AI Roadmap for Your University
One challenge is choosing to do nothing.
We talked about this malaise and the common causes here Why University's Will Miss The AI Boat
The other outcome is doing too many things that have no structure, no coordination.
This doesn't mean everything needs to be top-down command-and-control. That is a recipe for failure versus action.
But a high-level common roadmap to ensure resources are spent in the ways that first benefit the University as a whole is crucial.
And even if research is strong, unless this has already been a successful pillar, that should be left to the appropriate departments to pursue.
Starting an AI department from scratch, for example, would be difficult if there's no existing infrastructure or star personnel.
The goal of the initiative should be on attracting whatever is the core lifeblood to your university.
These are likely to be student enrollment, research grants, and alumni donations.
In this roadmap, I'm going to look at the proposals for each at a high-level. The details will need to be worked out based on your specific advantages in the respective market.
Growing Student Enrollment
Typically, student enrollment has focused on brands -- brochures, visits, so forth.
However, higher education is not immune from one of the biggest shifts that has occurred in software, which was Product Led Growth (PLG). The primary driver was self-serve: users could try the product and make progressively bigger decisions.
This currently is not something that has been done by Universities, in part because they haven't had to (they had brochures and rankings), but also because this wasn't possible before.
But with AI, this can be and the primary customers are still Parents (who are the primary customers for Universities), yet the Users should be prospective students.
How would a User experience a college or university well in advance of the one-day visit (which isn't really a good experience to begin with)?
Well, let's back into what parents and students do, in fact, struggle with on their journey to selecting a college?
And it has nothing to do with selecting the college.
It's often the learning experience of high-school.
Into this void has entered mediocre supplementary tutoring programs by humans. Most of them have focused on passing the SATs.
But with AI, it's actually possible to enable a chat or chat+voice experience based specifically on content from the universities.
Your first reaction might be that the content is not relevant; and it typically isn't.
Yet that is the problem for a number of areas: selection, as well as preparation.
But does it have to be this way?
The lowest, simplest most introductory courses at college can be made accessible to the aspiring high school student via an LLM trained on your own content.
This is the core experience, especially if the content can then be used as "credit" or helps in admission -- for both your own college or for any others, including your competitors (and I'll explain why in another letter).
The foundational basics may not be a differentiator. Physics 101 all tends to be the same in terms of concept, for example.
But the manner of presentation, tying things to courses or research different (such as the history of physics could be different); the introduction to psychology could seem like something not applicable for the typical high school student, but increasingly we are seeing examples where it can be by being the motivating application of a core skill -- writing, research, for example).
LLM's can give a chat-like experience which, unlike the more dry courses like Coursera or fixed learning experiences like YouTube become a relationship.
YouTube should not be dismissed, by the way; it is showing a clear attack vector against less nimble educational experiences.
On this playground of short videos, on-demand instruction, branded instructors and community, it's very compelling. In fact, I would argue that, long before you have AI, you should be taking every new professor through a program to expand their content now.
Imagine high school students coming to a chat interface and asking about topics or career interests, and being given an experience to help them to focus on their own thing.
This is a core differentiator which builds an early relationship while also expanding on the brands of your instructors, their research.
Yes, the other atmospheric issues are also very important: quality of other students, in-person education, dormitories, and food.
But don't miss this opportunity to now start to build out a general purpose experience where one persona is for prospective high school students to be able to go to their branded chat experience, where the content is not available general-purpose AI training models but only through yours and a curated network.
Alumni Benefits
Once you have a foundation of LLM's for prospective students, the next persona in your roadmap is for existing alumni.
This addresses a benefit to encourage more giving by addressing their own interests for career advancement or, to a lesser degree, personal development.
Educational utility to alumni can be a powerful way to extend the brand and increase giving.
One of the ways to do this is by providing through alumni, not through a direct aggregator like Coursera, LLM-based coursework.
The reason why this is a meaningful way to roll-out this type of coursework based LLM is because the initial surface area can actually be very sall (versus enrolling it within a school).
For example, a very popular course or one that is useful for anyone's career, could be rolled out for free to fine tune the content and training experience.
The ideal case would be it is initailly for free, and alumni clubs can then up charge as part fo their membership for in-person meet ups.
But this regular interface can also be one for on-going requests for alumni giving.
The challenge with most alumni giving is that these tend to be done in "drives" which do have their place.
But far more effectively would be on-going asks in context of activity, learning, assisting children to prepare for their college, or retirees finding new ways to engage with their alumni for advanced learning.
Once a well-crafted bespoke course is rolled-out to alumni, the model can be repeated for paid programs or to upsell into other alumni experiences which encourage higher levels of "giving" for lower-costs.
Those who are truly giving the big checks will not likely be influenced by this; what this is creating is a long-tail market for giving by a broader alumni base for whom small checks might not be a top priority.
However shifting to a product-led experience (could be one off, could be membership) creates an additional revenue stream that can serve as a pipeline for larger, traditional giving.