Seeking Sustainable AI, Not "Magic Beans"

Higher education today faces an existential crisis around enrollment and financial stability. New AI technologies hold tremendous promise as a way for institutions to further automate routine tasks, raise the bar for student support, drive engagement and enrollment, and save institutions staff time and money. In fact, the 2021 Horizon Report, Teaching and Learning Edition, identified AI as a key technology to closely watch.^Footnote1

IT leaders are being asked to play critical roles in selecting and deploying AI solutions that are scalable, secure, and, above all else, sustainable. Success hinges on picking the right tools. And a thoughtful assessment of various technologies requires separating hype from reality. The following nine rules of thumb can help schools identify AI solutions that can solve real business problems over the long term.

Implement versus Build

Ask yourself, "Do I want to implement a technology that someone else is highly incentivized to continuously improve upon?" or "Do I prefer to build something myself and evolve it over time?" While the latter may have some appeal, seasoned IT professionals know the burdens and pressures that have already been placed on their departments. Identifying a solution and a technology partner that meets your needs and working with them to implement that solution is far more scalable and manageable than building and maintaining something in house.

Speed to Launch / Ease of Maintenance / Technology Advances

These three themes are interlinked. How much of a lift is required of your institution when adding an AI technology? It is important to know how quickly you can launch the technology and how the "lift" will impact your institution's ability to operate the technology in an agile manner. Moreover, when it comes to AI technology, is your institution expected to handle all of the training and support on an ongoing basis, or will a strategic partner handle the bulk of the ongoing support? Finally, know whether your technology partner will build the new technology, update the existing infrastructure over time, or both.

Ease of Use for Functional Areas

Does the technology require non-IT staff to possess IT-like skills and knowledge in order to leverage and work within the AI technology? The solution absolutely has to be one that functional-area staff can use with ease. Otherwise, there is a risk that there will not be the necessary buy-in from staff and departments that regularly use the technology.

Multi-Tenant versus Single-Tenant Technology Partner Approach

It is important to understand whether a technology partner takes a single-tenant or a multi-tenant approach to their AI technology. There are opportunities and challenges to each. A single-tenant AI approach means building your own AI models using your own data. With a single-tenant approach, a partner can provide the institution with unfettered control over the AI. This may or may not be what you want. The flip side is that without the appropriate level of focus, supervision, support, and maintenance, the AI may very well decay over time. As the number of AI tasks increases and becomes more complex and varied, AI requires both data and supervised machine learning to properly understand the nuances of human thinking and interpretation to effectively turn it into action. A multi-tenant approach allows the AI to pool anonymized data and learn across a whole network of instances while still allowing for some customization. In this scenario, you may not have 100 percent control over the AI, but you will benefit from taking more of a big-data approach to machine learning, which can bring more depth and consistency to AI actions. This model can be achieved with a common, shared platform that supports a multitude of institutions. This ensures that the advances and learning at each individual school or "tenant" benefit the broader community of users.

Underlying AI Technology

Four of the world's leading technology companies—Google, Amazon, Microsoft, and IBM—are pouring billions of dollars into artificial intelligence. Who knows where the innovations that are made in the coming years will lead higher education. A number of AI providers have built their solutions on these technology platforms. Riding the wave of these global innovators is likely to take your institution further than you would get through a partner that has built a homegrown solution.

Learning on Its Own

One of the great promises of AI is that it can "learn on its own." It's important that we recognize what is realistic versus what we see in science fiction movies. A wise college administrator once said, "Do not try to sell me 'magic beans.'" There is a great opportunity with AI, but we are still in the early days. We can start today with some quick wins: launching AI technology that learns from interactions and matures through guided machine learning to better understand the intents of questions and is able to have constructive conversations with students and prospects.

Openness to Integration with Other Technologies

It's important to understand how open a particular AI technology is to integrating with your other systems. Whether it's through APIs or more advanced integrations, AI can help you leverage the data and capabilities from your other systems. The more data you can tap into, the easier it will be to break down silos that have historically slowed institutions down. The more data you can access, the more business problems you can solve.

Built for Higher Ed

Higher education is a very large market, and it has unique qualities, terminology, and processes. While vendors of all different types may have compelling technology, some will stand out for having a singular focus on higher education. It is worth assessing how much that knowledge and insight can make a difference for your institution and the students you serve.

Joining a Community

Do you want to be part of a broader community of similar institutions that share common challenges and are willing to help support one another? It is good to know if the partner facilitates a community that enables clients to effectively collaborate, share best practices, share content when it's not proprietary, and help each other with tools that build upon the success others are having.

The Wrap

IT leaders are central figures in digital transformation across higher education. Colleges and universities need to embrace technology more than ever before. Institutions' financial health and future success depend on it, and artificial intelligence can be a constructive part of a school's innovation and advancement moving forward.

As IT leaders consider the various AI options available to their institutions, the nine rules of thumb above can help them pursue a path that best suits their institution's needs. A key goal is finding a sustainable path. Sustainability is vital to whether or not an investment in a particular AI technology is simply "magic beans" or truly delivers on its promised benefits over time.

Note

1. Kathe Pelletier, et al., 2021 EDUCAUSE Horizon Report, Teaching and Learning Edition, research report, (Boulder, CO: EDUCAUSE, April 2021). Jump back to footnote 1 in the text.↩

---

Damon Vangelis is the founder of Ocelot.