Campus Technology Insider Podcast January 2024
Rhea Kelly 10:10
Noble, how did those definitions resonate with you?
Noble Ackerson 10:13
I look at it, I look at AI, so AI to me, for the longest — I lost this battle — was a marketing term. It was a way for me, then, to explain how we transformed data. And so the way I explained it was, you know, as humans, we learn from past experience. You know, a baby's walking, falls down, hits their head, they know not to, you know, shuffle their feet past that rock. AI is essentially the same thing. Instead of past experience, it's data. It's past data. And so, or real-time data, which is essentially a snapshot of the past, depending on how you look at it. And so when we start abstracting what possibilities, or how you can use this data transformation, we start looking at different ways to teach a machine based on its, on our past experience. So AI is us. It is an inter, it's sarcastic in its way, it repeats what we do, it repeats our biases, as you as you rightly put it. AI is us. My daughter tries to fight Midjourney, a stable diffusion based generative AI, by asking for an anime or Pokemon that looks like her. She does not look Asian, she does not look white. However, I observe silently that when she types in "beautiful lady standing by a pole," it returns a list of options, she'll click on the option that looks closely to her definition of beautiful, at a given point in time. Sometimes it doesn't look like her. But what she doesn't realize is that by selecting that thing, she's fed, you know, something confirming that this is a beautiful lady, and then she gets upset when she's trying to make…. Actually, a funny story on that. She's trying to make a Pokemon that looks like her, so she wants a black Eevee. If you know Pokemon, you know what an Eevee is. And we got a warning. It's like, you know, offensive, da da da da. And so we said BIPOC, and then it worked. Right? So long, long and short is that AI is just a data transformation. You get data, you swirl it around, and data scientists say, "Did it generalize? Does it look like the real world? No." So if the data is dirty, if the data is biased, unfortunately, that's bias. That, that's how AI works.
Rhea Kelly 13:03
So this kind of goes to what Dave was saying about people conflating the concept of generative AI and AI. Because for the past year, we've heard a lot about AI in teaching and learning, and you know, questions about plagiarism or things like that. So can you all talk about other ways that AI is going to impact education, and kind of think beyond teaching and learning — administrative, IT? Dave, I thought I'd throw that one to you first.
David Weil 13:32
Sure. Thank you. So I think we as, you know, institutional leadership, when we think about AI, a lot of our attention for the past year has been on its impact in teaching and learning. That's all been all the headlines. It's been, you know, the, the anxiety on campus from faculty and students. But I think that's just the tip of the iceberg. I really think the, in some respects, that's interesting. And equally interesting is its impact on all the administrative functions at the institution. I think we're already seeing companies that are marketing to admissions teams for AI that will help them assess applications, that will read it, the applications, it'll score the applications. There's other companies that are marketing to philanthropy and engagement teams or our marketing communication teams for customized messaging that they would do for, to outreach to alumni or prospective students. I also think that, you know, if you look at the announcements from Microsoft for the Copilot, those will have a big impact, I think, on a number of the functions that people will be doing for financials. So I think, you know, really, there's an opportunity to look at the impact of this across the institution. One thing we're doing at Ithaca College is starting in December, my deputy CIO and I are meeting with every vice president. And we are going to have a customized presentation for them to show some examples from their space that we've read about, or we've heard about, just to start planting seeds to say, "These are some things we think you should be thinking about." We also will be talking about definitions, here, to give people a sense of that. And then we're sort of going to walk them through thinking about it in three categories: culture, workforce, and technology. So culture: What's the impact on their organization in terms of how they do their work and their norms there? Workforce: What are the skills that their staff will need? In the future, you know, a year from now, two years from now, what will this technology free up that will allow them to focus on other value-added propositions? And then the technology: What are the applications that they should be looking for? What's the data that those applications will need access to? And things like that. And so really, I really think that that's the next wave of conversations that needs to be happening on our campuses.