Bryant Gumbel uttered the now-famous words nearly 30 years ago on the TODAY Show couch.

“What is the internet, anyway?” was part of a befuddled conversation on morning television about this new network of connected computers that was about to reshape the world. That 1994 conversation is now part of, well, internet lore.

Now, some 29 years later, you’re hearing similar conversations surrounding the advent of consumer-grade AI — artificial intelligence for the uninitiated.

“Artificial intelligence leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind,” IBM’s website reads.

The explosion of applications like ChatGPT has thrown many businesses into a new industrial revolution as new efficiencies could streamline workflow. That’s great for everyone but the humans who could be replaced by computers.

Few industries escape at least the conversation of AI disruption and college football is far from exempt.

And that’s where DJ Lee and his team at BYU enter the conversation.

Lee, the director of the Robotic Vision Laboratory in the Electrical and Computer Engineering Department at the Utah university is the unlikely driving force for what could mean changes to how college football programs operate.

What schools like Alabama and Georgia have on the competition are deep pockets and office space for a legion of analysts to break down film for as long as it takes to find tendencies and patterns within an opposing team’s film.

Lee and Co., are working on AI that could replace that human labor with machine-learning algorithms. The idea is to upload mountains (or terabytes) of film into a computer that would diagnose the formations and identify the tendencies of plays run from certain looks.

It’s third-and-4 at their own 35 in the third quarter on the road on a warm afternoon, what play is this team likely to run?

That’s some big-brain work right there and coaches are still trying to wrap their heads around it.

“To me, I’m just dealing with all the changes and all the things going on right now,” Kentucky football coach Mark Stoops said. “I mean, we’re holding onto our ass. So that’s like a whole ‘nother entity there that I’m not saying won’t come in very quickly. Obviously, with AI in itself, it’s exploding faster than we can all handle it. But for coaches, at least for myself in that area, there are so many different analytics and so many things to look at. In this department, I just haven’t got any grasp on it at all.”

Tennessee’s Josh Heupel is monitoring the situation.

“AI is infiltrating in so many different areas from business, I’m sure in sport, too,” he said at SEC media days. “Eventually it becomes a resource probably for everybody inside of college football. You know, how quickly that part takes inside of our program, I can’t give you an answer on. But I’m sure that at some point it will.”

And Missouri’s Eli Drinkwitz?

“I have no thoughts on AI in college football,” he said from the podium in Nashville last month.

Right, well LSU’s Brian Kelly does.

He sees the potential for the new technology to influence more than just game planning.

“We spend a lot of time on profiles, recruiting profiles and the intangibles or the traits that we’re looking for particular players,” Kelly said. “I think AI has an opportunity maybe to create that kind of model for us relative to who is that kind of player out there all over the country, without making a mistake. So I think that there’s going to be room for it, and I think it’s exciting and something that we’re about to venture into.”

That conversation hasn’t reached the Oxford, Mississippi office of Lane Kiffin, however.

“Well, I’m the first to say up here, when I know a lot about an area, I’m going to answer. If I don’t, I don’t,” Kiffin said at SEC media days. “The artificial intelligence and using that in recruiting, I don’t know about that. I can give you coach-speak and pretend that I do.”

For Lee and his team at BYU, the research is more about the Xs and Os.

Finding uniform game film was their initial logistical problem since they couldn’t identify each player before every snap. The solution to prove the concept was to use the Madden 2020 video game where they could achieve a drone-like view 30-40 degrees above the offensive formation.

“We selected five formation families and then for each family we have five different formations,” Lee said. “So total, we have like 25 formations that we collected a lot of videos for our experiment. It works well. We were able to get good results to recognize the formation.”

There are still plenty of kinks to work out with getting camera angles that show full fields of play that allow for unobstructed views of all 22 players on the field.

So what’s Lee’s big-picture goal here?

Time for back-office analysts to find a new career path?

Lee said he’s still working to get a program with whom to partner and gather real-world data. With it, he sees the ability to take 2D drawn play sheets of the past and turn them into dynamic models visible on tablets.

“I think that will be a more interesting training tool than just reading a static play chart,” Lee said.

Ultimately, Lee said he’d like to commercialize this technology.

“I would like for this tool to be available for teams who want to use it,” he said. “That’s my ultimate goal. In the end, this algorithm could be acquired by certain companies or a team. The immediate goal is to get videos and test it out on our algorithm to see how much data we can collect and then go to a football team.”

They’d for that to be BYU but he’s open to using it with anyone at the high school, college and/or professional teams.

Bottom line: It’s still up in the air but the possibilities are fascinating and could be revolutionary to a sport where the margins between success and failure are thin while the stakes are high.

Stoops in Lexington can see the allure of finding an edge.

“We’re all looking for any advantage,” he said. “I’m not saying it can’t. But it has to be somebody who is already there that’s working into the office who can dive into it because we’re all exploding with hiring people so to go dive into that area, it’s not something I’ve gotten into yet.”

So stay tuned.

Bookmark this story on the once-mysterious internet and, perhaps, a few of these coaches will have their Bryant Gumbel moment in retrospect.

Michael Casagrande is a reporter for the Alabama Media Group. Follow him on Twitter @ByCasagrande or on Facebook.