AI Changes Learning, Teaching and Research
"AI Changes Everything" is the tectonic shift across every industry.
In the same way that the Internet allowed new ways to live and work, so will AI.
Typically, higher education has been either insulated from or slow to address these changes. Some of it has to do with the seemingly limited market impacts (which I'll address in more detail on Higher Education - The Cost of Ivory Towerism in Market-Driven Economies.
For example, the rise of the Internet and YouTube as a means to provide similar instruction without paying hundreds of thousands of dollars in tuition has not yet eroded the business of major Universities. But the tide is shifting and may capitulate quickly rather than gradually.
The encroachment by Big Tech, such as Google, to provide education for free that can improve job prospects, the rise of cohort courses, and the more competitive job market is creating pressures on colleges that may not yet be felt by them. But applying Product principles, where college is the product, reveals weaknesses for traditional higher education that are being seized as opportunities by upstarts.
AI is going to continue to encroach on universities, and do so very quickly.
The educational experience, the instructional value, and perhaps even the research output is going to be changed.
The good thing is that, if higher education is able to capitalize on this as well as blockchain Blockchains give University's Superpowers, they actually can be in an important position socially and economically.
I'm a big fan, mostly, of universities based on their potential and form-factor: credible institutions with deep learning in collaborative environments fostering the next generation of creativity.
Unfortunately, much of the content, of late, has not been so appetizing, but I'm going to leave the politicizing out. When educational value is democratized, out in the open, socialized in a rigorous, critical manner, good things happen, and I think one of the artifacts of AI + blockchain for Universities when applied the right way matters.
Note: I think the implementation models probably make more impact than the initiative, and I think it's worth examining those as well. Execution is 99% - But It's Why Higher Ed Often Fails.
The number one reason higher ed must properly and strategically engage with AI is because they have content. Already, social media such as Reddit, Twitter, and Stackoverflow are trying to make their content core their AI strategy, whether to license it out or to use for themselves.
UGC has been the gift that keeps on giving to the platform owners -- they first monetized through advertising; now they are monetizing through AI licensing.
Higher education has an opportunity to repurpose the content in a way that achieves multiple goals -- but only if they have a comprehensive strategy and a cohesive way to execute.
What is odd (and perhaps endemic) was the primary narrative to come out of colleges: that students will use AI to cheat on their essays.
This feel like one of those compliance to retain the old guard versus looking at the opportunity to improve the learning experience.
It begs a few questions, such as:
- How are the writing and reviewing of essays a means of enabling or incentivizing learning?
- How can the assumption that generative AI will be used by the foundation for a better learning experience?
To me, answering this question is akin to how some universities (sadly, not enough) have leveraged video and YouTube to change the engagement model.
For example, many instructors and universities have believed that it was the lectures that was the product families have been paying for. Lectures as a learning modality is poor -- it's bad UX, it's bad retention, it's bad for innovation.
However, because universities treated the lectures as the product, many have not be serving their students and have been overcome by video-first models.
Some universities are seeing lectures as something that is best done asynchronously. Even better, can be given away for free.
By providing lecture asynchronously, self-directed students are more likely and able to schedule their own ingestion periods which could aid in better learning. For example, some may want to stop and pause to review sections, take notes. Some might like to listen multiple times at 2x while lifting weights or running. Whatever.
The lectures aren't the product. The more delivery mechanisms to meet different preferences the better.
What is the product?
Critical thinking, collaboration, and engagement with other students and the professor. When this is the product, it's much easier to justify the high price points.
The best experiences are likely with professors who can engage their students in a seminar format, and there are different ways to structure this. Far better than the lectures for those who care about their students.
I recognize this would be the bane of existence for professors who just update their notes once a year and dial it in from the lectern.
And this is where I think the execution models of Higher Education constraint themselves.
It's akin, but actually worse, than Kodak holding onto to traditional film-based cameras and turning a blind eye to digital.
It's actually why I think in terms of a strategy, the Product mindset is critical, versus the Political mindset: who is the primary Customer?
For a good number of Universities, the primary Customer are the students, the ones coming to study, learn and develop their minds.
I recognize that there are other Stakeholders, such as the Professors themselves and the corporations that provide research grants. Those should be considerations as well.
But, I would argue, that great student experiences contribute to strong alumni relations which contribute to alumni giving and can be a basis for endowments and grants.
Great student experiences contribute often to word of mouth growth, enrollment, even rankings (for those universities that care about them).
My point is: technology shifts can be viewed only as threats when the old ways are retained for their own sake rather than reframing first on the Customer: what benefits students.
So back to AI: if a student cares so little about their educational experience (and I believe there probably are quite a bit of those who get a free ride from their parents) then they will dial it in with generative AI. But that's not any different from them going to a website and asking someone else to write the paper for them.
Removing the in-person lecture environment in favor of smaller, more interactive groups with professors and graduate students doesn't enable the unmotivated student from choosing to not attend an in-person lecture.
To me, it is far better to design for the happy path -- the motivated student -- and design around a better instructional experience, one that makes the student feel they got more than their money's worth.
To properly tackle the challenge and promise of AI means resetting the goals and fixing eyes on the prize: students who learn.
Once we have that perspective, the design space opens up, and I'd love to hear from you what you think those could be.
Here are some spitball ideas (but I think following Product principles could be a real unlock):
- Generative AI that provides an instructional experience -- asking questions, reviewing answers -- based on a given lecture. Think of it as the personal tutor...even better if capturing the star lecturer's tone, voice, and humor!
- Generative AI that responds to student writing (writing is important), and is also able to detect and challenge writing that was entirely unoriginal and copied and pasted; instead, it provides critical thinking to ask questions about the quality of the writing.
- AI which can help foster real time discussion and collaboration with the professor (or perhaps on occasion, the agenic version of the professor)
On the topic of student writing: I wish I could write as well as many college students write on Twitter or on their own blog.
I think that this social model of writing content makes students better writers, thinkers, and more importantly helps to launch them into the world of ideas.
Are there risks?
Yes, and I think there should be a way to protect students from bad ideas haunting them for the rest of their life. After all, college should be a time for them to explore bad ideas, even things that can potentially seem horrific twenty years later when they are candidates for the Supreme Court.
There are things that can enable both a form of social discourse and self-sovereignty over the content in a world where everything lives forever on the Internet.
But that's a digression worth discussing later.
Right now, the focus is on AI to make a much better learning experience, and I think this can be done with the right framing.