Not a subscriber? Buy a Day Pass or just this story using our new Digital Wallet
To subscribe, click here.
Sign up for our daily newsletter here
Welcome to “Illini Insider,” your regular dose of University of Illinois news from beat writer Luke Taylor. Fresh out of college himself, he’s always looking for story tips, photo ideas and social media mentions. Email him at ltaylor@news-gazette.com and he’ll give chase.
A few weeks ago, I wrote about Gies Business chief disruption officer Robert Brunner and his upcoming online class created with the help of generative artificial intelligence.
Brunner and I didn’t have a chance to talk before that column, but we’ve since been able to get in touch and dig a little deeper into the process behind creating that course and the future of AI as Brunner sees it.
“I’m a big believer that if you want to understand how things will affect the world, you have to try them,” Brunner told me.
That’s a call to action for his fellow professors, but also one reason he was willing to try using AI to put a class together.
He’s pretty confident that AI usage on that level isn’t about to become the new norm; he used it to generate a script from his course outlines as well as replicate his voice and image.
While it made editing processes easier and cut down on scripting time, training the programs and creating the audio and visual elements were new processes to the entire team, so it was still a lot of work.
Brunner was still in the thick of the project when we talked, so that may have been affecting his opinion, but he was starting to think it was actually more work than just recording the class in a traditional way would’ve been.
There are some potential benefits to creating classes this way: the fact that it’s a lot easier to translate the text and quickly record courses in other languages would make it possible to get these classes to people across the world.
Brunner said that’s a big deal for a land grant university especially.
“That would be so in line with our mission that it’s kind of exciting,” he said.
That’s where he sees AI changing things for professors, not for standard in-person instruction.
In general, Brunner expects it to take three to five years for generative AI technologies to really hit their stride and become truly prevalent in day-to-day life.
“In the short term, people overestimate what these technologies can do, but in the long term, they underestimate it,” Brunner said.
It’s like the iPhone: when it first dropped in 2007, the little touchscreen device was groundbreaking, but compare that to the smartphones of today.
That means we’re in a window to put up some “guardrails” on generative AI, Brunner said, and answer complicated questions.
“What does it mean to have an AI version of me that can be programmed to say things?” is Brunner’s immediate question.
Sure, the AI Brunner is providing helpful instruction now, but could someone take it and make it appear that the real Brunner is saying something insulting or offensive?
Could someone do that with any public figure with enough audio and visual content available?
(Short answer: they could, but it does take a lot of work and know-how to pull it off in a convincing way with current technology. To be fair, deepfakes have been around for a while already and caused many of the same fears when they first hit the scene.)
That could easily be considered a form of slander, but should written law expand to cover eventualities involving AI?
Those kinds of questions will take time to answer, but for now Brunner says it’s time to start learning how to use AI.
Like some other professors at UI, he assigns homework that involves ChatGPT usage, hoping to teach them how to use it well and elevate the results they turn in.
“It’s really been a multiplier in the skills I’m confident in students taking away from the course,” Brunner said.
At the same time, he’s watching the students experiment and figure out what exactly generative AI programs are capable of.
“I don’t know the best way to use ChatGPT,” Brunner said. “That’s some of the beauty of it.”
Brunner’s job as disruption officer is to shake things up, always looking for the next big thing.
He’s excited by the ways he can imagine a lot of the newest tech working together in the future – one example could be using AI to generate worlds to explore in an artificial reality headset.
That sounds simple, but AI generation could be big for video games; a big selling point of Bethesda’s latest release, Starfield, is that it uses procedural generation to create seemingly endless unique planets to explore.
Brunner’s looking to the future with hope, seeing science fiction become reality.
“We often get wrapped up in negativity, and there are real concerns there, but technology has done amazing things for people,” Brunner said. “There are a lot of problems in the world, but things are going to get better.”