAI). Based in Abu Dhabi, the university has a simple
but bold mission: establish and continually evolve interdisciplinary,
collaborative research and development capability in the field of AI,
while educating innovators and leaders with the breadth and depth of
knowledge to grow technology and enterprise in the UAE and globally.
To achieve this ambitious, holistic directive, the board of directors
selected Dr. Eric Xing, an international leader in AI research and
entrepreneurship, to serve as its inaugural president.
Prior to his appointment, Xing was a professor of computer science at
Carnegie Mellon University in Pittsburgh, Pennsylvania, and the founder,
chairman, and chief scientist of Petuum Inc., a pioneering company that
builds standardized AI development platforms and operating systems for
industrial AI applications.
Xing pioneered the development of fundamental technologies in
machine learning, including metric learning, graphical models, and time-
evolving and sample-specific networks. More recently, he propelled theemergence of the new field of SysML—systems and machine learning—
and the technology for controlled generation of texts.
“Inside the academic community, everyone is trying to do something
exciting with AI, but there are limits in the current environment,
particularly the lack of agility and flexibility in industrial collaboration, tech
transfer, and entrepreneurship,” he says.
“There is also tension between applied and fundamental research, both
in terms of peer recognition and financial reward. Coming to MBZUAI gave
me the opportunity to start from scratch and think about big questions, like
finding the optimal system for management and administrative support
to create a healthy, vibrant organization able to do the most creative and
effective science. I am inspired to experiment with new ideas and open
opportunities in research, education, and entrepreneurship,” Xing adds.
But he has concerns about the speed and “the somewhat chaotic
development” of the AI field.
In Xing’s conception, “many things need to be rethought. This goes
beyond a choice of scientific approach and into style, taste, and public
influence.” He encourages consideration of what a university can offer to
enable the best minds to do the best work, and of what type of AI research
the academic community, rooted in the university, is in a unique position
to advance.
“Only by clarifying the scientific principles underlying AI, and solving
the problems of reproducibility and standardization of AI production, can
the real ‘engineering era of AI’ start. This is the same path trodden by
many highly mature disciplines.”
Xing believes that a clearer theoretical framework will give innovation
a better foundation, ensuring that AI implementations have a moreAdvertorial
A new paradigm in AI research and education
IMAGE COURTESY OF MBZUAI1224Product.indd 1538 12/17/21 9:45 AM