Since 1975
  • facebook
  • twitter

AI can bridge gap between industry and academia, Riyadh summit told

Long viewed as having competing interests, the two sectors could work in harmony, opening major opportunities for both, panelists said on the final day of the summit. (GAIN/File)
Long viewed as having competing interests, the two sectors could work in harmony, opening major opportunities for both, panelists said on the final day of the summit. (GAIN/File)
Short Url:
13 Sep 2024 04:09:05 GMT9
13 Sep 2024 04:09:05 GMT9

Tamara Aboalsaud

RIYADH: AI could be the key to breaking the long-running rivalry between industry and academia, experts have told the Global AI Summit in Riyadh.

Long viewed as having competing interests, the two sectors could work in harmony, opening major opportunities for both, panelists said on the final day of the summit.

Ahmed Serag, professor and director of AI Innovation Lab at Weill Cornell Medicine in Qatar, said that though academia and industry often operate in parallel, their differing “incentives and priorities” can create challenges for collaboration.

“They both have different incentives and priorities. Academia has been, most of the time, if not all the time, about advancing knowledge — which seems to take long time frames.

“(The outcome) also gets measured in publications and peer recognition. Looking at industry, on the other hand, they tend to build products that will provide return on investment — basically, generating profit,” he said.

Serag attributed the delay in applying academic research to industry’s domination of resources.

This is evident in the AI field, where talent, data and infrastructure are heavily concentrated in the private sector, he said.

However, Chuck Yoo, executive vice president for research affairs at Korea University, said that there is potential to reverse the trend.

“These days with the AI era, I’m seeing a huge change in how academia and industry collaborate,” he said, adding AI’s rapid development is the key to bridging the gap between academia and industry.

Serag highlighted the importance of effective communication in solving the issue.

“One of the solutions to this (communication problem) is, for example, to have programs or fellowships where interns or Ph.D. students could spend some time in the industry,” he said.

This would “expand their perspectives and give them a taste of how their work could apply in the real world,” he added.

A common trap in academia is falling into what academics call an “endless loop of research,” a problem that industry rarely faces due to financial incentives, Serag said.

More collaboration could prevent the issue by giving researchers a clearer picture of targets, he added.

“There have also been very good initiatives like building joint research centers and research labs,” Serag said, highlighting facilities formed between the Saudi Data and Artifical Intelligence Authority, King Abdullah University ofScience and Technology, and King Fahd University of Petroleum and Minerals.

Establishing shared intellectual property agreements at an early stage is key to bridging the gap, the panelists said.

“This is a fundamental part of why the industry, the company, wants to protect the rights to use this technology, and on the other hand the university wants to publish, to get recognition, which is why we call publications ‘the currency of academia’,” Serag added.

One solution is to “have a buffer where you agree on a patent on this (technology) first, and then for the university just to publish that after,” he said.

Abdulmuhsen Al-Ajaji, vice president of cloud software and services at Ericsson Saudi Arabia, said that more and more academics are taking examples from the industrial world.

“Universities are now launching their own accelerators, their own incubators and VCs (venture capitals), and investing directly in companies and startups to not only be part of the research, but also more toward owning that IP, commercializing that IP and just launching it for the public,” he said.

But industry’s exploitation of academic research is a long-running trend that will prove difficult to break, Serag said.

“The first leap that happened in 2012, where we managed to get most of the advanced algorithms for AI vision based on neural networks, actually started from academia; with the ImageNet competition that was organized by researchers, and then Google took over and scaled it up with their resources, and it is now part of all of the models we use for self-driving cars, medical diagnosing and so on.”

topics
Most Popular
Recommended

return to top