educationtechnologyinsights

Will the AI Winter End for Higher Education?

By Tom Andriola, CIO, University of California

Tom Andriola, CIO, University of California

Artificial Intelligence (AI) is all the rage, and it is being designated as a transformer for every industry – from financial services to healthcare to farming. Higher education is not immune to this hype, and more and more vendors are talking about the AI in their products. Plus, new entrants are promising to transform education through their AI-infused tools. 

To cut through the hype, we need to understand what AI really is. It is not one single thing but rather a collection of no less than 19 technologies that can be leveraged individually or in combination to achieve a differentiating capability for an organization. Natural language processing (NLP), virtual agents, robotic processes automation (RPA), deep learning – yes, they are different aspects of the whole AI space.

For me, AI follows Amara’s Law on the effect of technology: We tend to overestimate the effect of a technology in the short run and underestimate its effect in the long run. I believe that the long-term effect on higher education will be transformative, but it will take time – both for these capabilities to become mature and also for student, faculty, and staff expectations and comfort levels of AI to drive a high degree of adoption.

I believe we will see an evolution versus a disruption. Indeed, some institutions are starting to experiment now with chatbots and virtual agents to handle basic interactions. Others have already integrated virtual reality into their educational settings – an early AI win.

I predict the pressure to make education more affordable will influence more institutions to use RPAto lower the costs of back-end processing. For example, maybe we’ll see more AI-powered “knowledge worker assistants” take on the rote work in a queue, allowing staff to focus on more complex cases. Eventually, we’ll see intelligent machines using calibrated and validated models for simple, repetitive user cases.

As time goes by, and as this “new normal” sets in, we will have worked through privacy and other concerns so that they ultimately do not impede AI’s progress. At that point, none of this will seem like such a big deal. In much the same way, if you go back in memory, the World Wide Web and even mobile devices seemed scary and gave us trepidation about embracing them. Now can we even imagine a world where we wouldn’t have help them in all aspects of our lives? In just the same way, we’ll move more aggressively over time into using more complex AI capabilities.

"I do believe AI will ultimately have a profound and beneficial impact on higher education, as it will on many aspects of our society"

Once we do, the powerful possibilities of AI will kick in. The traditional analytics work on student retention and graduation rates will be calibrated by algorithms that have been well-validated across diverse populations and data sets, allowing consistent results across institutions. Admissions and financial aid processes may work very differently in a world where the combination of RPA and deep learning can offer more clarity and consistency in decision making.

Sure, arguments can be made about biases in the models. Those discussions will only improve the models. And the truth is, biases exist in our human-based processes today. Maybe we will find that, AI can do a better job in determining what courses meet an institution’s criteria for the transfer of credits. Or maybe by 2025, combinations of different AI technologies will truly enable the unbundling and re-bundling of educational credentials?

By 2030, does anyone expect that the educational experience won’t be transformed? (Compare what higher education looked like in 2009 versus what it looks like now.) I see the future student experience as being heavily supported and facilitated by a “digital education assistant.” After all, by that time think how much more we’ll understand about the human brain and thus be able to shape education, not only to enhance the learning experience but also to personalize it by learner type (e.g., auditory or visual learners). The journey will be informed by data – structured and unstructured, institutional and student supplied. And underlying machine and deep-learning models will only facilitate and optimize the whole educational process by leveraging each student’s “digital twin.”

So yes, I do believe AI will ultimately have a profound and beneficial impact on higher education, as it will on many aspects of our society. Like so many of the recent technology movements, it will happen over time and in uneven ways. Our institutions will evolve as they always have, but the result for me is clear. When we look back, we’ll ask how could we ever have done without it?

Read Also

Taming the Wild IT Ecosystem

Taming the Wild IT Ecosystem

Vince Kellen, PhD, CIO, University of California San Diego
The 3 Biggest Challenges of Managing Development Across Time Zones

The 3 Biggest Challenges of Managing Development Across Time Zones

Saunak Ranjitkar, CIO/Chief Software Architect, Spiralogics
IT Must be at the Decision-Making Table in Higher Education

IT Must be at the Decision-Making Table in Higher Education

Dr. Mira Lalovic-Hand, CIO & SVP, Rowan University
Innovation and Technology: The Crossroads of Education and Industry

Innovation and Technology: The Crossroads of Education and Industry

John Wensveen, Ph.D., Vice Provost of Academic Schools, Miami Dade College

Weekly Brief

New Editions